1
00:00:06,050 --> 00:00:08,650
Hello, and welcome to Drivechain Twitter Spaces.

2
00:00:08,650 --> 00:00:10,390
Today is Friday, July 28th.

3
00:00:12,810 --> 00:00:15,110
Here at LayerTwo Labs, we're thrilled

4
00:00:15,110 --> 00:00:20,110
to have one of our favorite Bitcoiners

5
00:00:20,390 --> 00:00:24,429
as our featured guest today, Dr. Adam Back,

6
00:00:25,670 --> 00:00:30,550
co-founder and chief executive officer of Blockstream,

7
00:00:30,550 --> 00:00:32,710
one of the coolest people to ever live.

8
00:00:33,710 --> 00:00:38,710
Dr. Back is a famous Bitcoiner for many reasons,

9
00:00:39,169 --> 00:00:42,210
among them that he is the only person

10
00:00:42,210 --> 00:00:45,050
whose name appears in the body of the Bitcoin whitepaper.

11
00:00:47,410 --> 00:00:48,950
Welcome to the stage, Paul.

12
00:00:51,509 --> 00:00:53,989
We're just opening up the space right now.

13
00:00:55,469 --> 00:00:56,309
Hey.

14
00:00:58,790 --> 00:00:59,630
Hey.

15
00:01:03,710 --> 00:01:04,550
Hey.

16
00:01:06,950 --> 00:01:09,210
I see some familiar faces in the audience.

17
00:01:10,350 --> 00:01:13,249
It is so awesome to have so many people

18
00:01:13,249 --> 00:01:17,530
in our Telegram group and tweeting about Drivechain lately.

19
00:01:17,530 --> 00:01:21,810
We have over 700 members in the Drivechain Telegram group,

20
00:01:21,810 --> 00:01:24,609
and ah, Dr. Back is here.

21
00:01:25,689 --> 00:01:30,230
Why don't I invite him up, our featured speaker today.

22
00:01:32,710 --> 00:01:37,710
Welcome, Adam.

23
00:01:40,810 --> 00:01:42,609
Thank you for being our guest today.

24
00:01:52,210 --> 00:01:55,910
Adam, try speaking to make sure that we're connected.

25
00:01:59,289 --> 00:02:00,410
I don't hear you.

26
00:02:00,410 --> 00:02:02,289
Paul, do you hear him?

27
00:02:02,829 --> 00:02:04,469
Yeah, I don't hear you either, unfortunately.

28
00:02:04,469 --> 00:02:05,829
Oh.

29
00:02:05,829 --> 00:02:08,069
Oh, wait, I saw blue lines appear that time, though.

30
00:02:08,069 --> 00:02:09,469
Yes, me too.

31
00:02:10,590 --> 00:02:12,710
Well, you know, people like to get the special headphones

32
00:02:12,710 --> 00:02:13,550
or whatever to plug in.

33
00:02:13,550 --> 00:02:16,270
Yes, yeah, take your time getting set up.

34
00:02:16,270 --> 00:02:17,749
We can't hear you yet, Adam.

35
00:02:25,589 --> 00:02:30,589
I have really enjoyed Adam's statements on sidechains.

36
00:02:31,010 --> 00:02:34,570
The Zcash sidechain, the Thunder large block sidechain,

37
00:02:34,570 --> 00:02:37,550
and others, his tweets and his comments

38
00:02:37,550 --> 00:02:39,249
at the conferences are so awesome.

39
00:02:41,089 --> 00:02:43,589
Adam, try speaking now to see if you're connected.

40
00:02:47,089 --> 00:02:49,169
All right, how about now?

41
00:02:49,169 --> 00:02:51,249
Yes, awesome.

42
00:02:51,249 --> 00:02:55,189
All right, Paul, take it away, if you'd like.

43
00:02:56,130 --> 00:02:56,969
Yeah, sure, sure.

44
00:02:56,969 --> 00:03:00,210
Yeah, well, I mean, basically one thought in my head

45
00:03:01,210 --> 00:03:06,210
is you've been a very helpful supporter,

46
00:03:07,309 --> 00:03:08,870
even for a very long time.

47
00:03:08,870 --> 00:03:13,870
I remember Consensus 2017, we had a table.

48
00:03:13,870 --> 00:03:17,050
You organized a little table and an email chain,

49
00:03:17,050 --> 00:03:21,210
and then there were also some very supportive emails

50
00:03:21,210 --> 00:03:26,029
over the years about certain large blocker projects

51
00:03:26,029 --> 00:03:28,910
should instead focus on this or whatever.

52
00:03:28,910 --> 00:03:31,050
So I guess the main thought in my head

53
00:03:31,050 --> 00:03:35,050
is you obviously sort of have gotten it,

54
00:03:35,050 --> 00:03:39,570
or at least you saw some kind of potential

55
00:03:39,570 --> 00:03:42,010
in this idea for a really long time.

56
00:03:43,169 --> 00:03:45,930
And so the sort of funny question would be,

57
00:03:45,930 --> 00:03:48,609
why don't all these other people get it, or whatever?

58
00:03:48,609 --> 00:03:52,850
Something like that, or I don't know if that's a good,

59
00:03:52,850 --> 00:03:54,930
but I just don't.

60
00:03:54,930 --> 00:03:56,410
How do you explain the difference, I guess,

61
00:03:56,469 --> 00:03:59,949
between all these different reactions to it or something?

62
00:03:59,949 --> 00:04:04,949
I mean, we had a conversation on a panel

63
00:04:06,229 --> 00:04:11,229
at Bitcoin Amsterdam, I think it was last year, right?

64
00:04:12,350 --> 00:04:13,309
There was a panel.

65
00:04:13,309 --> 00:04:18,309
And one of the concepts, I think,

66
00:04:18,749 --> 00:04:23,750
is that Bitcoin is good at adapting-

67
00:04:24,729 --> 00:04:26,409
Is anyone able to hear Adam?

68
00:04:27,609 --> 00:04:29,210
Yes, I hear him.

69
00:04:29,210 --> 00:04:31,650
Okay, good, still working, yeah.

70
00:04:31,650 --> 00:04:36,650
So I think Bitcoin is better at adopting changes

71
00:04:36,930 --> 00:04:41,690
that are kind of simple, localized, win-win,

72
00:04:46,329 --> 00:04:49,530
but for something which has a lot of trade-offs

73
00:04:49,590 --> 00:04:54,389
or a sort of more complicated game theory,

74
00:04:54,389 --> 00:04:59,030
or more open-ended, it tends to go

75
00:04:59,030 --> 00:05:00,590
into a big discussion topic

76
00:05:00,590 --> 00:05:03,150
about whether it's a good trade-off

77
00:05:03,150 --> 00:05:05,009
or different ways to do it.

78
00:05:07,190 --> 00:05:11,030
I think Bitcoin is generally a little bit resistant

79
00:05:11,030 --> 00:05:11,870
to change, right?

80
00:05:11,870 --> 00:05:13,990
Because people are defensive of their Bitcoin.

81
00:05:13,990 --> 00:05:18,150
So I think that made something like small signatures,

82
00:05:18,150 --> 00:05:22,409
which was a topic on that panel, easier to adopt

83
00:05:22,409 --> 00:05:24,930
because it's just a better signature, right?

84
00:05:24,930 --> 00:05:28,210
So there's not really that much to say

85
00:05:28,210 --> 00:05:30,289
about why somebody wouldn't want it.

86
00:05:30,289 --> 00:05:33,009
Like, you don't have to use it if you don't like it.

87
00:05:33,009 --> 00:05:34,930
It's a better signature, it's more compact,

88
00:05:34,930 --> 00:05:36,849
it's a little bit better for privacy, and so on.

89
00:05:36,849 --> 00:05:39,650
And so, after a number of years,

90
00:05:39,650 --> 00:05:43,449
that was added to Bitcoin with Taproot.

91
00:05:43,449 --> 00:05:46,509
And drivechain discussion, sorry,

92
00:05:46,509 --> 00:05:49,750
drivechain, sidechain discussion has been around

93
00:05:49,750 --> 00:05:51,949
probably for a similar amount of time, actually,

94
00:05:51,949 --> 00:05:53,310
to small signatures.

95
00:05:53,310 --> 00:05:56,229
It was talked about in, like, maybe 2013.

96
00:05:58,789 --> 00:06:01,889
But I think it's more complicated.

97
00:06:04,270 --> 00:06:07,310
Now, the reason I originally got interested

98
00:06:07,310 --> 00:06:11,829
in sidechains and drivechains was for modularity.

99
00:06:11,829 --> 00:06:16,210
So, you know, if you think about other systems,

100
00:06:16,210 --> 00:06:18,690
like the Linux kernel or something,

101
00:06:20,449 --> 00:06:21,569
it's a monolithic kernel,

102
00:06:21,569 --> 00:06:24,770
so getting changes into it is complicated.

103
00:06:26,789 --> 00:06:31,009
Whereas when they added sort of user-land modules

104
00:06:31,009 --> 00:06:33,009
and kernel modules and things,

105
00:06:33,009 --> 00:06:35,770
it became more permissionless, I guess you could,

106
00:06:35,770 --> 00:06:37,490
you know, you could implement one yourself

107
00:06:37,490 --> 00:06:38,329
or something like that.

108
00:06:38,349 --> 00:06:42,750
So, it enables this kind of more permissionless innovation

109
00:06:42,750 --> 00:06:44,349
that you get on layer twos,

110
00:06:44,349 --> 00:06:46,590
like Lightning or Arc or things like that.

111
00:06:46,590 --> 00:06:49,990
So, I was just thinking, you know,

112
00:06:49,990 --> 00:06:53,789
it's a pity Bitcoin's not more modular in that way,

113
00:06:53,789 --> 00:06:55,470
and how could you achieve that?

114
00:06:55,470 --> 00:07:00,050
And so, the sidechain idea was, like, a potential,

115
00:07:01,829 --> 00:07:06,310
because the, I guess, the main concern

116
00:07:06,310 --> 00:07:11,310
with adding consensus logic, like with an extension block,

117
00:07:11,930 --> 00:07:15,810
is that, which is a different kind of way

118
00:07:15,810 --> 00:07:18,569
to extend Bitcoin, hypothetically.

119
00:07:18,569 --> 00:07:21,870
If somebody accidentally makes a bug in an extension block,

120
00:07:23,770 --> 00:07:26,609
it could, like, split the network or fork the network

121
00:07:26,609 --> 00:07:28,329
or create problems for Bitcoin itself,

122
00:07:28,329 --> 00:07:31,609
whereas a sidechain or a drivechain is, you know,

123
00:07:31,609 --> 00:07:32,770
it's its own little world,

124
00:07:32,770 --> 00:07:34,090
and something goes wrong in there,

125
00:07:34,090 --> 00:07:36,470
has little impact on Bitcoin itself.

126
00:07:36,470 --> 00:07:40,470
So, yeah, no, pause there.

127
00:07:42,430 --> 00:07:47,150
Yeah, like, so, the idea of it being open-ended,

128
00:07:47,150 --> 00:07:48,550
I think, is where it really goes,

129
00:07:48,550 --> 00:07:51,909
where people's creativity kind of runs wild or something.

130
00:07:51,909 --> 00:07:56,009
They just think, like, anything is possible,

131
00:07:56,009 --> 00:07:59,150
whereas with something narrow, it's because it's not as,

132
00:08:00,270 --> 00:08:01,430
it's because it's not as cool

133
00:08:01,569 --> 00:08:05,409
and maybe it's not as controversial or something like that.

134
00:08:05,409 --> 00:08:07,490
Well, I mean, I think you can see it recently

135
00:08:07,490 --> 00:08:09,770
in the discussion about covenants,

136
00:08:09,770 --> 00:08:14,289
because there are lots of different proposals

137
00:08:14,289 --> 00:08:16,990
and say, you know,

138
00:08:19,409 --> 00:08:23,490
maybe there's a reasonable amount of interest in covenants,

139
00:08:23,490 --> 00:08:27,849
but there are different ways to do it,

140
00:08:28,690 --> 00:08:31,210
and so there's a big design space.

141
00:08:31,990 --> 00:08:34,629
And then I guess there's a sort of feeling of scarcity,

142
00:08:34,629 --> 00:08:37,349
like you can't add, you know,

143
00:08:37,349 --> 00:08:40,349
it would be odd to add half a dozen different ways to do it,

144
00:08:40,349 --> 00:08:42,550
just throw them all in a bucket, right?

145
00:08:42,550 --> 00:08:45,269
So I think people expect, you know,

146
00:08:45,269 --> 00:08:48,310
if you think about, like, a line of CPUs,

147
00:08:48,310 --> 00:08:50,870
like the Intel CPU and cross-generations,

148
00:08:50,870 --> 00:08:53,269
maybe they add a few CPU instructions,

149
00:08:53,269 --> 00:08:55,149
and presumably they have,

150
00:08:55,149 --> 00:08:56,310
they don't want to add too many instructions,

151
00:08:56,310 --> 00:08:57,349
so they're thinking real carefully

152
00:08:57,370 --> 00:09:01,810
about which instructions can be combined

153
00:09:01,810 --> 00:09:04,450
with other instructions, provide good value,

154
00:09:04,450 --> 00:09:09,450
and, you know, not sort of diverge massively

155
00:09:09,649 --> 00:09:11,490
from the program model of the CPU,

156
00:09:11,490 --> 00:09:14,070
so I think Bitcoin opcodes are a bit like that.

157
00:09:15,729 --> 00:09:18,750
Maybe that's another way that the drivechain

158
00:09:18,750 --> 00:09:20,050
or sidechain is different.

159
00:09:20,050 --> 00:09:24,729
It's sort of, when you're actually gonna implement it

160
00:09:24,750 --> 00:09:28,109
using opcodes, and maybe the opcode doesn't do it

161
00:09:28,109 --> 00:09:33,109
in one shot, but it enables you to build a sidechain

162
00:09:33,229 --> 00:09:37,269
using some existing operations, operators,

163
00:09:37,269 --> 00:09:40,950
and, or opcodes, and the new ones,

164
00:09:40,950 --> 00:09:45,950
but it feels like a kind of a higher-level concept, right?

165
00:09:46,750 --> 00:09:48,310
Like a low-level opcode is like, okay,

166
00:09:48,310 --> 00:09:52,409
verify Schnorr signature instead of an ECTA signature,

167
00:09:52,409 --> 00:09:55,269
and then, you know, the rest of it works similarly

168
00:09:55,269 --> 00:09:57,390
to how it works now, whereas the drivechain

169
00:09:57,390 --> 00:09:59,870
is a kind of meta opcode, or it seems like

170
00:09:59,870 --> 00:10:02,190
a higher-level programming concept, right?

171
00:10:02,190 --> 00:10:05,070
So maybe that's, it's a more novel thing

172
00:10:05,070 --> 00:10:06,950
for people to get their head around

173
00:10:08,390 --> 00:10:09,669
implementing an opcode.

174
00:10:11,789 --> 00:10:13,130
There's certainly much more novelty

175
00:10:13,130 --> 00:10:14,310
than I would have thought.

176
00:10:14,310 --> 00:10:17,149
Like, to me, it was not that complicated of an idea.

177
00:10:17,149 --> 00:10:21,269
Back in November 2015, it was just like

178
00:10:21,269 --> 00:10:24,289
the SPV proof was already established

179
00:10:24,289 --> 00:10:28,349
by the Blockstream October 2014 paper,

180
00:10:28,349 --> 00:10:30,450
so it was kind of like, it was already established,

181
00:10:30,450 --> 00:10:33,409
I thought, throughout the year 2014,

182
00:10:33,409 --> 00:10:36,349
that the sidechain would, like, prove the work,

183
00:10:37,390 --> 00:10:39,930
but it would not prove the validity rules,

184
00:10:41,050 --> 00:10:43,370
and then it just seemed pretty straightforward

185
00:10:43,370 --> 00:10:45,209
that you could just, you'd have a counter.

186
00:10:45,209 --> 00:10:48,870
It was basically, instead of having the sidechain headers,

187
00:10:48,870 --> 00:10:50,409
it was just, on the main chain, you just have, like,

188
00:10:50,409 --> 00:10:54,029
a counter, like someone'd count up with a little clicker,

189
00:10:54,029 --> 00:10:58,190
count up to some number of uploads or something,

190
00:10:59,149 --> 00:11:01,430
and it actually seemed like it was,

191
00:11:01,430 --> 00:11:03,430
when it first was proposed,

192
00:11:03,430 --> 00:11:05,930
it made it to the top of rBitcoin,

193
00:11:05,930 --> 00:11:08,529
and then it, a lot of people seemed to be

194
00:11:08,529 --> 00:11:10,990
very, very positive towards the idea,

195
00:11:10,990 --> 00:11:14,089
and even giving it attention and stuff,

196
00:11:14,089 --> 00:11:17,390
and the way I tell the story is sort of like

197
00:11:17,409 --> 00:11:21,490
the block size war immediately made,

198
00:11:21,490 --> 00:11:22,890
first of all, anything involving miners

199
00:11:22,890 --> 00:11:25,169
became, like, much more controversial,

200
00:11:25,169 --> 00:11:27,729
but just, like, the idea of any large blockers

201
00:11:27,729 --> 00:11:30,729
getting what they want, or any talk

202
00:11:30,729 --> 00:11:34,010
as if Bitcoin Core weren't perfect,

203
00:11:34,010 --> 00:11:35,669
so maybe this is me just reading

204
00:11:35,669 --> 00:11:39,010
a bunch of emotional baggage into it,

205
00:11:39,010 --> 00:11:41,169
but this is, like, kind of like,

206
00:11:41,169 --> 00:11:45,690
it seemed like pre, it seemed like before

207
00:11:45,709 --> 00:11:49,510
the block size, like, before SegWit2x,

208
00:11:49,510 --> 00:11:52,190
it actually seemed like there was much more support

209
00:11:52,190 --> 00:11:55,469
for drivechain among literally everyone,

210
00:11:55,469 --> 00:11:58,789
and then the sort of SegWit2x Bitcoin Cash happened,

211
00:11:58,789 --> 00:12:00,149
and then the last thing that people,

212
00:12:00,149 --> 00:12:04,289
once it's something we brought up in the Amsterdam thing

213
00:12:04,289 --> 00:12:06,630
is that if someone wants something with Bitcoin,

214
00:12:06,630 --> 00:12:08,469
and you say, oh, we can do that as a sidechain,

215
00:12:08,469 --> 00:12:11,149
do a zk-SNARK sidechain, but then,

216
00:12:11,149 --> 00:12:13,310
as soon as they leave, and then do zcash,

217
00:12:13,330 --> 00:12:17,529
or they do bch, then, like, they don't really want,

218
00:12:17,529 --> 00:12:19,330
the last thing they want is for Bitcoin to add it now

219
00:12:19,330 --> 00:12:22,690
that they have their own thing that they think is the best,

220
00:12:22,690 --> 00:12:25,890
so I kind of thought it was going better before,

221
00:12:25,890 --> 00:12:27,649
and then Bitcoin Cash split,

222
00:12:27,649 --> 00:12:30,450
and that kind of distracted everyone for a while, I think,

223
00:12:30,450 --> 00:12:32,810
including me, and I was just kind of like,

224
00:12:32,810 --> 00:12:35,010
but this is my weird interpretation

225
00:12:35,010 --> 00:12:39,450
of how people allocate their scarce,

226
00:12:39,450 --> 00:12:41,310
because everyone's review time is very scarce,

227
00:12:41,310 --> 00:12:43,130
like, just even a regular person,

228
00:12:43,830 --> 00:12:45,310
it's just a Bitcoin user,

229
00:12:45,310 --> 00:12:46,789
but I don't know how people look at it now

230
00:12:46,789 --> 00:12:50,130
if they joined recently, they've been around for so long.

231
00:12:54,149 --> 00:12:57,870
Yeah, I mean, I think there are obviously new contributors

232
00:12:57,870 --> 00:13:02,870
who didn't, you know, weren't paying attention

233
00:13:03,550 --> 00:13:08,550
to technical stuff through the block size drama,

234
00:13:08,649 --> 00:13:13,169
but I think you're right that that was,

235
00:13:14,130 --> 00:13:18,050
you know, some kind of mixture of dramatic,

236
00:13:18,050 --> 00:13:21,169
traumatic, and exhilarating, and profitable,

237
00:13:21,169 --> 00:13:22,690
depending on what you did,

238
00:13:22,690 --> 00:13:26,969
and so I think people were a bit hesitant

239
00:13:26,969 --> 00:13:30,690
about, you know, how the next soft fork would go, right?

240
00:13:30,690 --> 00:13:33,050
They were hoping it wouldn't be,

241
00:13:33,050 --> 00:13:34,370
it wouldn't, you know, there wouldn't be

242
00:13:34,370 --> 00:13:38,430
another dramatic event like that activation process, right?

243
00:13:39,370 --> 00:13:42,250
And then, yeah.

244
00:13:42,250 --> 00:13:45,130
Well, yeah, one of the many things it did, of course,

245
00:13:45,130 --> 00:13:48,969
is that it made everyone, like, basically afraid

246
00:13:48,969 --> 00:13:52,849
to do any soft fork like in the usual historical way,

247
00:13:52,849 --> 00:13:55,570
where it was really not a big deal at all before.

248
00:13:57,849 --> 00:14:02,010
And then after SegWit was so weird with the blockade,

249
00:14:02,010 --> 00:14:03,609
and then user-activated soft fork,

250
00:14:03,609 --> 00:14:06,769
and all the questions that entailed as to, like,

251
00:14:06,789 --> 00:14:08,789
who should ultimately be trumping what,

252
00:14:09,789 --> 00:14:13,550
and why, those weird governance questions,

253
00:14:13,550 --> 00:14:17,269
I think that was, that was huge, I think.

254
00:14:17,269 --> 00:14:19,990
That was bad for every single soft fork.

255
00:14:19,990 --> 00:14:22,789
I mean, to this day, basically, like, now, people,

256
00:14:22,789 --> 00:14:25,669
isn't it intriguing that, like, so many pull requests

257
00:14:25,669 --> 00:14:28,149
could be merged into Bitcoin Core every day,

258
00:14:28,149 --> 00:14:30,669
and any of them could have, like, an assert zero,

259
00:14:30,669 --> 00:14:33,750
or some kind of, like, key logger, or some kind of thing?

260
00:14:33,769 --> 00:14:36,990
Any pull request could have something really bad in it.

261
00:14:38,089 --> 00:14:40,289
But many of them fail right through,

262
00:14:40,289 --> 00:14:42,729
and not that they fail right through, but it's like,

263
00:14:42,729 --> 00:14:44,849
if one of these pull requests is a soft fork,

264
00:14:44,849 --> 00:14:49,849
then, like, treated with an enormous amount of anxiety.

265
00:14:52,130 --> 00:14:56,130
Well, right, I mean, yeah, the rate of changes,

266
00:14:56,130 --> 00:15:00,890
in terms of the amount of check-ins per major version

267
00:15:00,909 --> 00:15:03,990
is kind of large.

268
00:15:03,990 --> 00:15:05,390
It's a fairly active project, right?

269
00:15:05,390 --> 00:15:10,390
So, yeah, I mean, that is, you know,

270
00:15:10,510 --> 00:15:13,070
you want to feel like it's being reviewed very carefully,

271
00:15:13,070 --> 00:15:16,830
and of course, there's a lot of lessons learned

272
00:15:17,789 --> 00:15:18,950
since the early days of Bitcoin,

273
00:15:18,950 --> 00:15:23,669
and there is a pretty cautious verification of everything,

274
00:15:23,669 --> 00:15:27,709
but consensus logic is much riskier,

275
00:15:27,729 --> 00:15:32,729
and, I guess, more subtle and possible to make stakes on.

276
00:15:33,690 --> 00:15:35,930
I mean, I think even, you know,

277
00:15:35,930 --> 00:15:37,609
there have been small issues,

278
00:15:37,609 --> 00:15:42,609
or issues that had to be worked around with a fix

279
00:15:43,050 --> 00:15:45,890
that were not seen at the time they were added, right,

280
00:15:45,890 --> 00:15:47,450
over the last five years.

281
00:15:47,450 --> 00:15:52,169
But, yeah, I mean, so I think, like, in a way,

282
00:15:52,169 --> 00:15:56,450
the Schnorr and Taproot were, like,

283
00:15:56,450 --> 00:16:00,630
a super uncontroversial, localized, you know,

284
00:16:00,630 --> 00:16:02,070
opt-in upgrade to a signature,

285
00:16:02,070 --> 00:16:05,870
so that was a kind of arguably a good

286
00:16:08,510 --> 00:16:11,149
sort of type of feature to see

287
00:16:12,070 --> 00:16:15,109
how activation would be viewed,

288
00:16:15,109 --> 00:16:18,550
and it was, like, largely, you know,

289
00:16:18,550 --> 00:16:21,109
uncontroversial, and it activated.

290
00:16:22,709 --> 00:16:25,870
I mean, the speedy trial thing,

291
00:16:26,130 --> 00:16:28,010
that had a little bit of controversy in it, too,

292
00:16:28,010 --> 00:16:30,409
but ultimately, it progressed.

293
00:16:32,649 --> 00:16:33,969
I mean, I think the other thing

294
00:16:33,969 --> 00:16:38,469
which might be a bit different is,

295
00:16:40,089 --> 00:16:45,089
you know, the security of a signature

296
00:16:45,890 --> 00:16:47,649
is very binary, right?

297
00:16:47,649 --> 00:16:49,450
You know, as long as you don't lose control

298
00:16:49,450 --> 00:16:51,430
of the private keys, and you don't make mistakes,

299
00:16:51,430 --> 00:16:54,630
and, you know, serialization, and what have you,

300
00:16:54,630 --> 00:16:58,250
it's extremely robust, and secure,

301
00:16:58,250 --> 00:17:01,609
and has a huge security advantage for the defender,

302
00:17:01,609 --> 00:17:02,890
only the point owner,

303
00:17:04,570 --> 00:17:07,570
whereas a drivechain or a sidechain

304
00:17:07,570 --> 00:17:10,090
is a more fuzzy kind of game theory thing,

305
00:17:10,090 --> 00:17:12,409
where you're kind of, you know,

306
00:17:12,409 --> 00:17:15,889
trying to think about different adversarial scenarios

307
00:17:15,889 --> 00:17:16,970
with miners and things,

308
00:17:16,970 --> 00:17:21,970
so I think that that causes people more trepidation,

309
00:17:22,350 --> 00:17:27,350
because there's not a sort of very simple and direct,

310
00:17:28,869 --> 00:17:30,989
you know, this is a drop-in replacement,

311
00:17:30,989 --> 00:17:32,830
it's a signature, you know,

312
00:17:32,830 --> 00:17:35,470
what's not to understand kind of thing, right?

313
00:17:35,470 --> 00:17:39,950
Whereas a drivechain is like a new piece of game theory

314
00:17:39,950 --> 00:17:43,509
about, you know, incentives,

315
00:17:43,509 --> 00:17:46,350
and miner behavior, and so on.

316
00:17:46,350 --> 00:17:50,310
So, you know, I think it's being discussed a lot,

317
00:17:50,330 --> 00:17:53,409
and it probably works,

318
00:17:53,409 --> 00:17:56,570
but it's not as binary,

319
00:17:56,570 --> 00:18:01,570
and, like, convincingly, robustly secure, let's say, right?

320
00:18:02,850 --> 00:18:05,330
Yeah, it's a more complicated thing.

321
00:18:05,330 --> 00:18:06,930
I completely agree with that,

322
00:18:06,930 --> 00:18:09,730
in that it is more of a game theory thing,

323
00:18:09,730 --> 00:18:12,230
but I think this leads to another,

324
00:18:13,649 --> 00:18:15,310
this is another weird thing that has, like,

325
00:18:15,310 --> 00:18:19,970
kind of dogged or plagued this idea for a long time,

326
00:18:20,669 --> 00:18:21,509
which is this whole,

327
00:18:21,509 --> 00:18:23,629
people are kind of mixing up layer one and layer two,

328
00:18:23,629 --> 00:18:25,230
like, you know, like,

329
00:18:25,230 --> 00:18:30,190
lightning is also much more clear-cut than drivechain,

330
00:18:30,190 --> 00:18:31,909
but it's like,

331
00:18:31,909 --> 00:18:33,789
it should be that,

332
00:18:33,789 --> 00:18:37,830
you would think it's not anyone's concern on L1

333
00:18:39,430 --> 00:18:43,590
how insecure an L2 might be,

334
00:18:43,590 --> 00:18:45,230
because you don't have to use it.

335
00:18:46,310 --> 00:18:47,149
So I think that there's,

336
00:18:47,369 --> 00:18:48,210
I mean,

337
00:18:48,210 --> 00:18:50,889
people take responsibility for it.

338
00:18:50,889 --> 00:18:54,090
They take, you know, the Bitcoin Core developer

339
00:18:54,090 --> 00:18:56,970
might feel responsible for the coins,

340
00:18:56,970 --> 00:18:57,810
and they might think,

341
00:18:57,810 --> 00:19:01,409
oh, we're gonna let this person use a drivechain,

342
00:19:01,409 --> 00:19:04,409
and then maybe the coins will be lost

343
00:19:04,409 --> 00:19:08,570
by the withdrawal, the miner withdrawal process.

344
00:19:08,570 --> 00:19:10,810
The coins may be lost,

345
00:19:10,810 --> 00:19:12,330
but it's kind of like,

346
00:19:12,330 --> 00:19:14,169
shouldn't there be a little bit more,

347
00:19:14,169 --> 00:19:16,570
shouldn't there be a little bit less paternalism?

348
00:19:18,090 --> 00:19:19,789
Well, maybe.

349
00:19:19,789 --> 00:19:22,169
I mean, on lightning,

350
00:19:22,169 --> 00:19:23,129
that's kind of the case,

351
00:19:23,129 --> 00:19:24,409
and actually,

352
00:19:24,409 --> 00:19:27,610
I think the security model of lightning

353
00:19:28,529 --> 00:19:33,129
had a few surprises along the way, right?

354
00:19:33,129 --> 00:19:36,009
There were issues that were discovered,

355
00:19:36,009 --> 00:19:38,210
you know, years after the first version was coded,

356
00:19:38,210 --> 00:19:39,649
and we had to, like,

357
00:19:39,649 --> 00:19:41,970
make workarounds and fix that up

358
00:19:41,970 --> 00:19:42,810
and change the clients,

359
00:19:42,810 --> 00:19:44,450
change the protocols a bit.

360
00:19:44,470 --> 00:19:47,310
So it wasn't without its surprises,

361
00:19:47,310 --> 00:19:49,710
but I guess,

362
00:19:49,710 --> 00:19:51,789
you know, it was opt-in,

363
00:19:51,789 --> 00:19:55,110
and once you have the opcodes to enable it,

364
00:19:56,149 --> 00:19:57,190
and I guess you could have done it

365
00:19:57,190 --> 00:19:59,450
without any opcodes in a simple way,

366
00:20:00,989 --> 00:20:01,869
you couldn't stop it.

367
00:20:01,869 --> 00:20:02,950
So I think, you know,

368
00:20:02,950 --> 00:20:05,149
surprisingly, the fact that you can't stop something

369
00:20:05,149 --> 00:20:07,509
makes people pretty relaxed about it,

370
00:20:07,509 --> 00:20:09,470
you know, if you could build it anyway.

371
00:20:09,388 --> 00:20:17,148
I've seen that in, like, the C language evolution. There was, you read the history of the

372
00:20:17,148 --> 00:20:22,428
concatenate operator. The C preprocessor didn't used to have it, and there was a, apparently,

373
00:20:22,428 --> 00:20:26,828
I was reading about the history of it, there was a big debate about, you know, would it be a

374
00:20:26,828 --> 00:20:31,628
sensible thing to add this concatenate operator, and people, some people didn't like it, some people

375
00:20:31,628 --> 00:20:37,468
did, and then finally, somebody proved that with some enormous amount of hackery, you could, like,

376
00:20:37,468 --> 00:20:42,188
make your own using the existing preprocessor stuff. Some people shrugged, it's like,

377
00:20:42,188 --> 00:20:47,468
oh, you can already do it, so they added the operator, right? So, yeah, that's a curious

378
00:20:47,468 --> 00:20:53,628
effect, right? So it makes me think that, yeah, go ahead. You and I both know that,

379
00:20:54,907 --> 00:21:00,907
like, you and I both know that you already can do, like, the equivalent, what BIP300 is,

380
00:21:00,907 --> 00:21:06,028
this is what's, I think, very rarely understood, is that BIP300 is actually putting, like, a short

381
00:21:06,028 --> 00:21:10,508
leash on the miners, but you can already, if you want, you can already do, the miners could just

382
00:21:10,508 --> 00:21:17,707
make a wallet, and they could say, we promise we'll give these back if only two people who

383
00:21:17,707 --> 00:21:23,387
withdraw will make their own EVM chain, and we promise to give the coins back and be, like,

384
00:21:23,387 --> 00:21:31,787
custodial with the miners as custodians, and so, because many of the drivechain,

385
00:21:32,588 --> 00:21:36,268
I think many of the drivechain supporters try to do an argument like that, they say, well,

386
00:21:36,268 --> 00:21:42,027
listen, like, that's what we do with merge mining, we say, like, merge mining is already

387
00:21:42,027 --> 00:21:46,667
possible with altcoins, it's already possible to just, and everything the miners do, because this

388
00:21:46,667 --> 00:21:51,468
is where, because it's true that we do the opt-in and you can't stop it, that's for lightning,

389
00:21:52,588 --> 00:21:56,908
and then it's the same for drivechain and for maybe miners can steal or for whatever,

390
00:21:56,908 --> 00:22:01,067
but then where people start to get uncomfortable is they say, well, wait, what about the miners?

391
00:22:01,787 --> 00:22:08,747
The miners have to run this software, you know, but then we in the drivechain community,

392
00:22:08,747 --> 00:22:13,468
we say, like, well, you know, miners have to do all kinds of things all the time, they don't really

393
00:22:13,468 --> 00:22:17,067
have to do anything, they could just stop mining or they could not mine in a certain way, they don't

394
00:22:17,067 --> 00:22:21,228
have to use immersion cooling, you know, they could do, they don't have to use it, whatever

395
00:22:21,228 --> 00:22:25,787
they want, whenever they want, so it's really not the same, they can merge mine altcoins or they

396
00:22:25,787 --> 00:22:32,187
can switch pools, it doesn't, so we try to, I think people try to do that, but it doesn't seem

397
00:22:32,187 --> 00:22:40,748
to be working, but maybe we should, they try to do what, sorry? Well, we try to say that the type

398
00:22:40,748 --> 00:22:45,228
of thing you could do with drivechain is all, all of it is already being done, you know, like

399
00:22:45,228 --> 00:22:50,987
we could have a terrible custodial sidechain where you just send bitcoin over, oh right, you could

400
00:22:50,987 --> 00:23:00,588
emulate, yeah, well, I mean, liquid, liquid, I mean, we can do even worse than, we could make it

401
00:23:00,588 --> 00:23:05,228
like fully custodial, where I just like, you just send it to a bitcoin wall created by me,

402
00:23:05,787 --> 00:23:12,668
Paul Sztorc, and I promise to give everyone back the coins, and that is, oh, and, but this is a

403
00:23:12,668 --> 00:23:18,987
question that I asked, you know, I asked Andrew Polstra and, and Peter Weil in Amsterdam, several

404
00:23:18,987 --> 00:23:24,108
Amsterdams ago, I don't remember exactly when, but I said, this is my question about, like, people are

405
00:23:24,108 --> 00:23:30,907
allowed to spend their bitcoin on stuff, right, and they're allowed to sell their bitcoin for an altcoin,

406
00:23:32,028 --> 00:23:38,028
and this is like, you know, this is very similar to that, we're just saying, we're sending, sending

407
00:23:38,028 --> 00:23:43,148
the bitcoin to a script that unlocks a certain way, you know what I mean, we can already send our coins

408
00:23:43,148 --> 00:23:49,388
to miners directly, I can already send my coins to whatever, Boundary, or Jihan Wu, or whatever, whoever

409
00:23:49,388 --> 00:23:54,987
it is, pick your person, I can already, Marty Bent even, I can already send my coins to a miner, someone

410
00:23:54,987 --> 00:24:00,907
that is involved with mining, and I can already hope that they would come back, and I can already

411
00:24:01,868 --> 00:24:06,748
sell my coins for, like, Solana, or something that we would all kind of agree is probably not a great idea,

412
00:24:07,468 --> 00:24:12,028
but the end user can already do that, so these things seem to work, I mean, I guess what you're

413
00:24:12,028 --> 00:24:16,668
thinking about is something like, with the CTV can just emulate all the BIP300 parts, but I don't

414
00:24:16,668 --> 00:24:24,748
think that people appreciate that the design of BIP300 is to, is really to make life difficult

415
00:24:24,748 --> 00:24:32,267
for the miners, because it's kind of like, if they put themselves in a difficult spot, like a spot

416
00:24:32,267 --> 00:24:36,188
where it's hard for them to steal, then it's easier for people to trust them in the first place,

417
00:24:36,908 --> 00:24:40,507
moving the coins into the, into the sidechain in the first place.

418
00:24:41,547 --> 00:24:46,827
That's, like, where I was trying to go with it, but I think a lot of people have this view that

419
00:24:47,628 --> 00:24:52,748
this will decentralize mining, or something, I don't know, they have, like, they have mysterious views,

420
00:24:52,748 --> 00:24:55,707
but I try to say, well, this is already, all the stuff that's happening here, we're always, we're

421
00:24:55,707 --> 00:25:02,107
already, so, like, I think someone says something like, people will say something like, merge mining

422
00:25:02,107 --> 00:25:07,227
is bad, so we shouldn't do any more of it via sidechains. I think that's something that I still

423
00:25:07,227 --> 00:25:14,267
hear from time to time. Well, I mean, another comparable, because you were talking about,

424
00:25:15,308 --> 00:25:21,068
I mean, having, having a different security trade-off in something, I mean, it's curious,

425
00:25:21,068 --> 00:25:26,267
because Bitcoin itself, you know, when it, when it was first announced to the world,

426
00:25:26,267 --> 00:25:32,908
and then, after a few years, academics started to comment on it, and it was usually fairly

427
00:25:32,908 --> 00:25:37,068
derogatory, right, because they were like, well, it wasn't academically published, that didn't seem

428
00:25:37,068 --> 00:25:44,427
to sit well with them, and then they'd misunderstand how it works, or they would say it is, you know,

429
00:25:44,427 --> 00:25:52,267
its security model is weird, right? Like, the old eCash systems were using digital signatures,

430
00:25:52,267 --> 00:25:58,028
so they were extremely secure, but they had, whereas Bitcoin is like game theory, right? So,

431
00:25:58,028 --> 00:26:02,507
you know, the good guys versus the bad guys, like 50-50, plus some economic incentive,

432
00:26:03,308 --> 00:26:10,107
and so it was just different, and it took a while to, to get over that. So, but it's,

433
00:26:10,107 --> 00:26:17,227
it's interesting that, you know, now, these days, I think people use Bitcoin as the

434
00:26:18,188 --> 00:26:26,667
litmus test, right? Like, Bitcoin is secure, it's the gold standard, and so anything you want to

435
00:26:26,667 --> 00:26:34,588
introduce which has a slightly different security trade-off, they'll be immediately suspicious of

436
00:26:34,588 --> 00:26:39,148
it. It's like, well, it's not as secure as Bitcoin. And my usual answer to that is, well,

437
00:26:39,148 --> 00:26:45,308
of course, like, you know, no layer two, or, you know, sidechain, et cetera, is typically going to

438
00:26:45,547 --> 00:26:52,267
be as secure as Bitcoin, or Bitcoin, whatever, you know, clever security technique you come up with,

439
00:26:52,267 --> 00:26:56,107
Bitcoin would adopt it, like in the main chain, right? And so, like, Lightning is making a

440
00:26:56,107 --> 00:27:02,348
trade-off, right? You've got hot wallets, you've got keys online, and so it provides an advantage,

441
00:27:02,348 --> 00:27:08,667
but it also has security disadvantages. And so, another comparable is the statechains that

442
00:27:08,667 --> 00:27:15,947
Ruben Sampson proposed and is implemented in the Mercury wallet. So, as I understand that,

443
00:27:15,947 --> 00:27:26,588
it's another kind of, I guess, sort of layer two, right? With something vaguely like Lightning with,

444
00:27:26,588 --> 00:27:33,707
like, delayed settlement or something. And you saw, it has weird security trade-off where

445
00:27:34,667 --> 00:27:40,748
as long as there's a sort of server that's sort of co-signing or something, and as long as it's

446
00:27:40,748 --> 00:27:46,748
not trying to steal from you, it's okay. You know, if somebody comes to the co-signer afterwards and

447
00:27:46,748 --> 00:27:55,068
says, take this coin or freeze this coin, they can't. But if they were, you know, if they weren't

448
00:27:55,068 --> 00:28:00,667
trustworthy to start with, they could, you know, get involved in the payment path and just steal

449
00:28:00,667 --> 00:28:03,788
any coin that passed through their hands historically, and there's not much you can

450
00:28:03,788 --> 00:28:09,947
do about it. And, you know, so it's not as widely discussed as well known, but it's implemented,

451
00:28:09,947 --> 00:28:15,387
and you can use it, and it doesn't need any opcodes to make it work. So, I don't know,

452
00:28:15,387 --> 00:28:21,628
you know, if that had more flexibility or scope, maybe it would elicit the same kind of

453
00:28:22,827 --> 00:28:27,628
security trade-off discomfort, let's say, right?

454
00:28:27,788 --> 00:28:35,387
Yeah, Statechains, I think, I'm pluralist, I think we should try all these ideas. But yeah,

455
00:28:35,387 --> 00:28:40,267
the Statechains do have the federation that can steal from you, of course, I mean, lots of things

456
00:28:40,267 --> 00:28:48,187
do. And they have the one UXXO size thing, but I think there's kind of like, I mean, I tell this

457
00:28:48,187 --> 00:28:52,108
story over and over again, and I don't know if anyone, people are tired of, probably everyone

458
00:28:52,108 --> 00:28:56,747
is tired of hearing this story, but I think Statechains is another example of a good example

459
00:28:57,228 --> 00:29:03,627
of a data point on this story, where it's kind of like, because of the block size war, and because

460
00:29:03,627 --> 00:29:12,988
of the dispute, there was like, people really grouped up, and there was a lot of like, it wasn't

461
00:29:12,988 --> 00:29:18,028
as easy to be critical of Bitcoin, because people thought that anyone who is, anyone who said

462
00:29:18,028 --> 00:29:23,867
anything bad about Bitcoin Core, Lightning, Statechains, whatever, it was like, these people

463
00:29:24,507 --> 00:29:31,468
are spies sent from, you know, the Bitcoin Cash community to spread FUD or something.

464
00:29:32,028 --> 00:29:36,668
It says both, I think, oh, sorry. Well, I mean, it does happen, right? I mean,

465
00:29:37,548 --> 00:29:42,108
all coins, both of them, both of those are examples of things where,

466
00:29:44,267 --> 00:29:50,267
like, we seem to have become like very monocultury or something. Don't you think?

467
00:29:50,668 --> 00:29:55,707
Like, where it's like, there's like a set, well, it's weird, because in Bitcoin, there's so many

468
00:29:55,707 --> 00:30:01,548
ideas, but they're not, like, none of them, there's different expectations of like, how they

469
00:30:01,548 --> 00:30:09,307
will be taken seriously, and then there's like, will this idea, like, should they be done in a

470
00:30:09,307 --> 00:30:15,067
certain order or something? Well, I mean, I think the, you know, there seems to be a lot of excitement

471
00:30:15,067 --> 00:30:24,028
about layer two things, right? So, the ARK proposal that Burek made only a few months ago

472
00:30:24,588 --> 00:30:31,148
seems to create a bit of buzz, and the Fedimint protocol, and there's, you know, there's a company

473
00:30:31,148 --> 00:30:37,148
with making wallets for it, seems to create a lot of excitement in terms of its trade-off. But I

474
00:30:37,148 --> 00:30:45,148
mean, that is, like, also, that's federated, but arguably very scalable and very private

475
00:30:46,108 --> 00:30:51,627
as the trade-off. So, yeah, I give up something. I feel an immense sense of dread with Fedimint.

476
00:30:51,627 --> 00:30:56,908
I don't know about you, but I think that it's like, you're just asking, first of all, I don't

477
00:30:56,908 --> 00:30:59,867
think, like, people will sign up, and then they're going to have to pick, like, a list, or they're

478
00:30:59,867 --> 00:31:04,108
going to have to pick a group. I think they'll lose a lot of people on that. I mean, I hope I'm

479
00:31:04,108 --> 00:31:09,548
wrong. I hope it's a big success, but I think they'll lose a lot of people at the first step,

480
00:31:09,548 --> 00:31:13,788
which is choose your group, and then the group will, it's just choose who's going to rug pull

481
00:31:13,788 --> 00:31:21,067
you, like, very soon. Well, the rug pull is the, yeah, I mean, the rug pull risk is the trade-off,

482
00:31:21,067 --> 00:31:27,228
yeah. And I mean, their proposal for that is sort of more inclusive because it's a community thing,

483
00:31:27,307 --> 00:31:34,187
right? So, you know, different people can run alone, whereas Liquid is sort of a similar model

484
00:31:34,187 --> 00:31:40,507
in a way, except it's, you know, a group of exchangers and people like that, right? So,

485
00:31:40,507 --> 00:31:51,387
it's sort of more identified and less community. But, yeah, I mean, and in a way, the rug pull

486
00:31:52,348 --> 00:32:00,908
risk of, I mean, the theory is that, well, firstly, if there's any, like, individuals or

487
00:32:00,908 --> 00:32:06,747
groups of individuals that could steal coins, that worries people, right? And in Bitcoin itself,

488
00:32:06,747 --> 00:32:16,187
that's pretty hard because, you know, a rollback attack or something is probably very hard to

489
00:32:16,588 --> 00:32:23,867
achieve at this point, right, the amount of miners and stuff. But it's sort of,

490
00:32:26,028 --> 00:32:32,428
people think about that in terms of sidechains and drivechains because

491
00:32:35,627 --> 00:32:43,307
the main chain miners, well, a user's full node is not validating the sidechain rules,

492
00:32:43,307 --> 00:32:46,988
or if it did, there's not much it can do about it because it's not a consensus rule, right?

493
00:32:47,707 --> 00:32:54,668
So, the miners are kind of, you know, a miner could get away with, you know, mining Bitcoin

494
00:32:54,668 --> 00:32:58,827
in accordance with Bitcoin rules and stealing from the sidechain or drivechain. And so,

495
00:32:59,468 --> 00:33:05,307
the concept that miners are anonymous, which is, you know, sort of important conceptually as

496
00:33:06,267 --> 00:33:16,668
a reason why Bitcoin can have sort of finality and no censorship of transactions and things like that,

497
00:33:19,067 --> 00:33:25,627
means that people are nervous about a sort of distributed wallet that is controlled by

498
00:33:26,747 --> 00:33:30,908
whoever's mining recently, right? And some of the proposals are relatively fast

499
00:33:31,867 --> 00:33:36,908
to peg out, and drivechains takes another approach, which is intentionally slow to peg out

500
00:33:36,908 --> 00:33:40,668
so that humans have time to react. But fast versions...

501
00:33:41,788 --> 00:33:47,627
Yeah, I think it has a history of not working, like Ethereum and similar versions or something,

502
00:33:47,627 --> 00:33:50,908
where it's very fast, and then everything's just taken out of them.

503
00:33:51,947 --> 00:33:55,468
I mean, in terms of the bridges, yeah, those are kind of infamous. I don't know,

504
00:33:55,548 --> 00:34:01,867
people just keep using them anyway. But there are so many altcoin bridges that have been emptied

505
00:34:01,867 --> 00:34:05,867
of hundreds of millions of dollars over the last few years. It just keeps happening and happening.

506
00:34:05,867 --> 00:34:11,788
And you never quite know with those, whether the people who operate it or the software.

507
00:34:14,428 --> 00:34:20,588
This is a bridge. What it really is is to give us all your coins, wallet, and then they just move

508
00:34:20,588 --> 00:34:32,908
them. Yeah. So I think that the fact that, sort of conceptually, you could portray

509
00:34:34,507 --> 00:34:39,148
the wallets in a sidechain as behaving that way is...

510
00:34:44,827 --> 00:34:48,028
It's less sort of cut-and-drive ranks. It's not consensus enforced. I mean,

511
00:34:48,827 --> 00:34:57,788
I guess it has some similarities with Fedimint's trade-off in an abstract way, which is Fedimint

512
00:34:57,788 --> 00:35:01,947
is not able to audit things because of the privacy, and that's what gives its capability.

513
00:35:02,827 --> 00:35:10,908
And with a sidechain or a drivechain, the full nodes that, like the consensus rules of the main

514
00:35:10,908 --> 00:35:14,908
network, don't really care too much about the sidechain, so they're not enforcing its rules,

515
00:35:14,908 --> 00:35:18,267
and they don't want to enforce its rules, because if it had bugs in it, then it could fork the main

516
00:35:18,267 --> 00:35:29,067
network. And so you have... It's sort of by design, but it has a cost, right? So by design,

517
00:35:30,028 --> 00:35:33,468
you don't want the consensus rules to enforce the rules of the sidechain.

518
00:35:36,588 --> 00:35:41,548
Because that's what makes it permissionless and makes it safe for people to try experiments on

519
00:35:41,548 --> 00:35:47,307
sidechains that won't bleed through to the main chain. But the trade-off is,

520
00:35:48,187 --> 00:35:52,028
well, that's the side effect, right? The consensus rules are not going to protect you.

521
00:35:53,788 --> 00:36:00,668
So then you're relying on this kind of looser binding, right, of sort of with the fraud-proofs

522
00:36:00,668 --> 00:36:06,507
version, but you better hope the fraud-proofs are not censored, or with the slow version.

523
00:36:07,387 --> 00:36:11,548
Yeah, see, that's why I never thought the fraud-proofs... I thought this is, like, always

524
00:36:11,548 --> 00:36:17,307
pointless, because if 51% is against you, they can filter any message they want out. So I always

525
00:36:17,307 --> 00:36:22,028
thought it's pointless, and in a way, the drivechain has the fraud-proofs thing, because

526
00:36:22,028 --> 00:36:27,947
what we programmed the sidechain template to do is the sidechain template reduces all the

527
00:36:27,947 --> 00:36:33,307
outstanding activity to just the one hash, and then it puts the hash in every block header on

528
00:36:33,307 --> 00:36:40,028
the sidechain. So it's very, very easy to find, very, very, screaming the correct hash, like,

529
00:36:40,028 --> 00:36:46,187
as loud as possible, in the easiest-to-find way on the sidechain. And that's kind of like

530
00:36:46,187 --> 00:36:51,148
the fraud-proof, is just to say anyone runs the software in SPV mode, or they run the software in

531
00:36:51,148 --> 00:36:56,908
full mode, or whatever, they'll certainly know that fraud has been done. But you see,

532
00:36:56,908 --> 00:37:00,747
the idea of putting the fraud-proof on L1 is, like, just fundamentally impossible,

533
00:37:00,747 --> 00:37:05,148
because there's eventually going to be a point where the blockchain either goes left, if a certain

534
00:37:05,148 --> 00:37:11,548
message is in the blockchain, or it'll go right, and the miners can just censor that message, and they can

535
00:37:11,548 --> 00:37:14,987
always control it. So I always thought that it was always just going to come down to the miners

536
00:37:14,987 --> 00:37:21,548
themselves. This is what I thought in 2015, when I was reading some of... I was reading the Blockstream

537
00:37:21,548 --> 00:37:24,908
paper, I was reading all the other stuff, and I just thought, well, just push comes to shove.

538
00:37:25,867 --> 00:37:30,028
This is just the way it's going to be. And everything else, even when these people say

539
00:37:30,028 --> 00:37:36,507
ZK-SNARK, CoinWitness, whatever, I just thought, none of that will work, because the miners will

540
00:37:36,507 --> 00:37:40,828
just censor. They'll be ready for that, and they'll just censor anything they don't like. And

541
00:37:40,828 --> 00:37:46,428
the only way is to just give the miners the fees, give the miners the price appreciation,

542
00:37:47,068 --> 00:37:53,708
and then just make it as absolutely easy as possible to tell, for everyone to tell,

543
00:37:53,708 --> 00:37:57,308
whether or not this theft is happening, and do something about it.

544
00:37:59,148 --> 00:38:10,507
Yeah. Do you mind summarizing the blind merge mining, and why that exists, and how it works?

545
00:38:11,867 --> 00:38:21,148
Absolutely. It exists because I thought the idea... I was like, drivechain November 2015,

546
00:38:21,148 --> 00:38:27,467
then I was kind of like, whatever. Then Scaling 3 came up, and I wanted to speak there,

547
00:38:27,467 --> 00:38:31,148
because of how prestigious it is to speak at Scaling Bitcoin. So I thought, okay,

548
00:38:32,188 --> 00:38:38,987
I'll talk about sidechain scaling. And then we had some conversations with yourself,

549
00:38:38,987 --> 00:38:41,948
and Matt Corallo was there, and all kinds of other people were there.

550
00:38:43,467 --> 00:38:48,828
And people were worried about the miners running this software. People didn't like that.

551
00:38:48,828 --> 00:38:53,548
And then people brought up one funny story that I've heard over the years, which I think is

552
00:38:53,548 --> 00:38:58,188
probably true. I've heard a couple of different people tell it. So it's like,

553
00:38:58,188 --> 00:39:01,068
they used to merge mine Namecoin, and it was a tiny amount of money.

554
00:39:02,987 --> 00:39:06,507
But you had to run the Namecoin software, because the way merge mining works, traditional merge

555
00:39:06,507 --> 00:39:12,107
mining, you have to run the Namecoin software. You get like an extra $5 or something every 10

556
00:39:12,107 --> 00:39:19,308
minutes. Whereas the Bitcoin block was worth like $7,000 or some huge amount, 1000x more.

557
00:39:20,747 --> 00:39:26,828
And then the Namecoin software would crash, and the whole mining farm would shut off.

558
00:39:26,828 --> 00:39:32,428
And this would take like 35 seconds to restart or something. Namecoin was buggier.

559
00:39:33,788 --> 00:39:37,308
And then even there's some cases where power is flowing into these machines,

560
00:39:37,308 --> 00:39:41,227
but they're off. And so the capacitors just like explode and burst into flames or something.

561
00:39:41,227 --> 00:39:45,068
Something funny like that would happen. So I was thinking about this after,

562
00:39:45,068 --> 00:39:49,308
because it was October 2016. And then we had Construct. People were talking about blah,

563
00:39:49,308 --> 00:39:52,507
blah, blah about the miners, mining centralization. So that's really this

564
00:39:52,507 --> 00:39:59,467
mining centralization work came up. And I decided to publish in January 2017, right before Construct.

565
00:39:59,930 --> 00:40:10,810
Blind merge mining, which is this thing I had sort of invented, is the mining threat model equilibrium analysis. This is all my thoughts. I've just been thinking about this question the whole time.

566
00:40:11,170 --> 00:40:28,730
When someone tweets about it today in 2023, they don't know that from October 2016 to like January 2017, I already thought more about it than, you know, whatever it could possibly have been done by people who have just joined recently.

567
00:40:29,329 --> 00:40:52,369
And so I was really interested in this idea. And I actually thought that the mining centralization argument wasn't good. Like it wasn't a real, it's not a real thing. And which I think we should probably talk about because it's a fascinating idea as to what exactly do people mean when they or when they complain about miners or when they complain about mining centralization, which I think is like an absolutely fascinating question.

568
00:40:52,369 --> 00:41:05,650
But I published this essay, The Mirage of Mining Centralization, that was sort of syndicated, because I think people found it interesting, that this mining equilibrium model that was also long and people didn't really care about.

569
00:41:06,129 --> 00:41:20,209
And then I published blind merge mining. And this was all the culmination of all this, this thought or whatever. So the blind merge mining idea is that you can cheat, the miners don't necessarily need to run the sidechain full node, they can just partner with someone who is.

570
00:41:20,690 --> 00:41:39,650
That person has the magic bridge that makes it all work is that there's a person who owns layer one coins, have a layer one Bitcoin wallet. And the same person is already using the sidechain for something else. So they're just a sidechain user, they have a sidechain node, these people already exist.

571
00:41:40,610 --> 00:42:00,529
Basically, everyone who's a sidechain user has some layer one usage, because in order to run the layer two sidechain node, just like in order to run a lightning node, you must run layer one node. So this is actually a very easy act. So you have these people who are using the layer two sidechain, they probably have some coins, they probably store most of their coins on L1.

572
00:42:01,489 --> 00:42:16,610
They probably only use the L2 for whatever feature it is. Or it's like a hot wallet, like for payments or for whatever. So I was like, well, this is the magic is you have the same person standing on both sides, and they they build the sidechain block paying themselves the fee.

573
00:42:16,610 --> 00:42:22,510
fee. And then on layer one, the miners just see a message that says, put this code

574
00:42:22,510 --> 00:42:25,550
in, the coin base are put in a certain spot.

575
00:42:25,710 --> 00:42:26,670
Put this code in.

576
00:42:27,310 --> 00:42:33,310
And if you put this code in and move the block from sidechain block 14 to side

577
00:42:33,310 --> 00:42:36,370
chain block 15, I'll pay you this amount of money.

578
00:42:36,430 --> 00:42:39,030
So the sidechain block is worth five Bitcoin.

579
00:42:39,550 --> 00:42:45,450
Then they pay 4.99 layer one coins to get five sidechain coins when the

580
00:42:45,450 --> 00:42:46,610
block is the side block

581
00:42:46,610 --> 00:43:16,610
five sidechain coins when the block is mined, on layer one, they pay 4.99 layer one coins to the miner, the miner gets 4.99 coins, so they don't really care. And they also have no, they're relatively indifferent as to whether or not they themselves, it's like they're losing money by not running, they're not missing out. They may miss out slightly, but this is because there's a layer one to layer two exchange rate. And because running the sidechain node is actually literally inconvenient. So on one hand, they can either run the sidechain node and

582
00:43:17,610 --> 00:43:19,330
they don't have to, so they could be blind.

583
00:43:19,330 --> 00:43:27,009
And do they need like a pooling protocol to do that? Or can that transaction be broadcast so any miner would automatically include it?

584
00:43:27,690 --> 00:43:38,089
Yeah, I think we made it so that it would just be a broadcasted transaction. I think, I don't know if they, well, I mean, we have cryptex here, actually.

585
00:43:38,130 --> 00:44:07,649
Yeah, so you can create a special transaction that is called a BMM request. And that includes the hash of the sidechain hashmerkle route, and eight more bytes to identify it as being a BMM request, as well as a few bytes of the previous mainchain block hash. And then a miner who wants to include that transaction, they have to include it at a specific height because it has a lock time set. And then they also have to include the BMM

586
00:44:07,649 --> 00:44:24,050
commitment in the coinbase output of the block that they include the BMM request transaction in. So there's basically a way to use the memory pool to create bids for miners to create these BMM commitment hashes in their own coinbases.

587
00:44:25,810 --> 00:44:35,490
Okay, so it sounds like it's not like automatic for existing miners necessarily, but there's a way to do it. Or something like that.

588
00:44:35,810 --> 00:44:41,490
Yeah, they would have to update to recognize those transactions and include them correctly.

589
00:44:41,490 --> 00:44:42,930
Yeah,

590
00:44:43,302 --> 00:44:48,102
enforce something because we didn't want it, we can't have them collect all the bids.

591
00:44:48,342 --> 00:44:52,622
Like what if everyone you're at an auction house and I say, bid $7, I bid $7.50, I bid

592
00:44:52,622 --> 00:44:55,402
$8. The auctioneer doesn't collect them all.

593
00:44:55,822 --> 00:45:01,362
So it's actually kind of weird because we made a special kind of transaction where

594
00:45:01,362 --> 00:45:03,116
only one of them can

595
00:45:03,156 --> 00:45:08,296
be included per sidechain like per block and so the rest all just expire as invalid so

596
00:45:08,296 --> 00:45:14,256
there's it's kind of an it's very unusual whereas most bitcoin transactions do not

597
00:45:14,256 --> 00:45:18,476
have that property at all where if most of them if they're valid they're just always

598
00:45:18,476 --> 00:45:22,930
valid and they're just there for the picking kind of so this is a very unusual type of

599
00:45:23,569 --> 00:45:49,810
Yeah, so in our testnet, you'll see every block, there'll be a little bit of a flood of BMM request transactions. And then immediately after the sidechain block is connected, those will all be removed from the memory pool because after they've missed their chance to create the BMM block with the lock time set and the previous mainchain block bytes, the BMM request basically expires, and they don't lose their money because it's removed from the memory pool.

600
00:45:50,770 --> 00:45:53,569
It becomes an invalid transaction at that point.

601
00:45:54,770 --> 00:46:05,969
Hmm, okay. And, and I mean, I guess the miner isn't really checking right and it's not really a specific of the blend merge mining versus

602
00:46:06,370 --> 00:46:09,090
they would choose whoever pays the highest fee.

603
00:46:09,090 --> 00:46:31,490
Right. But I mean, like, if, if, you know, I guess there's a potential, because the miner is maybe not verifying or, you know, they could just cheat that you can get a commitment that is somehow wrong, right? Like, it looks like a sidechain commitment, but the hash is wrong when there's no block with a hash.

604
00:46:32,129 --> 00:46:33,490
Oh, yeah, you could do that.

605
00:46:33,490 --> 00:46:40,530
Yeah, and then the sidechain would have to decide how it handles that. Our sidechain examples would ignore that.

606
00:46:42,530 --> 00:46:43,090
Yeah.

607
00:46:44,050 --> 00:46:45,169
So that would just be the miner.

608
00:46:45,169 --> 00:46:48,370
Like, somebody would just waste money, basically, right?

609
00:46:48,370 --> 00:46:49,810
Yeah, exactly.

610
00:46:49,810 --> 00:46:50,689
Yeah.

611
00:46:50,909 --> 00:46:56,709
exactly. Yeah. You'd be giving up space for real fee -paying transactions by doing that.

612
00:46:57,209 --> 00:47:02,249
person who has an invalid hash would be like, you tried to...

613
00:47:02,589 --> 00:47:07,389
The analogy is like, if it's included in L1 and it does the BIP301

614
00:47:07,389 --> 00:47:09,869
BMM rules, that would

615
00:47:09,945 --> 00:47:13,345
be as if the sidechain block met its difficulty requirement.

616
00:47:14,065 --> 00:47:17,325
then maybe the header meets the requirement

617
00:47:17,325 --> 00:47:26,625
or

618
00:47:26,625 --> 00:47:29,185
the hash meet the difficulty

619
00:47:29,201 --> 00:47:33,401
requirement, that doesn't necessarily mean that a block exists, or is valid, or is a

620
00:47:33,401 --> 00:47:34,761
real sidechain block or something.

621
00:47:35,121 --> 00:47:39,761
it's kind of like, that's kind of how I think of it in my head, is that that's the

622
00:47:39,761 --> 00:47:44,981
analogy, like you pay and you get included, now you met the difficulty adjustment on

623
00:47:44,981 --> 00:47:48,301
the sidechain. So that is, that one is valid, but if there's no

624
00:47:48,457 --> 00:47:54,277
vector of transactions or anything, there's, or if there's just no, the block is

625
00:47:54,277 --> 00:47:57,657
never broadcast, it's just, you would go back.

626
00:47:58,137 --> 00:48:01,357
This is the key thing is you have to be able to reorg the sidechain without reorging

627
00:48:01,357 --> 00:48:02,217
the main chain also.

628
00:48:02,657 --> 00:48:07,557
this is like each block kind of refers, each code only works for like

629
00:48:07,893 --> 00:48:13,133
spot in the main chain blocks chain and the sidechain blockchain.

630
00:48:13,293 --> 00:48:16,353
only... then if you try to move it somewhere else, doesn't work.

631
00:48:17,273 --> 00:48:18,773
Actually, what you said is interesting, right?

632
00:48:19,653 --> 00:48:24,213
I mean, the main chain can have reorgs, but the drivechain can too.

633
00:48:24,653 --> 00:48:26,953
So the main chain could be progressing without reorgs.

634
00:48:27,930 --> 00:48:33,690
And then there's a commitment to like one height in the drivechain and then a

635
00:48:33,690 --> 00:48:37,370
competing one, like, so it sort of gets an orphan block in the drivechain.

636
00:48:38,169 --> 00:48:41,010
And that's valid, I guess, because otherwise you were assuming very tight

637
00:48:41,010 --> 00:48:44,690
coordination of, you know, drivechain liners or something, right.

638
00:48:45,089 --> 00:48:49,890
So important that I made an image of it in the January, uh, 2017.

639
00:48:50,849 --> 00:48:55,690
And, uh, because the, remember when I sat down to design it, the whole reason

640
00:48:55,690 --> 00:48:59,010
that I, the whole reason that it came to exist at all was because of this

641
00:48:59,010 --> 00:49:04,289
feedback about from the sidechain idea.

642
00:49:04,329 --> 00:49:09,450
This is like a sidechain phobia or something, which is like 2016 after

643
00:49:09,450 --> 00:49:12,649
scaling three, people were like, okay, people were like, we admit that the

644
00:49:12,649 --> 00:49:17,490
nodes don't have to run this, but the, maybe the miners have to run it also.

645
00:49:17,729 --> 00:49:21,450
So I was thinking like, how do I make it so that the miners are never,

646
00:49:21,450 --> 00:49:24,089
ever harmed by not running this.

647
00:49:25,529 --> 00:49:29,089
That was the whole ground up building of the, the idea.

648
00:49:29,649 --> 00:49:33,490
So you have to maintain reorgs and everything above it reorgs, including

649
00:49:33,490 --> 00:49:37,930
lightning, whatever sidechains, state chain opens everything, everything

650
00:49:37,930 --> 00:49:44,329
reorgs at the main chain reorgs, but the sidechain can reorg, uh, and without

651
00:49:45,329 --> 00:49:48,289
for this reason, I have, I actually wonder about whether or not someone

652
00:49:48,289 --> 00:49:52,130
will actually bid up the whole amount, including the main chain miners themselves.

653
00:49:52,649 --> 00:49:56,370
So I don't know about the L1 to L2 exchange rate for this type of thing,

654
00:49:56,810 --> 00:49:58,570
because it'd be like a piece sniping or something.

655
00:49:59,289 --> 00:50:02,930
So for everyone, it might be the case that the sidechain block is worth eight

656
00:50:03,370 --> 00:50:07,930
Bitcoin in fees or something, whatever number you like, but then on layer one,

657
00:50:09,130 --> 00:50:12,890
even to a huge group of miners or even to anyone, it may only be worth.

658
00:50:14,769 --> 00:50:18,969
You know, six Bitcoin or something because people might think, well, the

659
00:50:18,969 --> 00:50:23,649
sidechain block, some of my reorg, the sidechain block late later than I just

660
00:50:23,649 --> 00:50:27,089
paid six on layer one, but now I get nothing.

661
00:50:28,570 --> 00:50:30,010
So you have the reorg risk.

662
00:50:30,089 --> 00:50:33,089
The reorg risk is like a big discount that falls over the whole thing.

663
00:50:34,010 --> 00:50:34,610
Oh, right.

664
00:50:34,610 --> 00:50:42,089
The person who is, uh, constructing the blind merge mine is paying for inclusion,

665
00:50:42,089 --> 00:50:44,690
but it could get unincluded.

666
00:50:44,690 --> 00:50:48,130
So they would pay and not receive the sidechain fees in that case, I guess.

667
00:50:48,169 --> 00:50:48,450
Yeah.

668
00:50:48,450 --> 00:50:54,610
So they pay on L1 in this block and then there's a sidechain block that pays them

669
00:50:54,610 --> 00:50:58,050
the fees, which may, they may not like as much or whatever, you know?

670
00:50:58,089 --> 00:51:04,169
So, I mean, I think the analogy is good of difficulty requirement, because that's

671
00:51:04,169 --> 00:51:07,450
the case in every blockchain is that your block might be reorged.

672
00:51:07,930 --> 00:51:12,289
So this is why the analogy, so I try to emphasize this analogy that what you're

673
00:51:12,289 --> 00:51:17,810
doing is when you, when you do the BMM block, you are meeting the difficulty

674
00:51:17,810 --> 00:51:21,089
requirement for that block, but you don't know that it won't be orphaned.

675
00:51:21,529 --> 00:51:23,329
So it's the orphan or everybody.

676
00:51:23,329 --> 00:51:25,610
Well, I mean, I think that's, that's unavoidable, really.

677
00:51:25,890 --> 00:51:31,450
I mean, I think it's good that you, you know, design it in a robust way because

678
00:51:31,729 --> 00:51:35,529
you know, otherwise the only way to fix it is centralized.

679
00:51:35,529 --> 00:51:43,329
Actually, it cracks me up when some of the big block Bitcoin forks historically

680
00:51:44,209 --> 00:51:50,969
kind of tried to fight against double spans and RBF, and they basically found

681
00:51:50,969 --> 00:51:56,250
out, like they eventually learned why Bitcoin allows orphans and why there's

682
00:51:56,250 --> 00:51:58,050
proof of work, it's the hard way.

683
00:51:58,490 --> 00:52:01,769
So it's like, it's inevitable, right?

684
00:52:01,769 --> 00:52:03,089
That is very important.

685
00:52:03,089 --> 00:52:07,769
We should explain this to the audience because it is actually, it's like a kind

686
00:52:07,769 --> 00:52:12,089
of a weird deep fact or something, but this is related to like, like imagine

687
00:52:12,089 --> 00:52:16,690
someone finds a block that meets the proof of work requirement, but they

688
00:52:16,690 --> 00:52:23,289
just never broadcast it and you don't know the block could contain an error.

689
00:52:23,329 --> 00:52:27,209
It could have an invalid transaction or it could be perfectly valid.

690
00:52:28,130 --> 00:52:30,370
You don't really know if you should build on it or not.

691
00:52:30,930 --> 00:52:35,289
And you have no choice, but to basically orphan this block until people say what

692
00:52:35,289 --> 00:52:38,169
it is, you have to, the orphan is always better.

693
00:52:38,209 --> 00:52:40,729
It's always better to just orphan the block and say, I don't know what's going

694
00:52:40,729 --> 00:52:45,089
on, but unless I see, unless I get a perfectly perfect block in front of me

695
00:52:45,089 --> 00:52:48,649
right now, I'm going to thumbs this down or something.

696
00:52:48,690 --> 00:52:49,370
It's a very.

697
00:52:53,450 --> 00:52:57,690
I mean, actually it was an empirical lesson that hopefully miners learned

698
00:52:57,690 --> 00:53:04,329
from in BIP 66 activation, where too many people were spy mining.

699
00:53:05,050 --> 00:53:08,969
I just, just pulling headers off the pool and mining on them without, before

700
00:53:08,969 --> 00:53:10,409
waiting to check if the block is valid.

701
00:53:11,169 --> 00:53:14,890
And yeah, they, they kind of raced away building on the wrong chain

702
00:53:15,529 --> 00:53:17,169
that was coincidentally invalid.

703
00:53:17,769 --> 00:53:19,729
And that had to be fixed in a hurry.

704
00:53:20,409 --> 00:53:25,050
The model is an example of, in fact, everyone who ran their own pool

705
00:53:25,050 --> 00:53:26,890
node was totally unaffected by that.

706
00:53:27,890 --> 00:53:28,289
Right.

707
00:53:28,450 --> 00:53:28,729
Yeah.

708
00:53:29,250 --> 00:53:30,050
It's kind of ambiguous.

709
00:53:30,050 --> 00:53:33,810
If you, if you, how, how quickly you upgraded or something, I suppose.

710
00:53:34,690 --> 00:53:35,130
Right.

711
00:53:37,010 --> 00:53:39,610
But if you, but then the only people who were harmed, the main people who

712
00:53:39,610 --> 00:53:42,810
were harmed were the people who were cutting corners while mining.

713
00:53:42,930 --> 00:53:46,729
So presumably that was beneficial to them for a long time and then they had to pay

714
00:53:46,729 --> 00:53:51,209
a cost and then it's like, what's intriguing about that is of course what

715
00:53:51,209 --> 00:53:56,329
they, they now do even better, I'm sure they must do, I don't actually know this,

716
00:53:56,370 --> 00:54:01,490
but it's just the reason that they would do a much better version where they've

717
00:54:01,490 --> 00:54:06,649
looked at that and they say, how do we get all the benefits of SPV mining while

718
00:54:06,649 --> 00:54:10,409
also not paying those horrible costs, which I actually think is just the,

719
00:54:10,409 --> 00:54:12,209
now they aggressively spy mine.

720
00:54:12,329 --> 00:54:15,050
They know if other people are on the same network and they probably also

721
00:54:15,050 --> 00:54:17,250
check it as fast as they can.

722
00:54:17,890 --> 00:54:21,050
And then yeah, I think they check after the fact, right.

723
00:54:21,050 --> 00:54:24,010
I think they check after the facts and if they turn out to be wrong, they'll,

724
00:54:24,169 --> 00:54:26,729
they'll try to often their own mistake kind of thing, right?

725
00:54:27,409 --> 00:54:27,890
Yeah.

726
00:54:27,890 --> 00:54:31,649
And so I think that we now live in a world, thanks to the fact that the nodes

727
00:54:31,729 --> 00:54:36,729
were vigilant in that case and punish the miners, we live in a world where

728
00:54:36,729 --> 00:54:40,370
everyone is, we get all the benefits of all the different worlds because we have

729
00:54:40,370 --> 00:54:44,010
every, there's no real, like everyone's, the whole mining network switches over

730
00:54:44,010 --> 00:54:46,810
to the next block, like basically as fast as humanly possible.

731
00:54:47,050 --> 00:54:48,169
So we have that benefit.

732
00:54:49,010 --> 00:54:53,529
That means less money is wasted on stale blocks or orphan blocks or

733
00:54:53,529 --> 00:54:57,529
whatever they're called these days, which means that network difficulty is

734
00:54:57,529 --> 00:54:58,130
higher.

735
00:54:58,169 --> 00:55:02,329
And it means that actually nothing, even though nothing is validated, like in

736
00:55:02,329 --> 00:55:06,769
time, it's still validated quickly enough that there's no reorgs even of the

737
00:55:06,769 --> 00:55:07,769
invalid blocks.

738
00:55:07,769 --> 00:55:09,610
And then everyone running their own node is still.

739
00:55:12,209 --> 00:55:16,130
I mean, I think that story does also highlight.

740
00:55:16,130 --> 00:55:23,370
So I think, you know, the thinking about the sequence of how you might want to go

741
00:55:23,370 --> 00:55:31,250
about activating a soft fork is that you ideally want lots of full nodes,

742
00:55:31,649 --> 00:55:36,089
preferably economically relevant ones, ones that people and services are

743
00:55:36,649 --> 00:55:42,690
transacting and verifying things through to be upgraded before the miners

744
00:55:42,690 --> 00:55:47,329
activate, because then if the miners make some kind of mining mistake, the

745
00:55:47,329 --> 00:55:52,050
network, you know, fixes it for them, ignores their mistake.

746
00:55:52,050 --> 00:55:53,849
Otherwise you create mayhem in the network.

747
00:55:53,890 --> 00:55:58,570
If the nodes are, like if the important nodes in the network are lagging and not

748
00:55:58,570 --> 00:55:59,849
upgraded, that can be problematic.

749
00:56:02,529 --> 00:56:03,810
So quite interesting.

750
00:56:04,930 --> 00:56:08,729
And I mean, it's very bad for running a full node if you don't, if you're like

751
00:56:08,729 --> 00:56:13,169
missing one of the, like you're missing like a file, like whatever, .h, whatever,

752
00:56:13,169 --> 00:56:14,329
.cpp, you know what I mean?

753
00:56:14,329 --> 00:56:21,089
Like there's like hidden validation rules that you don't have, that's not good.

754
00:56:22,250 --> 00:56:27,450
Well, I mean, I think there's a sort of another, I mean, there's a lot of defense

755
00:56:27,450 --> 00:56:31,729
in depth thinking that goes into Bitcoin soft forks as well.

756
00:56:31,729 --> 00:56:42,930
So I think usually the upgraded transaction is non-standard as well, so that it

757
00:56:42,930 --> 00:56:48,209
shouldn't propagate, like it will be accepted if it's in a block, but it won't

758
00:56:48,209 --> 00:56:52,089
propagate because it's not standard in a particular network.

759
00:56:52,089 --> 00:56:55,409
So like that sort of protects against some stuff, I think, indirectly.

760
00:56:56,169 --> 00:57:03,010
Yeah, I tried to, yes, I was saying in Lisbon, I tried to present something like

761
00:57:03,010 --> 00:57:06,729
this where I was trying to say we should stigmatize, there was like this meta

762
00:57:06,729 --> 00:57:11,290
consensus idea, and I was trying to say that sidechains are actually good because

763
00:57:12,569 --> 00:57:17,969
with a sidechain, you are putting the coins into a UTXO where it kind of is

764
00:57:17,969 --> 00:57:23,370
like being a non-standard where you're saying, we know on layer one that we

765
00:57:23,370 --> 00:57:24,650
don't know what these rules are.

766
00:57:26,490 --> 00:57:29,250
We know, we don't know what they are over there, but we know the BIP300

767
00:57:29,250 --> 00:57:30,929
rules, which are discounts of this number.

768
00:57:31,410 --> 00:57:35,890
So we were saying, don't complain over here, move the complaining over there.

769
00:57:36,410 --> 00:57:41,890
And I was thinking that that was the point I tried to make in my talk, consensus

770
00:57:41,890 --> 00:57:45,689
and dissent, but no one, I think understood it at all, unfortunately, except for

771
00:57:45,689 --> 00:57:47,569
maybe you, because I remember you asking a question.

772
00:57:48,089 --> 00:57:50,170
But other people, I think were confused.

773
00:57:51,290 --> 00:58:01,290
So going back to the like centralization risk argument, I think that was partly

774
00:58:01,290 --> 00:58:10,209
informed by the kind of arguments presented and the risks that were put as

775
00:58:10,209 --> 00:58:13,410
counter arguments to the big block proposals, right?

776
00:58:13,410 --> 00:58:21,290
And so I'm not sure whether it's Bitcoin Cash or one of the other ones, but one

777
00:58:21,290 --> 00:58:24,650
of the forks that didn't happen, but basically they wanted some kind of

778
00:58:24,650 --> 00:58:29,049
enormous block, like a hundred megabytes and a gigabyte or something, right?

779
00:58:29,209 --> 00:58:36,089
And my argument to some of the proponents of that, why that might be a bad idea is

780
00:58:36,089 --> 00:58:40,410
like, look, not that many people have fast enough machines and fast enough

781
00:58:40,410 --> 00:58:43,010
networks to keep up with that.

782
00:58:43,530 --> 00:58:45,530
Or it's like catch up with history and verify it.

783
00:58:46,449 --> 00:58:51,530
And so you'll end up with a kind of de facto federated system with, you know, a

784
00:58:51,530 --> 00:58:53,809
handful of nodes in high speed data centers.

785
00:58:54,370 --> 00:58:58,809
And, you know, potentially if that escalates over time, you get to a

786
00:58:58,809 --> 00:59:04,610
situation where it's more like a private database where nobody gets to see the

787
00:59:04,610 --> 00:59:07,089
history, like as a full node, as a power user.

788
00:59:08,130 --> 00:59:08,329
Yeah.

789
00:59:08,329 --> 00:59:10,650
We don't, you know, the miners don't give it to you basically.

790
00:59:11,130 --> 00:59:15,410
And they say, well, you know, trust the miner is the hash, is the mobile path for

791
00:59:15,410 --> 00:59:17,729
you and you don't need to know.

792
00:59:18,370 --> 00:59:24,290
And if that happens, you know, arguably the sort of centralized producers of,

793
00:59:24,490 --> 00:59:32,449
well, of blocks who would be collecting transactions together, they could take

794
00:59:32,449 --> 00:59:34,209
the fees and it'd be hard.

795
00:59:34,329 --> 00:59:38,650
It'd be a big barrier to entry to becoming a miner, you know, assuming they

796
00:59:38,650 --> 00:59:39,969
don't offer a pool service, right.

797
00:59:40,449 --> 00:59:43,209
Or even if they do offer a pool service, it'd be very hard to.

798
00:59:52,370 --> 00:59:57,049
Well, I mean, if somebody is offering a pool service, they can set the policy,

799
00:59:57,049 --> 00:59:59,410
right, so they could be censoring things or.

800
00:59:59,953 --> 01:00:13,953
You know, adding, increasing the fees or something. And so it's hard to sort of break in and add a new pool. So if you get, you want a certain amount of a low enough barrier to entry so that you can, so that people can start a new pool.

801
01:00:13,953 --> 01:00:16,953
That's one argument, right. And it's a form of.

802
01:00:16,953 --> 01:00:27,953
I've carefully been tracking this. I feel like I'm like a hunter in the wilderness and I have been tracking this argument also all around.

803
01:00:28,953 --> 01:00:37,953
What happened in 2015, I mean, the block size war started to heat up in 2015. And so I thought, well, this is interesting. I mean, surely it was a very interesting thing to happen.

804
01:00:38,953 --> 01:00:52,953
And I collect, like I did, like I would later do in, before the blind mining, I was collecting, I was like, what is this dispute really about? Because you probably remember my talk from scaling one when I had that thing about the principle.

805
01:00:52,953 --> 01:01:01,953
There was the principle that matter, like what is going on in this conversation was kind of like I was very interested in this topic.

806
01:01:01,953 --> 01:01:06,953
And so I was tracking down everything everyone had said, and I was trying to boil it down.

807
01:01:07,953 --> 01:01:16,953
And I did track it down with this big block. The problem with big blocks is it does, in fact, make it more expensive, more difficult to run the node.

808
01:01:16,953 --> 01:01:23,953
And the full node is checking the block for errors. So the error correcting on the block, I thought it was all about the node costs.

809
01:01:23,953 --> 01:01:34,953
And I wrote that post measuring centralization. And I said, this is all on the full node. This is all on the full node cost.

810
01:01:34,953 --> 01:01:41,953
And that is exactly how I came up with Drivechain, in fact, because I had worked all that out in September, and then we went to scaling one.

811
01:01:42,953 --> 01:01:48,953
And then I listened to everything in scaling one. And we did the breakout groups and we had all that stuff.

812
01:01:48,953 --> 01:01:58,953
And I was like, well, I didn't see any reason why my thing would be wrong. I thought it's all about the full node cost, because the mining cost, the miners already do a bunch of stuff.

813
01:01:58,953 --> 01:02:04,953
They already need ASICs. They need electricity. Half of them are getting fired all the time anyway.

814
01:02:04,953 --> 01:02:11,953
So I was just kind of like, the miners, you have chosen to go into an Olympic boxing ring or something.

815
01:02:11,953 --> 01:02:20,953
So I was like, you signed up to be a Navy SEAL or something. They're constantly raising the standards.

816
01:02:20,953 --> 01:02:25,953
You're going to try and play chess against the world champion.

817
01:02:26,953 --> 01:02:32,953
So I was thinking of miners kind of like that. But the nodes, I was like, oh, this node thing, that's something that everyone should do.

818
01:02:32,953 --> 01:02:40,953
That is the barrier of joining the network. So many of the arguments that people such as yourself made, I was putting into my head.

819
01:02:40,953 --> 01:02:48,953
And I was thinking, it's all about the node. And that's why Drivechain explicitly draws this exact line.

820
01:02:48,953 --> 01:02:53,953
It draws this completely like church and state separation line between the node and the miners.

821
01:02:53,953 --> 01:02:59,953
And it says, well, kind of the miners would be expected to do this. They farm this up for merge mining. No big deal.

822
01:02:59,953 --> 01:03:04,953
But then the nodes do not have to run the large block.

823
01:03:04,953 --> 01:03:14,953
And this is all related to the stuff that we were talking about before, because when you brought up by merge mining, I was kind of thinking like, I could come up with this idea.

824
01:03:14,953 --> 01:03:19,953
But now, looking back, I wonder if it actually just made people more and more confused.

825
01:03:19,953 --> 01:03:28,953
Because I really do believe it doesn't matter if the miners need to do something, because the miners need to do a lot.

826
01:03:28,953 --> 01:03:37,953
So now we've come full circle and come back to this intriguing idea, which we should talk about a lot, because I think it's a fascinating idea.

827
01:03:37,953 --> 01:03:43,953
Because plenty of people then went on to say at Scaling 3, they said, OK, we buy it, the node part.

828
01:03:43,953 --> 01:03:50,953
But they said the mining part is also important. And many, many people, David Harding, lots and lots of people that I could name, tons.

829
01:03:50,953 --> 01:03:54,953
Because I researched all, I was trying to find every time this had ever been mentioned.

830
01:03:54,953 --> 01:04:03,953
And to this day, it happens on Twitter, with Drivechain comes up, miner centralization is bound to also pop up.

831
01:04:04,953 --> 01:04:17,953
Well, I think it's fair to say that running a mining node is a lot more expensive than running a full node.

832
01:04:17,953 --> 01:04:29,953
Now, the node part is approximately the same, but miners are typically like a farm of them, like hundreds of kilowatts or megawatts or something, right?

833
01:04:29,953 --> 01:04:34,953
Even somebody's garage, they've upgraded their own power and they've got dozens of miners.

834
01:04:34,953 --> 01:04:45,953
And so they'll have more bandwidth because the pool protocols are super chatty and each individual miner is furiously exchanging work packets with the pool.

835
01:04:45,953 --> 01:04:52,953
So you're going to have a big power bill, you're going to have, each miner might cost $10,000.

836
01:04:53,953 --> 01:04:56,953
And so it's not a cheap activity.

837
01:04:56,953 --> 01:05:00,953
And so, you know, if…

838
01:05:00,953 --> 01:05:03,953
Yeah, yeah, completely different planet than what the node costs.

839
01:05:03,953 --> 01:05:14,953
Running the node might be expensive, but it's not going to cost, like one S9 will get you a full node for many years, even at probably a large block size.

840
01:05:15,953 --> 01:05:30,953
Right. So I think it's useful to sort of put a stick in the ground or try to articulate what's important and why.

841
01:05:30,953 --> 01:05:40,953
And I think it is important that people should be able to run a full node, catch up with the networks and catch the low cost.

842
01:05:40,953 --> 01:05:44,953
Nobody's arguing with that, right? Not your keys, not your coins, and it's important to be able to run a full node.

843
01:05:44,953 --> 01:05:53,953
And more people are running full nodes, I think, partly because of Lightning and partly because of the block size drama to protect themselves as a kind of shield, right?

844
01:05:53,953 --> 01:05:56,953
Your full node will keep you on the chain that you want to be on.

845
01:05:57,953 --> 01:06:04,953
If you're relying on some SPV thing, maybe some service disagrees with you or you just get taken along and you're not paying attention.

846
01:06:04,953 --> 01:06:07,953
So that's good.

847
01:06:07,953 --> 01:06:16,953
And then, you know, de facto, a mining farm is using a lot more compute and network resources for the non-mining part of it.

848
01:06:16,953 --> 01:06:37,953
But there are other examples of kind of incremental protocol proposals, some of them consensus critical, some not consensus critical, like Neutrino, let's say, which is putting some kind of bloom filter or map and associating with a block.

849
01:06:38,953 --> 01:06:41,953
So this potentially some kind of, or zero sync is another one.

850
01:06:41,953 --> 01:06:49,953
So there's all the idea of commitments to, which is those snapshots that are validated.

851
01:06:49,953 --> 01:07:02,953
So those are things that would, you know, cost a miner to create and verify and maybe cost the full node less because you're just consuming it, right?

852
01:07:02,953 --> 01:07:09,953
Yeah, I think the bloom filter is a very good, like, it's very similar to like how I think of the sidechain.

853
01:07:09,953 --> 01:07:16,953
It's like a service that you could do and it would cost something because the bloom filter costs whatever, something.

854
01:07:16,953 --> 01:07:20,953
I don't know what it is, but it costs like resources.

855
01:07:20,953 --> 01:07:26,953
Yeah, I mean, zero sync might cost more because it's doing like, you know, zero knowledge proof creation.

856
01:07:26,953 --> 01:07:33,953
And I think sometimes those proofs are more expensive to compute and verify depending on technology involved.

857
01:07:33,953 --> 01:07:35,953
I think it's true for that type.

858
01:07:35,953 --> 01:07:49,953
Where's the, you know, people, because if I bring up drivechain on Twitter, some people will sort of push back on it, but they don't push back on bloom filters, you know, like, so what's going on?

859
01:07:50,953 --> 01:07:57,953
Well, I mean, that's, I think it's just because thinking about security tradeoffs is complicated.

860
01:07:57,953 --> 01:08:00,953
It's unintuitive to most people.

861
01:08:00,953 --> 01:08:16,953
And so as a proxy for that, don't question everything and want nothing to materially change about the balance of things because you don't necessarily understand from first principles what would happen if those balances changed a bit.

862
01:08:16,953 --> 01:08:22,953
And maybe if some of them, you know, if we do think about the first principles, maybe some of them change a bit, that's perfectly reasonable tradeoff, right?

863
01:08:22,953 --> 01:08:27,953
And so you might argue that with, you know, some kind of bloom filter thing, right?

864
01:08:27,953 --> 01:08:41,953
Let's say it doubles the compute load for a mining tool node and adds 10% to a normal full node and it provides some benefit.

865
01:08:41,953 --> 01:08:50,953
Now you can sync to UTXO commitments or you can run a more, a nicer SPV client or something, right?

866
01:08:50,953 --> 01:09:01,953
So I think those are tradeoffs that people engage with without thinking, I mean, they're probably thinking about it, right?

867
01:09:01,953 --> 01:09:08,953
The people who work on this kind of protocols that it is a tradeoff, but they seem to be relatively uncontroversial tradeoffs as long as they're not enormous, right?

868
01:09:08,953 --> 01:09:17,953
You know, like the compute cost didn't go up by a factor of a thousand or something or make it too expensive to run a full node as a client.

869
01:09:17,953 --> 01:09:30,953
So I think that argument is, you know, that tradeoff is a reasonable thing to say that a mining full node could increase its cost a bit.

870
01:09:30,953 --> 01:09:36,953
And like verifying the sidechain of moderate size would fall into that bucket, if you ask me, right?

871
01:09:36,953 --> 01:09:44,953
Because as you say, they have typically far higher costs, even for somebody with, you know, a couple of miners in their garage or something.

872
01:09:44,953 --> 01:09:52,953
Yeah, you have to be very, very small. And then the other thing is, of course, each sidechain, you know, they could just drop it if it was too expensive.

873
01:09:53,953 --> 01:10:00,953
They don't have to do it. They could just say we only it's just like how Bloomfilter serve like the node services are not all mandatory.

874
01:10:00,953 --> 01:10:09,953
You know, like, yeah, so they can just drop them if they're too. And the other thing that I that I think is not is always some reason people commonly forget this.

875
01:10:09,953 --> 01:10:17,953
But because people imagine they would do a best worst case scenario, and they're going to imagine like a sidechain node that costs infinity or something.

876
01:10:17,953 --> 01:10:26,953
But but in order for the sidechain network really exist at all in any meaningful way, they have to be some people who are using it who are basically 0% miner.

877
01:10:26,953 --> 01:10:30,953
There's a sidechain full nodes. They have no hash rate and they have no interest in mining, you know.

878
01:10:30,953 --> 01:10:35,953
And the node has to be cheap enough for them or this network is just going to not exist.

879
01:10:35,953 --> 01:10:43,953
And so Solana or something, it has to it can't be like there is a limit before the sidechain will just die off, I think.

880
01:10:47,953 --> 01:10:57,953
Yeah.

881
01:10:57,953 --> 01:11:07,953
I mean, I think that is an interesting point about when, when it stops making sense to run a blockchain.

882
01:11:07,953 --> 01:11:15,953
I think like the whole the whole like the key value of a blockchain is that users can audit it.

883
01:11:15,953 --> 01:11:23,953
Not all users have to, but, you know, and for the main chain, you know, people want that audit cost to be relatively low.

884
01:11:23,953 --> 01:11:31,953
So it can be decentralized and lots of people can do it for a sidechain or drivechain and probably okay with that being a bit of a higher cost because opt in.

885
01:11:32,953 --> 01:11:44,953
And, you know, maybe people want to make a trade off where you need a desktop and a reasonably, you know, consumer bandwidth, but higher bandwidth, let's say.

886
01:11:44,953 --> 01:11:48,953
Right. And so I think people are okay with that trade off.

887
01:11:48,953 --> 01:12:04,953
But, you know, if if it gets to the point where a power user can't really validate the blockchain without moving into a data center, then it's not really a blockchain anymore, I would say, right?

888
01:12:04,953 --> 01:12:06,953
Or you're using the wrong data structure.

889
01:12:06,953 --> 01:12:09,953
If you got to trust the miners, you could make a more efficient protocol.

890
01:12:10,953 --> 01:12:18,953
That's your argument. Some of the big block camps, they actually argue that you should trust miners and try to like invert the security model.

891
01:12:18,953 --> 01:12:22,953
And because you do that, if you want to build something, have people opt into it.

892
01:12:22,953 --> 01:12:27,953
But I would argue it's kind of dumb because the wrong data structure, if that's what you want to do.

893
01:12:28,953 --> 01:12:32,953
I guess that's not really what they came up with.

894
01:12:32,953 --> 01:12:40,953
Because, of course, they were saying all those, you know, Roger Ver and all those people were saying, we don't mind trusting the miners a little bit more.

895
01:12:40,953 --> 01:12:51,953
And so so I was like, oh, that was that was partly why I was like, OK, well, we'll just they'll just use the system drivechain where.

896
01:12:51,953 --> 01:12:54,953
The miners are responsible for the withdrawal.

897
01:12:54,953 --> 01:12:58,953
They already they're already in for a little bit of trust in that direction.

898
01:12:58,953 --> 01:13:05,953
Well, much more really, because the drivechain is like a one hash every three months is a lot different than.

899
01:13:06,953 --> 01:13:12,953
Controlling the whole chain, like every message on the chain and the chain is completely dark and you can't see what's going on there.

900
01:13:13,953 --> 01:13:27,953
I mean, so I guess I guess the risk perception is that something could make a transition from being, you know, power user verifiable to not verifiable.

901
01:13:27,953 --> 01:13:34,953
And you can see the motivation, which is kind of like what is argued, right?

902
01:13:34,953 --> 01:13:38,953
Yeah, it's all good. But the the the validation is the node.

903
01:13:38,953 --> 01:13:47,953
It's all the node cost. So I always thought it was like you have to break it into different node costs and the miners, then they run whichever nodes they want.

904
01:13:47,953 --> 01:13:53,953
And if the sidechains are very, very giving them a lot of revenue and the revenue kind of covers their costs.

905
01:13:53,953 --> 01:13:57,953
And if other people, if that's where the big blockers want to go, then do I compare.

906
01:13:58,953 --> 01:14:01,953
See, this is important about that balance of power risks thing.

907
01:14:01,953 --> 01:14:10,953
Like I compare the large block sidechain to like Vita or Venmo or something like something where we have people who aren't using Bitcoin at all.

908
01:14:11,953 --> 01:14:17,953
They're probably not going to become cypherpunks tomorrow, but we can move them on the continuum.

909
01:14:17,953 --> 01:14:24,953
We want to move as many people on. And we especially want everything to be in Bitcoin, BTC, the coin, 21 billion coins.

910
01:14:24,953 --> 01:14:35,953
We want them to be owners. So I'm thinking, like, how do we just get those people in like right now and they can become small blockers next year?

911
01:14:35,953 --> 01:14:38,953
Because right now they're right now they're just they're fiat people.

912
01:14:39,953 --> 01:14:40,953
Infinite blocker.

913
01:14:42,953 --> 01:14:44,953
Well, I mean, I think you could see that.

914
01:14:48,953 --> 01:14:50,953
You know, a transition.

915
01:14:52,953 --> 01:14:54,953
From verifiable to non verifiable.

916
01:14:55,953 --> 01:15:11,953
Could be like moving from a big block drivechain to moving to Fedimint because you can't usefully audit what the federation is signing because some different technical reason, right?

917
01:15:11,953 --> 01:15:28,953
Because the blind signatures, you can never tell that the collusion of some threshold of the federation didn't like print extra coins themselves on the side and peg them out at a time before anybody noticed.

918
01:15:28,953 --> 01:15:29,953
Kind of thing, right?

919
01:15:29,953 --> 01:15:37,953
I mean, eventually you notice when it hits zero and there's still people with claims and complaining, but it could persist for a while.

920
01:15:37,953 --> 01:15:56,953
So I think that that would be that would be the risk rate that you could get to that situation because the drivechain block got bigger and bigger until then people running businesses that don't care or think, you know, some fractionalities in their interest to keep it quiet or something.

921
01:15:56,953 --> 01:15:57,953
Right.

922
01:15:58,953 --> 01:16:01,953
So I don't know if you I mean.

923
01:16:02,953 --> 01:16:03,953
I think sometimes people.

924
01:16:05,953 --> 01:16:06,953
Right. You know, the end users risk.

925
01:16:07,953 --> 01:16:12,953
That's you could tell a story about basically any any other L2, right?

926
01:16:12,953 --> 01:16:14,953
You could say there's a custodial thing.

927
01:16:14,953 --> 01:16:15,953
It works great for a while.

928
01:16:16,953 --> 01:16:18,953
The owner dies and gives it to his children.

929
01:16:19,953 --> 01:16:20,953
They rug pull it.

930
01:16:21,953 --> 01:16:22,953
Yeah.

931
01:16:22,953 --> 01:16:23,953
Or you could say lightning.

932
01:16:23,953 --> 01:16:24,953
Same thing.

933
01:16:24,953 --> 01:16:29,953
You could say people use lightning for a while, but over time, such and such happened.

934
01:16:29,953 --> 01:16:31,953
Custodial lightning took over or.

935
01:16:31,953 --> 01:16:38,953
Well, I mean, I guess I guess the lightning story could be there are so many channels open that there's not enough space in blocks to close them.

936
01:16:39,953 --> 01:16:53,953
Somebody starts like a mass, you know, hostile closes of, you know, hundreds of thousands of channels simultaneously, and there's not enough space blocks to to close them out to counteract them or something, you know, so you can make up.

937
01:16:54,953 --> 01:16:59,953
Would be if we had a magical technology that worked perfectly for everyone that everyone liked and that everyone understood.

938
01:17:00,953 --> 01:17:14,953
But if we don't have that magic technology, then it seems to me the obvious thing that we should do is that each end user should be like presented with like some kind of menu of options.

939
01:17:15,953 --> 01:17:16,953
Yeah.

940
01:17:16,953 --> 01:17:24,953
Well, I mean, I think that probably, you know, because what I'm describing in terms of centralization failure, I think people are kind of.

941
01:17:25,953 --> 01:17:37,953
Have learned or adopted a philosophical viewpoint informed by Bitcoin, blockchain, logic and consensus rules that any kind of.

942
01:17:38,953 --> 01:17:51,953
Desired outcome should be enforced by a smart contract, like the fact that your full nodes verifying the network should, you know, at least game theory or mathematically prevent the bad thing from happening.

943
01:17:51,953 --> 01:17:57,953
But in reality, many things you can't really do that to right just doesn't happen to work out.

944
01:17:58,953 --> 01:18:05,953
And so, you know, you could always say, well, you know, people could go go nuts and he's going in a bad direction.

945
01:18:06,953 --> 01:18:24,953
And then this, this, you know, this particular chain or layer two or arc or lightning or something could it's it's kind of self-sovereign, you know, natural withdrawal logic could get degraded by something excessive.

946
01:18:24,953 --> 01:18:25,953
Right.

947
01:18:25,953 --> 01:18:36,953
And, you know, question is like, yeah, you've probably construct arguments like that for many existing systems, right?

948
01:18:36,953 --> 01:18:45,953
Like probably you could make one for lightning or for two way peg sidechain with fraud proofs.

949
01:18:45,953 --> 01:18:52,953
You know, if I guess the unilateral withdrawals could get too many to take out or something like that's another risk.

950
01:18:53,953 --> 01:18:59,953
So I guess we just have to evolve incrementally.

951
01:18:59,953 --> 01:19:00,953
Yeah.

952
01:19:00,953 --> 01:19:19,953
I mean, I think like you've got to say like, you know, I mean, I think one argument is like technology is too complicated to expect a silver bullet for everything that is future proof for all possible, you know, evolution and the way that usage evolves and limits evolve.

953
01:19:19,953 --> 01:19:33,953
So realistically, technology evolves incrementally and, you know, when it gets close to problems or limits, you know, people get invented and they find new things.

954
01:19:33,953 --> 01:19:53,953
So it sort of evolves biologically. So we shouldn't, shouldn't, you know, give up or reject things because, you know, there are hypothetical things that could happen in the future because we don't know the future and new ideas are evolving all the time.

955
01:19:54,279 --> 01:20:08,279
They happen only to the people who have decided to use each L2, so to speak. So like, that is how it should be, right? Like that is, that's, that's a nirvana. I mean, is it not right?

956
01:20:08,279 --> 01:20:37,279
Yeah, well, I mean, I think that's, that was what was being discussed at scaling one. And it's unfortunate that, you know, these large block forks didn't actually get packaged up as a large block opt-in sidechain, because that would allow people who want that tradeoff to get what they wanted, right? And wouldn't need a default drummer. And it would be an okay tradeoff.

957
01:20:38,279 --> 01:20:44,279
You know, up to some size, you know, as long as our user can verify stuff, I think it's a fair tradeoff.

958
01:20:44,279 --> 01:20:56,279
And just so we can be clear, you're saying that BIP300 could have prevented the entire Bitcoin Cash civil war.

959
01:20:57,279 --> 01:21:17,279
Well, that's what I thought. Like a number of us tried to suggest that to some of the big block proponents. And I think the typical argument was that it's too slow and they wanted to, you know, everything was urgent. And that would require technology.

960
01:21:17,279 --> 01:21:20,279
Imagine where we could be today.

961
01:21:21,279 --> 01:21:43,279
I think the large blockers would be better off today, by far, like way, much, much better. I think everyone would be better. But I think one of the weird things they did with many of their, many of their, you know, whatever stunts or attempts, I think just discredited themselves, basically unfairly.

962
01:21:43,279 --> 01:22:01,279
So some, like a lot of them, I think, like they didn't have a better idea, but they just lost so hard. And also the other thing, they discredited many other things. Like, I think the hard fork is actually not as bad of an idea.

963
01:22:01,279 --> 01:22:28,279
I think it was like kind of like everyone gets free money. We see another idea in the wild. We try it. But because of the circumstances of the first big hard fork was like trying to name it Bitcoin and stuff, I thought was actually made it so that this institution or this idea of the hard fork could like basically never be used again or, or whatever.

964
01:22:29,279 --> 01:22:31,279
Kind of like was very confusing.

965
01:22:32,279 --> 01:22:52,279
So, so can I ask something related to you, Mr. Black? What do you think the, the risks are today of, of ossification? Let's just say that we, you know, I guess organically, so to speak, everything is, is ossified, whether or not we realize.

966
01:22:53,279 --> 01:23:02,279
What do you think the risks are of something like that happening again and not being able to be avoided or other risks?

967
01:23:02,279 --> 01:23:04,279
Can you still hear me?

968
01:23:04,279 --> 01:23:05,279
Yeah, I hear you.

969
01:23:05,279 --> 01:23:06,279
Yes, we do.

970
01:23:06,279 --> 01:23:32,279
Okay. Yeah. Yeah. The thing looks frozen, but it's still working. Okay. Well, yeah, I mean, I think Bitcoin still has insufficient modularity and lacking features on layer one for like, you know, bolts and covenants and extensibility and modularity.

971
01:23:32,279 --> 01:24:01,279
So I think, you know, with some of those things built in, it could, it could make sense to have an ossified, like fairly ossified base layer, because then there would be enough modularity and extensibility that other things could be, you know, evolved by programming without needing to change, you know, opcodes and, or add opcodes and things to be done in layer twos there.

972
01:24:02,279 --> 01:24:04,279
You know, make different security tradeoffs, I guess.

973
01:24:04,279 --> 01:24:18,279
So I think it would be unfortunate if, if, if ossification is too close, you know, if we're stuck where we are, I don't think that would be an unfortunate situation.

974
01:24:19,279 --> 01:24:43,279
Yeah, I think like people say, you were talking about that, that balance of power thing, which idea, which I think is 100% correct, which is like people say, say, well, we don't want to disturb the way things are too much. But I do really wonder about like, I kind of always think, well, isn't it disturbing to have to have a SegWit2x happen or to have a Bitcoin Cash exist at all or to have altcoins exist?

975
01:24:43,279 --> 01:24:52,279
Like, every day, the people in Ethereum, they wake up, and they have one thought on their mind.

976
01:24:53,279 --> 01:25:12,279
This is the same thought they've had since 2016, the whole Ethereum community. And people try to avoid acknowledging this, but to me, it's clearly the truth that they are thinking every single day, how do we flip Bitcoin and take over?

977
01:25:12,279 --> 01:25:20,279
And it's like, a lot of that if you just had the ETH sidechain, they would just be like, there would be no...

978
01:25:21,279 --> 01:25:47,279
Well, I mean, maybe they do, like they're probably thinking of how to pump and dump their ICO, like all altcoins. But, you know, personally, I think it's better to, you know, focus on innovation and not look over your shoulder at what crazy people are doing, like regulatory risk takers and stuff.

979
01:25:47,279 --> 01:26:03,279
But I also think it's implausible that something so pre-mined and centrally controlled and insecure and buggy could be digital gold, right? Like, I think Bitcoin is too far ahead.

980
01:26:04,279 --> 01:26:21,279
I said that type of thing in 2015, 2016, 2017. But now, years later, I think people show up, and to them, Bitcoin just looks pre-mined to them. They just say, 1991. They have no real – from their point of view, they don't really see a difference.

981
01:26:22,279 --> 01:26:39,279
Well, I mean, I think there is a phenomena, which is probably like, you know, human psychology or something, to look at something that has been around longer, and had a meteoric rise, and feel that you missed out, right?

982
01:26:39,279 --> 01:26:56,279
And so to try and jump on something that looks early and with some good marketing, you know, and try to get, you know, try to play with that. And so you see newcomers go through that experience sometimes and then get burnt by, you know.

983
01:26:56,279 --> 01:27:13,279
But I think it's also like a bit unstable because there are many, you know, there are always new chains trying to – so I think like these sort of layer two – sorry, not layer two, but these sort of smart contract chains are like kind of ballisticly competing with each other, you know.

984
01:27:13,279 --> 01:27:37,279
There's dozens of them, right? And they're all – you know, I think you have the same effect with Ethereum, right? The people that were in the Ethereum pre-mine, which literally was a pre-mine, are looked at, disfavorably, by newcomers who then are persuaded to go into the next pre-mine, like Solana or whatever. There are dozens of them, right?

985
01:27:38,279 --> 01:27:54,279
So, you know, it is what it is. But I think the other thing that's maybe more interesting is that, you know, when apparently in 2015 when the sidechain paper was released, the price of altcoins went down across the board.

986
01:27:54,279 --> 01:28:10,279
Now, I wasn't paying attention to altcoin prices, so I don't know if that was coincidental, but apparently it happened, right? And so I think that maybe people were expecting that, oh, no, you know, they believed the feature story, you know, the feature coin story that the value is because of the features.

987
01:28:10,279 --> 01:28:31,279
And if Bitcoin was going to take all the features, then it would be harder to market the coins. But I think ultimately these things are, you know, the altcoins, 20,000 more, 25, I don't know what the number is now. It's a kind of mutually understood lottery ticket or casino game, right?

988
01:28:32,279 --> 01:28:48,279
So, you know, so with the, you know, some of the drivechains that you've built as examples, you know, you might think, well, no chain would be able to bootstrap itself without feature now because you don't need to use it on Bitcoin, right?

989
01:28:48,279 --> 01:29:12,279
But I think that misses the point that, you know, there are people that want to gamble on those things, right? Or they want to, you know, get connected to the insiders, get cheap coins early, get discounts for promoting it, etc. And there's a whole ecosystem of people trying to make money off of that. And really what they're doing is, you know, trying to separate people from their Bitcoin or from their dollars, right?

990
01:29:13,279 --> 01:29:14,279
So I think that ecosystem…

991
01:29:16,519 --> 01:29:16,919
Well,

992
01:29:18,259 --> 01:29:20,979
lot of people, I think a lot of people are doing that where they know there's some

993
01:29:20,979 --> 01:29:25,579
people who are deep down their Bitcoiners and those people are trying to maximize the

994
01:29:25,579 --> 01:29:30,659
amount of Bitcoin, but there's different people who are the deep down who are Ethereum

995
01:29:30,659 --> 01:29:34,979
people and those people want to acquire Ethereum and then flip

996
01:29:43,279 --> 01:29:46,279
…does a lot of trial and error and a lot of R&D.

997
01:29:47,279 --> 01:29:55,279
Yeah. I don't think so, but, you know, they can hold on to their copium, right?

998
01:29:56,279 --> 01:29:58,279
But anyway, like…

999
01:29:59,279 --> 01:30:00,279
Sagan?

1000
01:30:01,279 --> 01:30:08,279
Well, I was just going to say, the first part of what you said seemed to say, like, bleeping the feature story and having the altcoin prices crash…

1001
01:30:09,279 --> 01:30:10,279
Yeah.

1002
01:30:10,279 --> 01:30:24,279
…when the sidechain paper came out, that I thought you were going in a different direction with your comments. And I thought you were going to say, aha, this proves that altcoiners know that once… the more progress sidechains make, the more doomed they are. And the only reason why they haven't crashed to zero is because…

1003
01:30:24,279 --> 01:30:25,279
Well…

1004
01:30:25,279 --> 01:30:28,279
…it's dragging its feet on the sidechains. So I thought you were going to go in a completely different direction.

1005
01:30:28,279 --> 01:30:42,279
Well, no. I mean, I think people's expectation maybe shifted over time. So I think that altcoin price crash was not coincidental. I think it was perception at the time.

1006
01:30:43,279 --> 01:30:56,279
But then, you know, I mean, of course, sidechains have been slower to, you know, go live and have new features on them, and have a variety of features, right, that are actually live.

1007
01:30:57,279 --> 01:31:14,279
But I think also, you know, probably that market has realized that it doesn't really matter, ultimately. I mean, like, some of these multi-billion dollar chains are literally copies of other chains, right?

1008
01:31:15,279 --> 01:31:32,279
Like, the technology is the same, or very small tweaks. There's a different set of promoters and pre-mine owners, and they're competing with each other. And, you know, some of them have taken, like, central use cases away from Ethereum and have higher volume than it.

1009
01:31:32,279 --> 01:31:56,279
And probably that will change when they start to get overloaded or something, right? So I'm thinking, my thinking about, you know, altcoin economics, because it concerned me from the point of view, you know, of efficient market hypothesis, that if you think these things are silly, and I do, why isn't the free market able to fix it, right?

1010
01:31:57,279 --> 01:32:20,279
And so that was perplexing. And I wanted to, like, figure out a way to short them and, you know, profit from correcting this, you know, defective market situation. And so eventually I came to the conclusion that, you know, it's just a different market. It's a lottery ticket market, and people like gambling.

1011
01:32:21,279 --> 01:32:40,279
And so my revised theory is that, you know, 95% or more people that are participating in it have no illusions that the altcoins they're speculating are going to be digital gold, or that they're going to last more than a few years, and they're just trying to, you know, play the musical game, musical chess game.

1012
01:32:40,279 --> 01:32:42,279
And, you know,

1013
01:32:43,878 --> 01:32:47,358
to do

1014
01:32:53,358 --> 01:32:59,178
you you to be able to do able to

1015
01:32:59,178 --> 01:32:59,558
able

1016
01:32:59,718 --> 01:33:03,758
sidechain, Counterparty sidechain, Ordinal sidechain, whatever you're going to it,

1017
01:33:03,778 --> 01:33:08,998
something with you have, you can just make stuff all the time, like you can make a

1018
01:33:08,998 --> 01:33:13,618
lottery or you can make NFT, ICOs.

1019
01:33:13,958 --> 01:33:15,479
I mean, in some sense we do see this all

1020
01:33:15,539 --> 01:33:20,859
the time, see ICOs NFTs which are just like for fun, but if that were

1021
01:33:20,859 --> 01:33:26,279
like, if that were existed, there's no reason why

1022
01:33:26,279 --> 01:33:28,819
there'd be no reason for the altcoins to exist.

1023
01:33:29,179 --> 01:33:31,199
In order for the story to actually work, has

1024
01:33:32,279 --> 01:33:51,279
I mean, I think they need, you know, because you could say, well, why waste the bandwidth and the disk space on a blockchain? You know, you can have, like, centrally operated, provably fair poker games and casino games. This is known, like, pre-blockchain technology, right?

1025
01:33:52,279 --> 01:33:54,279
And why don't they use that?

1026
01:33:54,279 --> 01:33:55,279
Awesome. Yeah.

1027
01:33:56,279 --> 01:34:15,279
So then the argument is, well, why don't they do that? It would be cheaper. You know, you could make, you wouldn't be restricted by technology, right? You could make up whatever crazy rules you wanted and have a central server, provably fairly enforce them in some kind of, you know, role playing, gambling game.

1028
01:34:16,279 --> 01:34:25,279
But I think that doesn't work because they probably want the sort of illusion of reality, even if they all know it's kind of, doesn't matter.

1029
01:34:25,279 --> 01:34:35,279
Yeah, that's what I think too. But I think that is still something that if sidechains became mainstream, that would actually still crush everyone. It would just not, it would be too clear and it would do it.

1030
01:34:36,279 --> 01:34:49,279
But I have another interesting idea, if you want, a theory that you could use on your side, that this is actually all, that it is not, it's all for gambling and actually sidechains wouldn't kill it necessarily. It depends.

1031
01:34:49,279 --> 01:35:01,279
So my theory is that it's actually too difficult for most people to just create their own, like, PokerStars server or something. But if there's, remember, like, didn't Matt Corral made, like, the altcoin generator or something?

1032
01:35:01,279 --> 01:35:04,279
Yeah, I mean, that also had the effect, actually.

1033
01:35:04,279 --> 01:35:15,279
Yeah, so it's actually like, people actually, it is possible, especially one, the same person can make like 50 altcoins, they can make 500 altcoins, you know, they figure it out, they just keep doing it.

1034
01:35:15,279 --> 01:35:29,279
And so it actually is just literally easier to launch an altcoin, like we, you and I might think of it as kind of incredible because we think of like the long period before Bitcoin and all the numerous failed attempts to create e-currency.

1035
01:35:29,279 --> 01:35:42,279
But we would think it's bizarre, but actually once Bitcoin and Litecoin are out, it's very easy to just, some guy can be making thousands and that could be the same single guy who made all 20,000 of these altcoins.

1036
01:35:42,279 --> 01:35:51,279
Oh, and there are certainly people that are serial altcoin creators, you know, there are people that infamously create half a dozen of them, right?

1037
01:35:52,279 --> 01:36:08,279
Yeah, they learn, of course, they're evolving in the direction of how to shill a worthless altcoin the best way. So their lore and their science is going in that direction.

1038
01:36:08,279 --> 01:36:24,279
Well, I mean, I have a theory that it's a kind of lemon market. So what for Bitcoin or normal technology would be defective actually becomes desirable for a casino coin.

1039
01:36:24,279 --> 01:36:39,279
So, you know, like if you say, well, these guys heavily pre-mined it, like 80, 70, 80, 90% pre-mined and they're completely ethical and they're wash trading it and they're untrustworthy.

1040
01:36:39,279 --> 01:36:50,279
And yet the market would say, well, that's great because we can really trust that they're going to pump the heck out of it. And then we can ride along for it and dump it, try and dump it when they do.

1041
01:36:50,279 --> 01:37:00,279
And so, you know, something that looks like a hive of villainy actually becomes an attractive thing for people to speculate on.

1042
01:37:00,279 --> 01:37:19,279
And so you get this kind of weird, perverse, like high level features of it, right? Where, you know, like a top trumps game where you're like comparing supercars, which one's better, you know, has more torque, more horsepower, more cylinders, etc, right?

1043
01:37:19,279 --> 01:37:40,279
But in this game, you know, a very, what is it, safety or somebody's, you know, it was Michael Saylor, the shark duck in a tree or something, right? So you get these crazy creations that can be successful casino tokens, but they make no sense.

1044
01:37:41,279 --> 01:37:57,279
Like, you know, defective technology, defective incentives, but it turns out they're not effective. They're just optimized for, they're optimized for gamified, you know, casino action, right? And so we're looking at it in the wrong, wrong way. And it's a different market.

1045
01:37:57,279 --> 01:38:04,279
Well, I think if this is the case, though, what's the end point where, when, when does that stop?

1046
01:38:04,279 --> 01:38:07,279
I don't know, like nowhere good, I would say.

1047
01:38:07,279 --> 01:38:33,279
I think I have a decent metaphor, which is like, you know, the Cambrian explosion, that metaphor has been used, but it's kind of like, once you have life, like, it will kind of like never stops to some extent, where it's like, you have some things evolve, like Bitcoin evolves, it exists, and then people do the hard work of proving, like, because even Bitcoin was kind of going nowhere before Silk Road.

1048
01:38:33,279 --> 01:38:44,279
You have to build up like the whole experience first, like someone actually using it for something, the whole gear of it, like WikiLeaks or whatever, and the word gets out.

1049
01:38:45,279 --> 01:39:10,279
And people, it doesn't just fracture into a million other pieces immediately, like there's a community behind it. So the Bitcoin evolves, it proves that crypto can work. And then it's like the Scambrian explosion is like you have bacteria evolving to infect the, I mean, the bacteria obviously existed before, but you have like infectious bacteria that infect larger animals or larger creatures.

1050
01:39:11,279 --> 01:39:27,279
And so the whole thing is like, you have different species. So like, I think like, you have like your Bitcoin, then you have like your copycat, like a Litecoin, then you have like people like maybe like a Dan Larimer, who is like a serial true believer, you know, when he's actually making something.

1051
01:39:27,279 --> 01:39:52,279
You and I don't love it, maybe per se, but he, he's like making stuff. And then you have like, whatever, Solana, and then you have like, whatever, you have like stuff where it's like evolving in different, it's there's different speciation, you know, like we have plants, we have fungi. You know what I mean? So that's what I'm trying to say is, it doesn't.

1052
01:39:53,562 --> 01:40:00,122
Ever stop, but you do get clear species, they have like their little like, they have like their

1053
01:40:00,122 --> 01:40:09,322
Well, I mean, the way I'm thinking about it is apparently some countries have kind of societal

1054
01:40:09,322 --> 01:40:16,762
scale problems with pyramid scams. You know, like, I guess, a lot of people are kind of a gambling

1055
01:40:16,762 --> 01:40:25,561
tendency, a lot of participation in penny stocks, and pyramid scams. And so it runs through the

1056
01:40:25,561 --> 01:40:29,482
country and the country has like social problem with it and trying to stop it and stuff, right.

1057
01:40:30,201 --> 01:40:38,361
And so, you know, I think those things, if they become too economically impactful, and burning,

1058
01:40:39,482 --> 01:40:45,402
you know, 10s or 100s of billions of dollars, at some point, somebody has to do something,

1059
01:40:45,402 --> 01:40:51,402
because it starts, you know, creating economic problems at a country scale or something, right.

1060
01:40:52,122 --> 01:40:59,642
So I mean, I think the Terra Luna collapse was like $60 billion of retail money. And that caused

1061
01:40:59,642 --> 01:41:03,801
a bunch of regulators to wake up and try to do something about it, right, of that particular one.

1062
01:41:04,441 --> 01:41:10,682
But I think, you know, the, the risk, you know, unless unless it reaches an equilibrium by itself,

1063
01:41:11,642 --> 01:41:21,242
if the kind of gambling thing is, you know, a scope to grow, it could start to create a problem

1064
01:41:21,242 --> 01:41:28,521
for like, you know, allocation of capital in the world, and lead to an idiocracy outcome. And so

1065
01:41:28,521 --> 01:41:31,642
I think there's somewhere along the way, if it just gets more and more excessive, like,

1066
01:41:32,521 --> 01:41:34,602
you know, somebody has to do something about it.

1067
01:41:35,002 --> 01:41:39,082
I have sympathy for the regulators, some, I mean, somewhat, I mean, I don't have very much,

1068
01:41:39,082 --> 01:41:43,082
but I because I think they mostly, they screw up, they let the bad projects through,

1069
01:41:43,082 --> 01:41:47,402
they harass good projects often. But like, if I go out into the real world, and then,

1070
01:41:48,201 --> 01:41:53,642
like, you must have experienced this also, where you're talking to someone, and they they have a

1071
01:41:53,642 --> 01:41:58,121
normal job, but they're, they're kind of smart, like they're like a realtor, or like an accountant,

1072
01:41:58,121 --> 01:42:03,401
even or something. And they're like, they'll say, like, they'll say, Oh, you know, about crypto,

1073
01:42:03,401 --> 01:42:07,882
you know, about crypto, like I'm buying doggy coin, they don't know it's pronounced Doge.

1074
01:42:08,601 --> 01:42:12,601
I'm buying doggy coin. And they're like, yeah, I just needed to hit 10 cents or something. I

1075
01:42:12,601 --> 01:42:17,321
just needed to. And yeah, like, these are like, that's like with normal people. Like,

1076
01:42:18,121 --> 01:42:23,321
it's like many, many normal people who would be considered successful adults.

1077
01:42:24,041 --> 01:42:30,201
Well, I mean, people buy lottery tickets to write, you know, they do, you know,

1078
01:42:30,761 --> 01:42:38,601
sell things for amusement, or, you know, it might work, or

1079
01:42:38,601 --> 01:42:42,922
you know, it's all negative expected value. Like, as a former professional statistician,

1080
01:42:42,922 --> 01:42:49,481
I walk into the casino, I see in my in my, in my mind, when I look at the different

1081
01:42:49,481 --> 01:42:55,002
games, I feel like negative 4%, negative 7%, negative 11%, like, negative 36%.

1082
01:42:55,802 --> 01:42:59,082
For like, whatever, for like crafts with blackjack. So I just see it all. They're

1083
01:42:59,082 --> 01:43:03,882
all negative numbers, though, you know? So it's like, you're the guy that wouldn't buy them,

1084
01:43:03,882 --> 01:43:08,121
right? Why is anyone here? Why is anyone in this room at all? Like, you should pick,

1085
01:43:09,082 --> 01:43:11,481
pick the highest numbers on screen, but they're all red.

1086
01:43:13,321 --> 01:43:18,522
You know, they get people, they're in the casino, they could go on TD Ameritrade or whatever,

1087
01:43:18,522 --> 01:43:23,961
and just start trading stocks. And it would be a casino where the on average, they'll make seven

1088
01:43:23,961 --> 01:43:31,481
or so percent. Yeah, well, actually, you can see some kind of in terms of the propensity for

1089
01:43:31,481 --> 01:43:35,561
people to do stupid stuff with stocks, you can see the kind of game stock thing, right? You get

1090
01:43:35,561 --> 01:43:42,041
some kind of defunct stock, which is heavily shorted, and rightly so. And some guys on Reddit,

1091
01:43:42,041 --> 01:43:48,601
like short squeeze it, and bankrupt some hedge funds. So, you know, if you want to do stupid

1092
01:43:48,601 --> 01:43:53,642
stuff, they can find ways to sort of create gambling opportunities out of the stuff, right?

1093
01:43:57,162 --> 01:43:59,242
Anyway, I mean, we should come back to...

1094
01:44:03,242 --> 01:44:08,282
Maybe we should talk about like, so I would love to, we didn't really get to the whole break into

1095
01:44:08,282 --> 01:44:13,962
the mining pool business, which is a thought that I have been hunting and tracking and thinking

1096
01:44:13,962 --> 01:44:18,122
about for a very long time. But another thing I wanted to be sure to ask you about is like,

1097
01:44:19,242 --> 01:44:24,362
you know, I can have like this certain, I don't have like a big, it's not like a crazy strategy,

1098
01:44:24,362 --> 01:44:28,842
but like for a drivechain, I've been like putting out the software, we're trying to put out the

1099
01:44:29,642 --> 01:44:34,682
BIPs, we're trying to put out a pull requests, we're trying to put out more info on our site,

1100
01:44:34,682 --> 01:44:39,802
we're collecting people's names for that layer2labs.com slash friend list. And like I'm on

1101
01:44:39,802 --> 01:44:48,122
Twitter, we do the spaces. And I just wanted you to maybe just like, what do you think we should

1102
01:44:48,122 --> 01:44:54,682
do differently, better, whatever? Is everything going perfectly? Is this just as fast as it goes?

1103
01:44:54,682 --> 01:44:59,642
Is this, should we do something else? You know, what do you think of that?

1104
01:45:00,601 --> 01:45:09,881
Well, I mean, bear in mind that I'm not really, I'm more like somebody that looks at protocols

1105
01:45:09,881 --> 01:45:15,161
and things. So I haven't actually, and of course I got like active because I got concerned about the

1106
01:45:15,881 --> 01:45:22,201
block size drama, right? But otherwise I'm more like applied crypto and other people are doing

1107
01:45:22,201 --> 01:45:26,201
that kind of stuff. But you know, I'm an observer of it, the same as you and everybody else that's

1108
01:45:26,201 --> 01:45:33,001
been in the system for a while. So my impression is that things can take, I mean, one aspect is

1109
01:45:33,001 --> 01:45:41,161
things can take a long time because Schnorr signatures were coincidentally first talked about

1110
01:45:41,161 --> 01:45:49,402
in 2013. I was talking about it with some of the developers on IRC, this is before Blockstream,

1111
01:45:50,041 --> 01:45:58,761
and that only got its way into Bitcoin with a tap route then quite recently, right? And that was,

1112
01:46:00,041 --> 01:46:06,522
you know, I guess what caused that to happen, it wasn't me pushing it, right? It was like,

1113
01:46:06,522 --> 01:46:10,601
you know, people got to thinking about it, like, well, that's cool. Look what you can do with it.

1114
01:46:10,601 --> 01:46:15,721
And then, you know, technical people start to implement things, implement it in libraries,

1115
01:46:16,522 --> 01:46:23,321
work out specs for how you could use it. And, you know, got enough people enthusiastic that

1116
01:46:23,321 --> 01:46:27,642
they made it happen, right? At least got the technical pieces together and

1117
01:46:29,481 --> 01:46:35,642
implementation and bits and stuff and able to get consensus to activate it

1118
01:46:36,601 --> 01:46:43,642
post, I guess, the most recent soft fork after the SegWit one, which got

1119
01:46:45,082 --> 01:46:47,721
sort of randomly strapped in the block size drama, right?

1120
01:46:49,642 --> 01:46:54,041
Otherwise, I would say that SegWit should have been non-controversial too, to my perspective.

1121
01:46:56,201 --> 01:47:04,122
So I think, like, the timeframe is kind of interesting, right? That that's quite slow,

1122
01:47:04,122 --> 01:47:10,281
the things to happen organically. And the other kind of observation is just more

1123
01:47:11,881 --> 01:47:19,881
like just looking at it, that it seems like you need, like, multiple independent people to be

1124
01:47:19,881 --> 01:47:28,682
excited about it and involved in discussions and things. I think in most technology fields,

1125
01:47:28,682 --> 01:47:34,522
there's a kind of, you know, trickle down of knowledge and viewpoints, right?

1126
01:47:35,321 --> 01:47:41,082
There's not that many people that have the expertise, maybe, to...

1127
01:47:43,481 --> 01:47:48,041
Yeah. Well, there's a little bit of a chicken and egg problem with that, which is that

1128
01:47:48,041 --> 01:47:52,441
I think more people would be excited about it if we could use it somewhere. Some people say,

1129
01:47:52,441 --> 01:47:56,842
oh, launch this all on Litecoin or something. I honestly think if we did that, then there would

1130
01:47:56,842 --> 01:48:02,362
be some probability that Litecoin would become instantly very, very cool, because it would have

1131
01:48:02,362 --> 01:48:08,122
a Zcash sidechain, ETH side. Who knows what would actually happen in that? It may cause chaos among...

1132
01:48:11,481 --> 01:48:18,842
But people have suggested something like that. Like, I think, what should, like,

1133
01:48:20,362 --> 01:48:25,961
is there anything, like, you just think we should just stay the course or what would you, you know?

1134
01:48:25,961 --> 01:48:32,762
Well, I mean, you can look at another conversation that's happening right now, which is the

1135
01:48:34,922 --> 01:48:42,202
covenants discussion, which is quite an interesting analogy because it does enable

1136
01:48:43,242 --> 01:48:46,682
quite a lot of new potential features, including potentially drivechains in a kind of,

1137
01:48:47,562 --> 01:48:53,802
you know, inefficient, many, many, many long scripts way, perhaps. But, you know, it does

1138
01:48:53,802 --> 01:48:58,601
seem like Bitcoin's, you know, the fact that Bitcoin covenants are not possible today seems

1139
01:48:58,601 --> 01:49:02,202
to be a kind of historical accident. Like, if certain opcodes hadn't been disabled, it would

1140
01:49:02,202 --> 01:49:08,202
have already been possible. It's just sort of side effects. So, you know, the limits of

1141
01:49:08,202 --> 01:49:14,442
programmability seem to be a bit ad hoc, and certainly covenants would unlock a lot of

1142
01:49:14,442 --> 01:49:21,322
potential, right? And so it's interesting to look at the, you know, how that's evolving,

1143
01:49:21,322 --> 01:49:26,762
discussions around that. So I think, you know, CTV kind of got pushed a bit too hard and

1144
01:49:28,041 --> 01:49:36,202
sort of turned some people off, but there's still an interest in seeing covenant functionality

1145
01:49:36,202 --> 01:49:43,562
for what it enables, right? I mean, the general wider extensibility plus vaults are quite interesting.

1146
01:49:44,521 --> 01:49:50,041
What lesson do you think I should learn from CTV experience? I mean, the way it looks to me is kind

1147
01:49:50,041 --> 01:49:59,081
of like that everyone is crazy, and what you need to do is quit the community for two years, and then

1148
01:50:00,041 --> 01:50:09,481
I think what made that controversial was it went from a discussion. I mean, I guess the annoying

1149
01:50:09,481 --> 01:50:14,841
thing for people that were interested to see covenants is that there wasn't, you know, there

1150
01:50:14,841 --> 01:50:19,081
wasn't that much of a critical mass of people tracking and commenting and trying things out

1151
01:50:19,081 --> 01:50:25,401
and proposing alternatives, right? So, but I think it's interesting to look at the way it's

1152
01:50:25,401 --> 01:50:31,641
evolving, right? I mean, there wasn't that much of a critical mass of people tracking and

1153
01:50:31,641 --> 01:50:40,281
proposing alternatives, right? So, but I think what caused that to flare up was that they kind

1154
01:50:40,281 --> 01:50:45,882
of skipped to, okay, well, let's activate it. Let's have a discussion about how to activate it.

1155
01:50:46,841 --> 01:50:51,882
And, you know, presumably discussion about how to activate it, you would normally expect that

1156
01:50:52,202 --> 01:50:58,922
this is the proposal that a lot of people think is the best option, right? The best method.

1157
01:51:00,041 --> 01:51:06,041
So, of course, that caused it. You know, the activation is kind of its own problem in a way,

1158
01:51:06,041 --> 01:51:12,601
though, because it's like, I mean, it's problematic. Enormously, they disagree enormously on what to do,

1159
01:51:12,601 --> 01:51:18,202
you know, Francis has had something like should just release a different client,

1160
01:51:18,202 --> 01:51:24,521
get people to run that like Jeremy Rubin kind of did what, what Francis Brio sort of said, but

1161
01:51:24,521 --> 01:51:28,762
I don't want to like ascribe too much. But I like some people have different views on exactly what

1162
01:51:28,762 --> 01:51:33,802
you should do to activate. Yeah, maybe there should be more. I mean, in the past, there was

1163
01:51:33,802 --> 01:51:38,922
nothing like this, but there was like, there was not as much controversy over anything. And even

1164
01:51:38,922 --> 01:51:43,002
the SegWit time, I don't think that was a real controversy that was that people didn't actually,

1165
01:51:43,961 --> 01:51:53,002
miners didn't actually dislike SegWit. What was thought was the miners needed some way,

1166
01:51:53,002 --> 01:51:56,762
the miners thought they weren't getting something that they wanted from the process,

1167
01:51:57,401 --> 01:52:01,481
is how I interpret it is that they... I think they were also lobbied.

1168
01:52:04,202 --> 01:52:09,562
People had convinced them that they need this two megabyte hard fork block size increase.

1169
01:52:10,442 --> 01:52:14,122
So they thought, how do we get that? And they thought this comes from the developers,

1170
01:52:14,122 --> 01:52:18,841
and they were sort of right about that. Like, you know, the developers put out the software,

1171
01:52:19,401 --> 01:52:22,762
I know that it would not have immediately like meant there would have been pushback

1172
01:52:22,762 --> 01:52:25,802
against the developers. But from their point of view, they were thinking like,

1173
01:52:27,481 --> 01:52:31,242
the developers have a gun pointed at me, I'm going to point a gun at them. And there was not

1174
01:52:31,242 --> 01:52:34,281
the point is, the point I'm trying to make has nothing to do with that. It's just about

1175
01:52:35,161 --> 01:52:41,322
SegWit was withheld because they thought, like, I'll give you your thing that you want when you

1176
01:52:41,322 --> 01:52:48,442
give me my thing that I want. Yeah. Like SegWit should be coded differently or something. 0%.

1177
01:52:50,281 --> 01:52:54,281
Well, I mean, it got a bit political and silly, really. But my impression was that a lot of the

1178
01:52:54,281 --> 01:52:59,882
miners didn't really care one way or the other. But I think people misunderstood the game theory

1179
01:52:59,961 --> 01:53:02,841
and thought that miners were kingmakers. And like, whatever the miners,

1180
01:53:03,802 --> 01:53:09,081
they thought that activation and the whole blockchain logic was enforced by miners.

1181
01:53:09,081 --> 01:53:12,442
And therefore, if they could just, you know, fly to China and persuade enough miners,

1182
01:53:13,242 --> 01:53:15,562
they could get the big blocks they wanted, or a hard fork.

1183
01:53:16,442 --> 01:53:23,481
This is true for every miner activated soft fork. They are sort of kingmakers. That's not how I

1184
01:53:23,481 --> 01:53:26,601
would describe it. But you know what I mean? Like, they could just unilaterally activate any

1185
01:53:26,601 --> 01:53:31,161
miner activated soft fork, right? And then it'd be a cat and mouse game. If people hated it,

1186
01:53:31,161 --> 01:53:41,481
they'd have to do the... Well, I think the thing with that is that it's not, I mean, technically,

1187
01:53:43,401 --> 01:53:47,802
it's the economic full nodes that enforce the consensus rules. And the miners just,

1188
01:53:49,081 --> 01:53:52,442
you know, also verify the consensus rules. Because if they don't,

1189
01:53:52,442 --> 01:53:58,362
the block will get ignored by the network and that'll be expensive, right? So from that point of

1190
01:53:58,362 --> 01:54:03,322
view... From the classical soft fork of like reassigning an op-nop or whatever, for the

1191
01:54:03,322 --> 01:54:08,202
script interpreter fails sometimes. They can... Because this is another point I tried to make in

1192
01:54:08,202 --> 01:54:14,122
Lisbon that I think no one understood. But it was like, if you do two soft forks for op-nop

1193
01:54:14,122 --> 01:54:20,122
three or something, and they're different, then that's actually a hard fork. So the order in

1194
01:54:20,122 --> 01:54:25,081
which you do the soft forks matters. Well, I mean, I think the thing is you could,

1195
01:54:25,882 --> 01:54:31,641
hypothetically, I mean, that's what you mean by miner activated soft fork. And that was the

1196
01:54:31,641 --> 01:54:38,601
terminology that evolved because I think one of the Bitcoin forks proposed to do it and it didn't

1197
01:54:38,601 --> 01:54:44,281
happen in the end. But that's where the miner activated soft fork came from, that term. But I

1198
01:54:44,281 --> 01:54:51,641
think technically miners could enforce some auxiliary rule that regular full nodes are not

1199
01:54:51,641 --> 01:55:00,682
looking at. But the problem is that would be a weaker model, right? Because what we were talking

1200
01:55:00,682 --> 01:55:07,721
about earlier, where you ideally want the full nodes to be upgraded before the miners activate

1201
01:55:07,721 --> 01:55:13,882
a new consensus rule. So you'd be kind of watering down the game theory and security of Bitcoin

1202
01:55:14,281 --> 01:55:20,442
right? Would be key because on day one, it would be weak. And we discussed earlier in this

1203
01:55:20,442 --> 01:55:27,641
conversation, the idea that it's not a good idea. The miners enforce a rule. We would then probably

1204
01:55:27,641 --> 01:55:32,521
want them to at least tell us about it so that at least we know what it is. So where one would be,

1205
01:55:32,521 --> 01:55:36,841
it would be the dark rule enforced secretly. And then they explain the rule. And ideally,

1206
01:55:36,841 --> 01:55:40,202
they would explain the exact open source code or whatever that's enforcing the rule,

1207
01:55:40,202 --> 01:55:43,961
which would be the full explanation of the rule. And then other people would maybe,

1208
01:55:44,601 --> 01:55:49,401
they would upgrade their node. They'd say, well, miners are enforcing this rule, then I want to be

1209
01:55:49,401 --> 01:55:56,762
maybe compatible. And then over time, it would harden into a user activated soft fork type thing.

1210
01:55:56,762 --> 01:55:58,521
Yeah, I mean, I think that's, I mean,

1211
01:55:58,521 --> 01:56:01,481
Should soft fork try to do that?

1212
01:56:01,481 --> 01:56:02,442
I think that's essentially dangerous.

1213
01:56:02,442 --> 01:56:08,202
For CPPs, you generally just go to the miners and say, well, you activate it. And we'll see,

1214
01:56:08,521 --> 01:56:11,562
if people say no, but I want to know, I would like to know like kind of why,

1215
01:56:11,562 --> 01:56:14,922
because I kind of also think the answer is no, but I wonder about that.

1216
01:56:16,682 --> 01:56:24,362
I mean, I think the BIP66 is an example of why not, right? Which is, if only the miners are

1217
01:56:24,362 --> 01:56:31,161
verifying something and people don't even know about it, or they don't agree, or they're not

1218
01:56:31,161 --> 01:56:38,041
enforcing it, then, you know, now you're relying on the current set of miners and they can make

1219
01:56:38,041 --> 01:56:40,122
mistakes and do some spy mining and stuff, right?

1220
01:56:40,122 --> 01:56:47,081
Isn't that an example where the miners only hurt themselves? Anyone who did nothing was

1221
01:56:47,081 --> 01:56:53,481
unaffected. If you had a full node, a regular full node, you were unaffected. I think

1222
01:56:55,882 --> 01:56:58,121
if they activate and then they tell everyone, like Jeremy would,

1223
01:56:58,842 --> 01:57:03,241
it's kind of implausible that they would secretly completely adopt a rule. And if they did,

1224
01:57:03,241 --> 01:57:06,682
we would never know about it anyway, by definition. So it would like with this

1225
01:57:06,682 --> 01:57:11,802
scenario, if Jeremy Rubin puts out his thing for discussion first, technically,

1226
01:57:12,361 --> 01:57:17,641
then he has, he goes to the miners and the miners say, well, why the heck not? And he convinces

1227
01:57:17,641 --> 01:57:25,082
them to, the pools click over, pools click over, you know, maybe that's far-fetched, but

1228
01:57:25,641 --> 01:57:27,082
I mean, I think, you know,

1229
01:57:27,881 --> 01:57:35,161
95% BIP-9 regular, what I consider the old regular style of soft fork activation.

1230
01:57:35,722 --> 01:57:37,722
The question is, what is wrong with that?

1231
01:57:39,562 --> 01:57:47,241
Well, I think it's a dramatic departure from Bitcoin security model in a way, right? Because

1232
01:57:47,802 --> 01:57:54,761
the op-nops are there to be reassigned though. They're kind of there to just be taken.

1233
01:57:54,761 --> 01:57:57,641
Suppose you put a four-byte version number in each transaction.

1234
01:57:57,641 --> 01:58:03,881
I mean, it's kind of like the BSV logic, which I think is, you know, they try to turn things

1235
01:58:03,881 --> 01:58:07,082
around and say, oh, you should trust the miners. It doesn't matter that you can't verify blocks,

1236
01:58:07,082 --> 01:58:12,281
you've got to trust miners anyway, miners are in charge. And, you know, if you take that

1237
01:58:12,281 --> 01:58:14,121
viewpoint, I mean, it's the consistent...

1238
01:58:14,842 --> 01:58:21,722
It's very different than, you know, for everything else, for everyone who's using it on Monday,

1239
01:58:21,722 --> 01:58:27,001
it stays exactly the way it was. And it was not a trust the miners BSV thing in any sense, right?

1240
01:58:27,001 --> 01:58:28,602
Not in the remotest sense.

1241
01:58:28,602 --> 01:58:38,041
Well, I mean, I think any, I think you're either SPV or you're not, right? So if you are not aware

1242
01:58:38,041 --> 01:58:44,361
of or not verifying something that is relevant, you could actually be accepting invalid blocks

1243
01:58:44,361 --> 01:58:45,241
and not know it, right?

1244
01:58:45,962 --> 01:58:51,241
Well, every block though, every protocol is an SPV protocol or some other protocol,

1245
01:58:51,241 --> 01:58:57,401
right? Some other like super protocol. Like if I'm a full node of Bitcoin Core 25,

1246
01:58:57,401 --> 01:59:03,482
I'm SPV of an infinite list of other protocol. So it can't just be...

1247
01:59:03,962 --> 01:59:12,602
So, but I mean, that is a reason why you want to upgrade, like preferably before

1248
01:59:13,962 --> 01:59:19,562
soft fork is activated, because otherwise you could come out of the wrong side of a BIP-66

1249
01:59:19,562 --> 01:59:22,521
like situation, right? I mean, I just...

1250
01:59:22,521 --> 01:59:26,281
If it's related to time, then that's also fixable though, because that's, and in fact,

1251
01:59:26,281 --> 01:59:29,881
it sort of was with BIP-9 where it's like, you have two weeks, you can make it longer,

1252
01:59:30,202 --> 01:59:34,761
you could say miners are going to signal over three months, and then you're going to have

1253
01:59:34,761 --> 01:59:40,682
three months to upgrade to a CPV, and then we'll know what to do.

1254
01:59:43,575 --> 01:59:53,575
I mean, I'm just kind of saying, I'm saying like, isn't the current process we have doesn't really seem to work, because it does seem like it's very ambiguous what the process is.

1255
01:59:53,575 --> 01:59:59,575
There's going to be a lot of people who can secretly delay things for years or veto things. It doesn't seem like...

1256
01:59:59,575 --> 02:00:03,575
Could you explain for context what the BIP-66 situation was?

1257
02:00:04,575 --> 02:00:16,575
So miners were spy mining, or I think over half of them, which means they were just pulling...

1258
02:00:16,575 --> 02:00:26,575
They were connected to a mining pool and to other mining pools, and when they would see that the jobs indicated a new block,

1259
02:00:26,575 --> 02:00:31,575
they would just start building empty blocks on top of it without verifying it, without waiting to verify it.

1260
02:00:31,575 --> 02:00:43,575
And so I think what happened is, by bad luck, you needed a certain amount of upgraded miners with activation signaling, but there were some stragglers that hadn't upgraded.

1261
02:00:43,575 --> 02:00:53,575
And so I think they spy mined on a block that was actually invalid and mined by a straggler that was below the 95% threshold or whatever it was at that time.

1262
02:00:53,575 --> 02:01:01,575
And then they all piled on top of it for a while. I think it was multiple blocks long before people figured out what was going on.

1263
02:01:01,575 --> 02:01:11,575
And so if you were running old software, you wouldn't necessarily have been able to tell what was happening.

1264
02:01:11,575 --> 02:01:15,575
Now, there was no malicious behavior there. It was just an accident.

1265
02:01:15,575 --> 02:01:22,575
But you could also construct malicious versions of that that could steal from people who hadn't upgraded, probably.

1266
02:01:23,575 --> 02:01:35,575
Yeah, I mean, the overall idea, I think it's not good for, you know, there's a lot of people's money invested in Bitcoin.

1267
02:01:35,575 --> 02:01:43,575
And so I think that would be a kind of dramatic variation of the security model and undermine confidence.

1268
02:01:44,575 --> 02:01:58,575
And therefore people would fight it. In reality, you'd end up with probably some kind of UASF-like variant that tried to block the miner-activated soft fork or something.

1269
02:01:58,575 --> 02:02:01,575
I don't know. It's not entirely predictable.

1270
02:02:01,575 --> 02:02:11,575
I think it's definitely the case that most people, you know, if they're faced with like a change, they probably say, well, you know, most people are actually concerned.

1271
02:02:11,575 --> 02:02:14,575
The devil you know versus the devil you don't or something.

1272
02:02:14,575 --> 02:02:20,575
So but I think it's also the case that I think it could be people who would worry the reverse.

1273
02:02:20,575 --> 02:02:26,575
And they would say, like, listen, like there's not a lot of competitive pressure right now.

1274
02:02:26,575 --> 02:02:33,575
But what if there was a different coin that was like moving faster or something?

1275
02:02:33,575 --> 02:02:47,575
You know, I'm saying it's not automatically the best thing that it's not automatically the best thing to be conservative because there are plenty of ways that the project could actually die off.

1276
02:02:47,575 --> 02:02:55,575
And I think people don't really know the details of like if for something to be a soft fork, like for your old node to still work to some extent.

1277
02:02:55,575 --> 02:03:00,575
It really is very limited in what it can do in that sense.

1278
02:03:00,575 --> 02:03:04,575
It depends on, you know, you can write weird soft fork, of course.

1279
02:03:04,575 --> 02:03:14,575
So, yeah, I don't know. But I'm just saying it's like it's not actually like a fully put together question to just say everything should always take 10 years.

1280
02:03:15,575 --> 02:03:22,575
Well, I mean, that that is an argument for extensibility and modularity, right?

1281
02:03:22,575 --> 02:03:25,575
So that more things can give up like lightning.

1282
02:03:25,575 --> 02:03:38,575
I think we can just go and do them in a modular way without needing a consensus change because, you know, I mean, you can think of a consensus change is a bit like IPv6, right?

1283
02:03:39,575 --> 02:03:47,575
It took years to get enough cross-compatibility and bridges that you can even rely on it, right?

1284
02:03:47,575 --> 02:03:54,575
So, of course, with Bitcoin, you need a kind of a defined activation point where the new rules apply.

1285
02:03:55,575 --> 02:04:11,575
So, I think, you know, I think consensus as a decision making process is inherently a bit slow because of people action is no action.

1286
02:04:11,575 --> 02:04:18,575
And that, you know, I mean, a consensus decision making process, there are organizations that use that for like hundreds of years, right?

1287
02:04:18,575 --> 02:04:20,575
So it's not a new phenomena.

1288
02:04:20,575 --> 02:04:30,575
And if you read about those things, it does seem like that's actually what happened, you know, if there would be something contentious that would just not, you know, take a long time, basically.

1289
02:04:30,575 --> 02:04:39,575
But the fact that it does take a long time is a reason why you want more extensibility and modularity.

1290
02:04:39,575 --> 02:04:44,575
And extensibility, I mean, like something like simplicity where you can implement more things directly.

1291
02:04:44,575 --> 02:04:49,575
And, of course, covenants enable more things to be implemented directly.

1292
02:04:49,575 --> 02:05:00,575
So as the expressiveness gets improved, you know, more things become possible to implement directly without needing new opcodes.

1293
02:05:01,575 --> 02:05:02,575
So I think it's converging on there.

1294
02:05:02,575 --> 02:05:10,575
And I hope that people, you know, because there are people that would say, you know, ossification is fine.

1295
02:05:10,575 --> 02:05:19,575
But, you know, there are other people that want and would benefit and value a feature that needs a new opcode, right?

1296
02:05:20,575 --> 02:05:23,575
I want to frame it as like, shouldn't we?

1297
02:05:23,575 --> 02:05:27,575
Because I think the framing, some people think, okay, like 50 people want this thing.

1298
02:05:27,575 --> 02:05:31,575
Like millions of people use Bitcoin, but 50 people want TPV or something.

1299
02:05:31,575 --> 02:05:33,575
How do I frame it like that?

1300
02:05:33,575 --> 02:05:45,575
I think most people, like 99% of the people, even like most people listening in or something, they're not going to know about, like, it's, let me try to put it this way.

1301
02:05:45,575 --> 02:05:48,575
Like we split the universe into two timelines or something.

1302
02:05:48,575 --> 02:05:55,575
The timeline with CTV or with, like there's big network effects.

1303
02:05:55,575 --> 02:06:00,575
So I'm saying in the future, everyone is onboarded with a drivechain or CTV or something else.

1304
02:06:00,575 --> 02:06:03,575
And that's the world where Bitcoin wins.

1305
02:06:03,575 --> 02:06:11,575
And it's not that Bitcoin loses in the other world because there are so many fans of drivechain like today or something.

1306
02:06:12,575 --> 02:06:24,575
It's like most people will never have heard of any of this, you know, but they, whether or not they can actually be a real user of Bitcoin kind of depends on whatever are colliding.

1307
02:06:25,575 --> 02:06:33,575
So that's what I'm trying to say is it's actually, it's not like, oh, we need to, what are we going to do if we lose?

1308
02:06:33,575 --> 02:06:38,575
If we lose Jeremy Rubin, then the project will be dead because that's not what I'm saying.

1309
02:06:38,575 --> 02:06:46,575
I'm saying like we have to, the project, the project competes is open source software.

1310
02:06:46,575 --> 02:06:54,575
And there is a limit to how many good ideas you can ignore before it just makes the project uncompetitive.

1311
02:06:54,575 --> 02:06:58,575
We're nowhere near that point now, but we will inevitably reach that point.

1312
02:06:59,575 --> 02:07:21,575
Well, I mean, I think the motive is that I think probably is meaningful to more people like in a positive way is that, you know, we all think that the properties of Bitcoin are really cool.

1313
02:07:21,575 --> 02:07:23,575
And over time, there are more people that want to use it.

1314
02:07:23,575 --> 02:07:27,575
And of course, lightning helps, but it's not enough, right?

1315
02:07:27,575 --> 02:07:31,575
And, you know, UTXO sharing schemes.

1316
02:07:31,575 --> 02:07:35,575
Okay, that's, that's, you know, maybe that works.

1317
02:07:35,575 --> 02:07:42,575
But ultimately, there is something higher assurance coming from a UTXO, right?

1318
02:07:43,575 --> 02:08:04,575
And so if, if the technology doesn't evolve to be able to provide more people with UTXOs, then, you know, as a Bitcoin active user base that wants to cold store savings and censorship resistant transactions, you know, goes from 100 million to a billion.

1319
02:08:04,575 --> 02:08:05,575
We have a problem, right?

1320
02:08:05,575 --> 02:08:06,575
With today's technology.

1321
02:08:06,575 --> 02:08:18,575
And so, you know, what, what will break if that if, if the technology, you know, if the adoption gets out of technology as well, probably more people will end up using custody or something kind of that will get degraded service, right?

1322
02:08:18,575 --> 02:08:33,575
So, to our point of view, as technologists that look at these trade offs, they're like, well, sure, we'd be better if they could get something in between, which is, you know, they can get a real UTXO, but it's on a drivechain or a sidechain that is, you know, got a different security trade off.

1323
02:08:33,575 --> 02:08:34,575
That would be better, right?

1324
02:08:35,575 --> 02:08:44,575
And so I think it's a good point that we want to find technical solutions for that, because it looks like the adoption is going to be coming.

1325
02:08:44,575 --> 02:08:51,575
And, you know, it can, it can happen relatively quickly too, right?

1326
02:08:52,575 --> 02:09:11,575
Yes, and whatever the end user demand will be the engine that pulls the train, so to speak, where it's like, whatever they want will be the thing that gets all the network effects, like if they want custodial, or if they want, you know, like, it'll just be.

1327
02:09:12,575 --> 02:09:22,575
So what I'm saying is like, on day one, you turn Jeremy Rubin away, or you just make it take five years instead of taking six months.

1328
02:09:22,575 --> 02:09:32,575
That's on day one, and then day later, you know, Brian Armstrong and Coinbase owns every Bitcoin or something, or something like a long timeline.

1329
02:09:32,575 --> 02:09:43,575
Whatever the end user is going to want will get the network effects, and it'll be self-reinforcing. Maybe it'll be a different, it could be a different coin other than Bitcoin, or it could be something worse.

1330
02:09:43,575 --> 02:09:44,575
CBTC, I don't know.

1331
02:09:44,575 --> 02:09:45,575
Well, I mean, I think the…

1332
02:09:45,575 --> 02:09:48,575
Whatever the user wants is going to be their thing.

1333
02:09:48,575 --> 02:09:52,575
Yeah, I mean, you can see that.

1334
02:09:52,575 --> 02:10:13,575
I mean, actually, with the fee spike, which seems to be caused by the inscripters, you know, putting JPEGs in the chain, which is a craze like a month or two ago, caused the fees to go up and created service problems for some lightning services.

1335
02:10:14,575 --> 02:10:32,575
And so, you know, what they ended up doing was, this is Bolts Exchange, so they have a submarine swap to add money to a lightning channel or take money out of a lightning channel, but it involves on-chain transactions, right?

1336
02:10:32,575 --> 02:10:46,575
And so the fees got high and it stopped working. So they made a trustless atomic swap, like a submarine swap, from liquid Bitcoin to lightning so that they wouldn't need to do an on-chain transaction.

1337
02:10:46,575 --> 02:10:55,575
And they did it really quickly. They did it in the space of a couple of weeks, and they had an internal alpha, and people were playing with it and stuff, and it was live.

1338
02:10:55,575 --> 02:11:10,575
But not all things are easy to do quickly, right? You know, like a layer two, like ARK or Fedimint, things like that, take quite a while to get enough adoption and wallet technology and stuff.

1339
02:11:11,835 --> 02:11:17,775
I think I'm going throw off a vacation to some extent,

1340
02:11:18,355 --> 02:11:22,235
but I just also think like, well, what should we do now is kind of where I'm kind of

1341
02:11:22,235 --> 02:11:25,315
boiling it down to, because this idea is not this certainly wouldn't say this is

1342
02:11:25,315 --> 02:11:29,355
necessarily quickly because this is from November 2015, the original

1343
02:11:40,575 --> 02:11:45,575
Jeremy Rubin or like, you know, like, what should I? That's kind of what I'm getting around.

1344
02:11:45,575 --> 02:12:09,575
I think, you know, with new technology, you want to see, like, multiple people engaged in, like, analysis and feedback and variance and stuff. Like, there's, I think, somewhere along the way, there have been spurts of, you know, constructive discussion about variance, right?

1345
02:12:09,575 --> 02:12:27,575
Other than Checksick from SAC, which was older and in liquid, there's TxHashed and a few variants, and I think one more recent variant. So, you know, whatever it was that triggered those conversations is good.

1346
02:12:27,575 --> 02:12:47,575
And, you know, if you get some people who are, you know, get interested in things and see them through to the end, like, you know, got the energy to write bits and be actively involved and involved for months at a time, and get fewer of them involved.

1347
02:12:47,575 --> 02:13:12,575
And I think that's usually what it takes to get soft forks, you know, into a stage where people will want to explore consensus and get, you know, get some kind of wider interest going. So now you can cultivate that, but that's what observationally, that's what seems to have worked in the past, right?

1348
02:13:13,575 --> 02:13:31,575
I'm thinking, like, for example, with Taproot and short signatures, there are a lot of people looking at those bits. I mean, you can look at, like, the GitHub and I guess the comments on the bits and the code and discussion forums.

1349
02:13:31,575 --> 02:13:53,575
And maybe conferences too, like there are some sort of, not really a proper conference, but like a hackathon or whatever the IETF calls those kind of things. I think those are constructive.

1350
02:13:53,575 --> 02:14:22,575
Yeah, and it's slowing down very meaningfully.

1351
02:14:23,575 --> 02:14:48,575
Of course, if we've reached a point where the size of the network or whatever, whatever it is that has grown past the scale of us to be able to really do anything, like say we've reached the point where ossification has kind of organically happened, it's not like we're gonna, not like there's a sign that goes up and everybody is made aware that that point has been reached.

1352
02:14:48,575 --> 02:15:07,575
It might be many years before it becomes obvious and there's kind of just a consensus around the fact that that point has been reached. And you mentioned, obviously, the downsides of that.

1353
02:15:08,575 --> 02:15:27,575
So in that case, there might not be value in trying to approach it from different ways. They were trying to push changes, soft forks, in other ways. It's not like that.

1354
02:15:27,575 --> 02:15:50,575
You want it to succeed, right? For Bitcoin to ossify in the long term, I think is probably good for the base layer to not be undergoing rapid changes as there's more value on top. But for it to get there, for that to be sensible, I think we need more changes, right?

1355
02:15:50,575 --> 02:15:59,575
We need extensibility, we need simplicity, we need sidechains, drivechains and covenants and maybe some more things, right?

1356
02:15:59,575 --> 02:16:25,575
But I think the other interesting thing is that maybe there are, I think my feeling is the natural evolutionary pressures are there for Bitcoin to want to reach that sensible ossification stage where the ossification isn't preventing generic innovation in a permissionless way, right?

1357
02:16:26,575 --> 02:16:42,575
But I think the other thing is there could be multiple paths to get there because it doesn't take that much more expressiveness on a chain because I said the limits on expressiveness of Bitcoin scripts are kind of partly accidental due to specific opcodes getting disabled.

1358
02:16:42,575 --> 02:16:55,575
Like cat was a quite powerful one and it had bugged coincidentally, right? But as a few more opcodes come in that are more generic or powerful like covenants, I think that might be enough to implement something, right?

1359
02:16:55,575 --> 02:17:15,575
Or the other thing is it seems like people are more inclined to accept, to go with extensibility because extensibility is a kind of positive thing, right?

1360
02:17:15,575 --> 02:17:42,575
Well, we can do more things with it. Great. Let's do it kind of thing, right? Whereas if it's about a specific use case, I'm like, well, I don't know, does that use case have downsides? And then they want to analyze that, right? But whereas if the extensibility, you can use that to build a drivechain, well, then you get to drivechains in a different way. And then maybe you get a drivechain, you know, you build a drivechain using that, using covenants, for example.

1361
02:17:42,575 --> 02:17:43,575
Yeah, that would be fine.

1362
02:17:43,575 --> 02:17:58,575
A more efficient way comes later, right? You know, maybe because it's kind of like that C prepressor, for example, right? The fact that it was possible to do it in a kind of hacky way seemed to make people completely okay with, oh, well, let's make it elegant then as it's already possible, right?

1363
02:18:00,295 --> 02:18:02,735
Well, that's what I hope though, because it's already...

1364
02:18:03,355 --> 02:18:04,135
Well, it's already...

1365
02:18:04,135 --> 02:18:05,975
Almost. I guess.

1366
02:18:06,655 --> 02:18:09,815
I mean, what you're saying, we could make like, we could make one that was only with

1367
02:18:09,815 --> 02:18:11,295
like, nLockTime or something.

1368
02:18:11,895 --> 02:18:12,655
It's with a...

1369
02:18:12,655 --> 02:18:13,908
I could like,

1370
02:18:13,908 --> 02:18:19,568
like make a make a mining consortium wallet and then collect

1371
02:18:19,568 --> 02:18:20,428
everyone's money.

1372
02:18:21,928 --> 02:18:26,928
I don't know it doesn't seem to be it seems to going to slow like not in a sense of

1373
02:18:26,928 --> 02:18:28,242
flow is good slow

1374
02:18:28,322 --> 02:18:31,602
slow and steady, seems to be maybe not

1375
02:18:31,602 --> 02:18:36,862
slow in the sense that the process is becoming

1376
02:18:36,862 --> 02:18:42,575
more defective than it was, more bureaucratic, maybe, which is sort of fine.

1377
02:18:43,575 --> 02:19:10,575
I mean, I think the progress on covenants is kind of not linear, right? Like, stop for a bit, and then there were like, people got excited, and there are multiple proposals, extra, like new proposals, or, you know, some people deciding actually maybe CTV was better than they thought, or trying to analyze the, you know, the equivalence.

1378
02:19:10,575 --> 02:19:29,575
Like, you know, could you implement CTV with, what is it, APO? That's the one for lightning. And apparently, I think you can do it in one direction, but not the other, which shows that one is slightly more powerful than the other or something.

1379
02:19:29,575 --> 02:19:41,575
So those conversations were happening, you know, more recently, which I think is useful. So, you know, I think, to get progress on sidechains and drivechains, peer-to-peer version,

1380
02:19:41,863 --> 02:19:46,503
we need to cultivate some compositions like that.

1381
02:19:46,503 --> 02:19:49,703
So even if people are, like, proposing alternatives,

1382
02:19:49,703 --> 02:19:52,903
like, at least they're talking about it, right? And it's a step forward.

1383
02:19:52,903 --> 02:19:58,343
Well, my concern is kind of like that, to really get the pressure,

1384
02:19:58,343 --> 02:20:04,502
you actually would need some kind of crisis or some kind of competitor or something.

1385
02:20:04,502 --> 02:20:08,343
I don't know if that's the case or not, but that's kind of like how I see it, is like, people...

1386
02:20:08,502 --> 02:20:13,143
You know, it's like the exam paper is due Thursday, and so you procrastinate until Wednesday.

1387
02:20:14,183 --> 02:20:20,183
If it was due earlier, they would do it earlier. I don't know, maybe that's not correct.

1388
02:20:22,023 --> 02:20:26,423
It's kind of like bounded by other things that we've done, like taproom.

1389
02:20:26,423 --> 02:20:33,223
But I think that, yeah, I mean, I think the problem is that, you know, if push comes to shove,

1390
02:20:34,183 --> 02:20:40,263
if you ask people who are, you know, have 90% of their savings in Bitcoin,

1391
02:20:41,383 --> 02:20:44,822
what features they want, get an interesting response, which is,

1392
02:20:45,543 --> 02:20:49,303
well, my number one feature request is just don't break it.

1393
02:20:50,903 --> 02:20:53,702
And then, you know, so if there's any risk...

1394
02:20:54,903 --> 02:21:02,023
Like that joke is seen as saying every soft fork has, like, an X% chance of breaking it or

1395
02:21:02,023 --> 02:21:03,623
whatever, which is really not true because...

1396
02:21:03,623 --> 02:21:08,502
Well, right. I mean, I think, yeah, yeah. So I'm not saying that, but I'm just saying that if...

1397
02:21:08,502 --> 02:21:11,223
Don't replace the software with, like, something else.

1398
02:21:11,223 --> 02:21:19,223
But like, I'm just saying that for the people who are interested in Bitcoin for different reasons,

1399
02:21:19,223 --> 02:21:25,543
and if they're just, you know, using it as an investment, and they might, you know,

1400
02:21:25,543 --> 02:21:30,583
apart from the custody, they might not feel bad about it being the ETF where they don't even have

1401
02:21:30,583 --> 02:21:33,863
to touch it. Like, instead of having a gold bar, they've got a gold ETF and they don't care,

1402
02:21:33,863 --> 02:21:39,543
right? So for those kind of people, you know, if you complain to them, like,

1403
02:21:39,543 --> 02:21:42,903
oh, Bitcoin is ossifying, it won't be able to have this feature, it won't innovate as

1404
02:21:42,903 --> 02:21:47,863
permissionlessly, they kind of shrug and say they don't care, right? But for the rest of us,

1405
02:21:47,863 --> 02:21:51,943
we think this is really bad because now, you know, what's the side effect going to be?

1406
02:21:51,943 --> 02:21:54,023
About all Bitcoin users.

1407
02:21:54,023 --> 02:21:54,823
Right, right.

1408
02:21:54,823 --> 02:21:59,223
I mean, the network effects mean whoever cares about the most users is just going to win,

1409
02:21:59,223 --> 02:22:03,622
and every other coin will eventually go to zero, is my view.

1410
02:22:03,622 --> 02:22:08,342
Right, well, I mean, I suspect that what's going to happen is Bitcoin is going to win regardless,

1411
02:22:08,342 --> 02:22:15,463
it's just that it will have the unfortunate effect of, you know, if people can't own UTXOs

1412
02:22:15,463 --> 02:22:20,502
and censorship and transact without censorship resistance, then it will be a degraded future

1413
02:22:20,502 --> 02:22:25,863
where, you know, a lot of users won't have as good of assurances they're getting from Bitcoin.

1414
02:22:26,663 --> 02:22:28,183
Well, they won't, you know, if...

1415
02:22:28,183 --> 02:22:31,943
Bitcoin would clearly beat everything that exists, like everything, I don't think any

1416
02:22:31,943 --> 02:22:37,463
of the competitors are that good, including Ethereum, including whatever, take your pick.

1417
02:22:39,463 --> 02:22:47,542
So, but I don't, I think this is, like, wrong to generalize from the current set are very weak,

1418
02:22:47,542 --> 02:22:53,223
I think. I don't think there has been strong competition, like people don't really,

1419
02:22:53,223 --> 02:22:56,183
they still were still, like, very early as far as, like, the...

1420
02:22:57,863 --> 02:23:02,342
So you're saying that maybe a competent altcoin might emerge eventually?

1421
02:23:02,342 --> 02:23:05,782
Right, at some point, yeah, and there have been, there's been very few that are that

1422
02:23:05,782 --> 02:23:10,583
competent, and if one emerged that was, like, half competent and more popular than Bitcoin,

1423
02:23:11,943 --> 02:23:16,743
then I would think, like, you know, then that would send the Bitcoin to zero eventually.

1424
02:23:17,062 --> 02:23:25,622
I mean, I think that people, you know, because, I mean, if that's true or not,

1425
02:23:25,622 --> 02:23:29,062
I don't think people are going to find it persuasive, because they are pretty much

1426
02:23:29,062 --> 02:23:35,863
convinced that Bitcoin will succeed in one way, even if just as a savings technology,

1427
02:23:35,863 --> 02:23:41,542
that's not much transacted, or not even, you know, that only 10% of people own UTXOs,

1428
02:23:41,542 --> 02:23:46,502
or only 1% of people own UTXOs, and the rest are using custodians or ETFs.

1429
02:23:46,502 --> 02:23:51,863
So, like, for some people, they would be okay with that outcome. Now, we wouldn't,

1430
02:23:51,863 --> 02:23:57,383
because that is not self-sovereign, you know, not your keys, not your coins, etc. So we

1431
02:23:57,383 --> 02:24:04,982
desperately want to find a way to scale it without degrading security or having opt-in security

1432
02:24:05,942 --> 02:24:12,583
layers or sidechains and things like that, right? So I think we need it to happen,

1433
02:24:12,583 --> 02:24:20,743
because otherwise the future is less good. But I think that it can happen incrementally in,

1434
02:24:21,542 --> 02:24:29,542
like, three or four different tracks, right? So one way is, if some of these covenant opcodes get

1435
02:24:29,542 --> 02:24:34,982
implemented, that might be enough new expressive power to implement a drag chain, you know,

1436
02:24:34,982 --> 02:24:38,503
maybe with some big, annoying scripts or something, but still, right?

1437
02:24:38,503 --> 02:24:41,783
I have a slightly different point of view, which is that the ETF people

1438
02:24:42,903 --> 02:24:48,662
are shooting themselves in the head, so to speak, because what they're going to do,

1439
02:24:48,662 --> 02:24:54,822
they're saying Bitcoin only needs to be good as a savings account, which is fine. It should be good

1440
02:24:54,822 --> 02:25:05,463
as a savings account and also other things. But because of this view, the key innovation in terms

1441
02:25:05,463 --> 02:25:14,102
of network effect, just build it all on network effect, will be slower to arrive in Bitcoin than

1442
02:25:14,102 --> 02:25:21,142
in the competent altcoin that doesn't exist yet. And then as a result, they will not even be

1443
02:25:21,942 --> 02:25:27,942
they will not even be owners when the competent altcoin becomes more popular. And then they are

1444
02:25:27,942 --> 02:25:32,743
actually... So now it didn't even work as a savings account. So even by their own criterion, they...

1445
02:25:35,942 --> 02:25:44,503
Yeah, I mean, I think it is a short sighted view in a sense that, you know, it's kind of like

1446
02:25:45,463 --> 02:25:48,982
people buying internet stocks, and they're not interested in the internet, right? So they're

1447
02:25:48,982 --> 02:25:53,463
like, well, the internet might, there seems to be people excited about the internet, I don't care,

1448
02:25:53,463 --> 02:25:58,183
but I'm going to buy some stocks because I think they'll put the price up. But the point is that,

1449
02:25:59,142 --> 02:26:03,942
you know, what's driving the price, right? It's adoption and why people are adopting because of

1450
02:26:03,942 --> 02:26:10,822
the censorship resistance and the bearer assurances and things. So, you know, the people

1451
02:26:11,382 --> 02:26:18,422
just investing in the asset class, I mean, it's, you know, they just, I mean, that's a clear,

1452
02:26:18,422 --> 02:26:24,822
strong part of Bitcoin's adoption driver. So we shouldn't knock that. But the value they're

1453
02:26:24,822 --> 02:26:30,023
investing in is because of the technology too, right? So you need the technology to

1454
02:26:30,982 --> 02:26:37,382
to keep working. And as the scale increases, and support different trade-offs,

1455
02:26:38,982 --> 02:26:42,743
or it's less valuable, right? It doesn't reach its potential, it's less valuable,

1456
02:26:42,743 --> 02:26:48,023
I think. Because it's not, it's not only that they want, I agree with that, people do want

1457
02:26:48,023 --> 02:26:52,503
the censorship resistant payments, they want the privacy, they want the self-sovereignty.

1458
02:26:52,503 --> 02:26:58,343
But they also, they really, really need to be on the network that ends up being the one that

1459
02:26:58,662 --> 02:27:04,102
everyone uses. Because I really think it's like, sort of like, speaking English versus Esperanto

1460
02:27:04,102 --> 02:27:10,023
or something, you know, it's like, they just don't realize that they want that, you know,

1461
02:27:10,023 --> 02:27:13,542
they bought, they're buying the internet stock, they're not interested in the internet, they don't

1462
02:27:13,542 --> 02:27:20,743
realize their counter-internet will just... But I mean, I think their assumption might be,

1463
02:27:20,743 --> 02:27:26,183
and it could be right, you know, I think it's probably a fairly fringe view, right? I mean,

1464
02:27:26,903 --> 02:27:33,142
I think most people who are interested in Bitcoin today care about the technology working.

1465
02:27:34,102 --> 02:27:39,222
I mean, maybe some macro investors don't, and it's like, you know, seeing something that's popular

1466
02:27:39,222 --> 02:27:43,702
and people are buying, but they don't need it, right, from their point of view. But I think

1467
02:27:43,702 --> 02:27:50,422
that's like, at least by usage, that's a small percentage. But, you know, they could be right too,

1468
02:27:50,662 --> 02:27:59,623
which is, you know, it's like TCPIP. Could something displace TCPIP? Yeah, maybe not,

1469
02:27:59,623 --> 02:28:08,903
you know, not enough incentive. Nobody cares to change. So, but that's not a reason for us. I

1470
02:28:08,903 --> 02:28:13,303
mean, we want the technology to reach its full potential and deliver value to anybody who wants

1471
02:28:13,303 --> 02:28:20,823
to own a UTXO and have transactions to be able to have them, right? Yeah, and of course, they

1472
02:28:20,823 --> 02:28:25,943
might think, which would be very clever of them, they may think, well, listen, the community is

1473
02:28:25,943 --> 02:28:32,022
important, so they will bring my UTXO along for the ride. The new thing will be whatever,

1474
02:28:32,902 --> 02:28:36,902
hard fork in Bitcoin or have they import the UTXO set or something like that. So they think,

1475
02:28:36,902 --> 02:28:40,823
I'm just going to own this and I'll probably be good, which would probably be a pretty smart

1476
02:28:40,823 --> 02:28:45,383
thing. I would actually say that's probably pretty smart. If a person doesn't have time

1477
02:28:45,383 --> 02:28:50,743
to look into all this minutia. I mean, people do that, right? I mean, that was the fork phenomena

1478
02:28:50,743 --> 02:28:57,383
and the airdrop phenomena, right? Trying to, because I guess a new coin, when there's 25,000

1479
02:28:57,383 --> 02:29:02,503
coins or however many there are today, primary problem is lost in the sea of noise, right?

1480
02:29:03,063 --> 02:29:06,582
Got to find a reason for anybody to care about it. So they need like a massive pre-mine,

1481
02:29:06,582 --> 02:29:14,582
a massive marketing budget, and to get anybody to care about it. So hence the airdrop trick, right?

1482
02:29:17,303 --> 02:29:25,222
Yeah, but I think that's difficult at this point. Anyway, I mean, I think that

1483
02:29:28,823 --> 02:29:31,462
multiple of these strands can get there too, right? So

1484
02:29:32,422 --> 02:29:37,863
covenants, maybe that gets there. I think it'd be interesting to analyze

1485
02:29:39,703 --> 02:29:43,943
if all of those covenant variants would enable drivechains, or if one is better,

1486
02:29:43,943 --> 02:29:53,143
or if there are changes that would make it easier to do. Interesting to cross compare.

1487
02:29:53,143 --> 02:29:58,342
Yeah, again, I could design the JIT 300 to be very efficient and whatever, but I think your

1488
02:29:58,582 --> 02:30:04,102
point would be, you get the actual drivechain, then you do fake, I mean, you get like the kind

1489
02:30:04,102 --> 02:30:12,902
of simulated or, you know what I mean, like a backdoor kind of, and then you just tell people,

1490
02:30:12,902 --> 02:30:19,143
by the way, you should just merge this because it just cleans up something that people already do.

1491
02:30:20,263 --> 02:30:24,823
Well, I mean, you can see that with lightning, right? In the sense that

1492
02:30:25,782 --> 02:30:29,623
the current lightning is made more efficient and it's free management, I think,

1493
02:30:29,623 --> 02:30:34,982
with the APO or other solutions to the current limitation.

1494
02:30:34,982 --> 02:30:39,383
People say that, but actually then other people say that without the punishment transaction,

1495
02:30:40,183 --> 02:30:41,943
it won't work. So they say,

1496
02:30:44,823 --> 02:30:49,863
Well, I mean, I think it's not actually as adversarial as they like to think it is,

1497
02:30:49,863 --> 02:30:53,782
but it's a fair point, right? If you can grief on something all day with no penalty, then

1498
02:30:54,982 --> 02:30:58,422
and you don't have an ethical boundary to that, then why wouldn't you just keep doing that?

1499
02:30:59,222 --> 02:31:01,623
You still promise once in a while, right?

1500
02:31:01,623 --> 02:31:04,582
Yeah. So the heads I win, the tails I break even.

1501
02:31:06,582 --> 02:31:12,982
Right. Yeah. So anyway, but I mean, I think it's the sort of analogous situation, right,

1502
02:31:12,982 --> 02:31:18,743
which is something is working, but not as efficient as it could be. And so it seems like

1503
02:31:19,863 --> 02:31:23,462
Yeah, that's an easy sell because this is an improvement on

1504
02:31:24,183 --> 02:31:27,462
inefficient thing that already, it's just an efficiency improvement at that point.

1505
02:31:28,422 --> 02:31:33,222
Right. So I mean, if you can, and I think the efficiency maybe doesn't even matter that much

1506
02:31:33,222 --> 02:31:36,823
if there are not that many commitment, if there's only one commitment transaction per block, or

1507
02:31:37,782 --> 02:31:46,902
if it's infrequent or what have you. And, but it does, I think, because of the sort of limit of,

1508
02:31:48,342 --> 02:31:51,143
you know, because it's open source and people are working on what they find interesting.

1509
02:31:52,102 --> 02:31:58,422
One thing that drives interest is usage, right? So if people are using something and it's cool,

1510
02:31:58,422 --> 02:32:03,222
and there are cool things happening over there, and they'll get interested to read about it and

1511
02:32:03,222 --> 02:32:07,542
then get enthusiastic for making it more efficient, right? So I think that's one path.

1512
02:32:09,303 --> 02:32:15,222
Another possible path is simplicity, which is, you know, if, I mean, it's not quite ready yet,

1513
02:32:15,222 --> 02:32:19,462
but you know, if Bitcoin had simplicity, then you could probably make a pretty efficient

1514
02:32:20,183 --> 02:32:24,582
drivechain or just implement, you know, BIP300 using it or something, right? So,

1515
02:32:25,542 --> 02:32:33,782
or close. So that would be another path. And that is, you know, the pitch for why liquid or

1516
02:32:33,782 --> 02:32:40,183
Bitcoin might want simplicity is that it's a candidate for ossification, which is it's enough

1517
02:32:40,183 --> 02:32:44,102
self-extensibility that you could implement something like Schnorr from scratch without

1518
02:32:44,102 --> 02:32:49,383
needing a soft fork, other than for kind of competition efficiency reasons. So

1519
02:32:51,542 --> 02:32:55,623
about 300 itself, it's sort of like it's simplicity, but only on L2.

1520
02:32:57,383 --> 02:33:00,342
Yeah. Yeah. But I mean, I think that because

1521
02:33:00,342 --> 02:33:04,503
there's all these people complaining about a crazy, they're weird theories about what

1522
02:33:04,503 --> 02:33:08,503
BIP300 might do. Shouldn't that all just apply double for simplicity or something?

1523
02:33:08,902 --> 02:33:14,582
Well, I mean, if we went this route, I would be happy for it all to be over. But I would just

1524
02:33:14,582 --> 02:33:20,422
think, I mean, it was simplicity changes, miner centralization or whatever.

1525
02:33:20,422 --> 02:33:24,582
Well, we haven't, we haven't had that debate yet. So, you know, we'll see how it works out,

1526
02:33:24,582 --> 02:33:33,143
but right. But I think it might be different because it's super low level. So people are

1527
02:33:33,143 --> 02:33:38,422
going to look at it and compare it to. I mean, there's a mistaken comparison

1528
02:33:38,422 --> 02:33:42,982
there because it would enable BIP300. So it's actually in a weird way, it's higher level

1529
02:33:42,982 --> 02:33:47,063
because it's. Right. But I think people don't have to be

1530
02:33:47,063 --> 02:33:52,102
logical in their inferences. They look at the increment, right? So it's kind of like this,

1531
02:33:52,102 --> 02:33:56,662
the C preprocessor example I gave. So, you know, when there was a question of adding

1532
02:33:57,383 --> 02:34:01,863
concatenate operator, big debate, when somebody shows it's already possible,

1533
02:34:01,863 --> 02:34:05,303
okay, great. Let's optimize it. Right. So it just completely changed the discussion.

1534
02:34:06,022 --> 02:34:12,503
So I remember you saying a couple of months ago that simplicity is at minimum five years away.

1535
02:34:12,503 --> 02:34:18,823
Can you explain then what that timeline, what is happening in those five years? Why is it five

1536
02:34:18,823 --> 02:34:25,863
years away? Well, I mean, it's not concrete. I'm just looking at, you know, the Schnorr signatures,

1537
02:34:25,863 --> 02:34:31,303
like how long things have gone from idea phase to adoption, but, you know, like there's a team

1538
02:34:31,303 --> 02:34:35,063
of people working on simplicity. So the technology is evolving pretty fast and we're hoping to get

1539
02:34:35,063 --> 02:34:44,422
it into liquid relatively soon, like this year or something in a life form. But, you know,

1540
02:34:45,623 --> 02:34:50,902
you have, it's the same challenges, covenants or drivechains, right? So now you have it

1541
02:34:51,542 --> 02:34:57,782
and then people have to, and simplicity is doing, you know, novel things that

1542
02:34:59,063 --> 02:35:04,823
in a sort of different program model. So people don't like to accept things they don't understand.

1543
02:35:04,823 --> 02:35:10,102
So probably take, you know, enough developers to get up to speed and understand how it works to

1544
02:35:11,222 --> 02:35:15,703
say that, okay, now I get it. Seems good to me kind of thing, right? So you need that kind of

1545
02:35:15,703 --> 02:35:20,503
viewpoint where I think a drivechain is easier to understand probably. So that's possibly a

1546
02:35:21,143 --> 02:35:31,943
downside, but, oh, and another direction which could, you know, change the game is more and more

1547
02:35:32,823 --> 02:35:39,143
compact and efficiently verifiable things of execution, like snarks and stocks and bullet

1548
02:35:39,143 --> 02:35:47,863
proofs and things. So the research guys at Blockstream recently published a paper called

1549
02:35:48,823 --> 02:35:54,982
Bulletproofs++, which has, you know, got some more interesting efficiency optimizations in it.

1550
02:35:54,982 --> 02:36:05,782
So, and an indication that you might be able to efficiently implement unlinkability using it. So

1551
02:36:05,782 --> 02:36:14,503
that's, it shows that that area is progressing. Of course, there are multiple kind of

1552
02:36:14,503 --> 02:36:20,263
zero-knowledge proof frameworks that, you know, implement whole circuits, but many of them are

1553
02:36:20,263 --> 02:36:26,263
using novel security assumptions. And so I think it'd be difficult for Bitcoin to adopt them,

1554
02:36:26,263 --> 02:36:32,823
but Bulletproofs is using similar kind of assumptions to Schnorr and ECUSA. So I think

1555
02:36:32,823 --> 02:36:41,462
it's the kind of security assumption that is more likely to be adoptable by Bitcoin, let's say.

1556
02:36:42,342 --> 02:36:50,263
So, you know, that path is, I don't know, I couldn't place a timeline on that because

1557
02:36:50,982 --> 02:36:58,023
I think those proof systems are not efficient enough yet. And there's also a kind of

1558
02:36:59,863 --> 02:37:05,942
question mark around them. But like, in theory, if they got efficient enough to do what Greg

1559
02:37:05,942 --> 02:37:12,023
Maxwell talked about on Bitcoin Talk in, I don't know, 2012, 2013, whenever it was,

1560
02:37:13,942 --> 02:37:18,342
about implementing a sidechain using, I mean, basically it was a sidechain, but before that

1561
02:37:18,342 --> 02:37:23,222
term was coined, but where the sidechain logic is enforced by the main chain using a

1562
02:37:23,222 --> 02:37:31,543
zero-knowledge proof or proof of execution, you could get scalability, you know, so it might be

1563
02:37:31,543 --> 02:37:38,982
expensive for miners, but we might have a different problem.

1564
02:37:39,702 --> 02:37:44,423
I think, right, because then don't you have to bike-shed like every detail of like the prover

1565
02:37:44,423 --> 02:37:52,423
and whatever? Like I haven't looked into it that much. And of course, every day that passes in

1566
02:37:52,423 --> 02:37:59,783
that timeline, we get almost certainly closer and closer to the point of authentication.

1567
02:38:00,263 --> 02:38:06,503
Yeah, I mean, I think that the zero-knowledge proof stuff, while it's evolving, is probably

1568
02:38:06,503 --> 02:38:14,663
further away than, I think probably the covenants path might be the quickest and then

1569
02:38:17,462 --> 02:38:18,982
drag chains or simplicity

1570
02:38:22,503 --> 02:38:26,982
in terms of potential path modularity.

1571
02:38:27,222 --> 02:38:28,503
Modularity.

1572
02:38:28,503 --> 02:38:33,302
Well, should we invite other people to speak maybe? We seem to have a lot of people here.

1573
02:38:33,302 --> 02:38:38,743
Sure. I have two quick questions of my own. Adam, thank you so much for being here.

1574
02:38:38,743 --> 02:38:43,143
You mentioned that drivechains are more peer-to-peer than the ordinary sidechains

1575
02:38:43,143 --> 02:38:48,423
that exist today. In some tweets, you used that phrase more peer-to-peer and you said that today.

1576
02:38:48,423 --> 02:38:54,822
Can you say in your words, how is a drivechain sidechain more peer-to-peer than the sidechains

1577
02:38:54,822 --> 02:38:55,783
Bitcoin has today?

1578
02:38:58,263 --> 02:39:04,263
Well, I mean, the sidechains Bitcoin has today are federated, I guess, you know,

1579
02:39:04,263 --> 02:39:11,702
liquids and Rootstock and statechains, although they're not exactly sidechains, and Fedimint,

1580
02:39:12,263 --> 02:39:19,543
also not exactly a sidechain, but also federated. And so drivechain and the peer-to-peer sidechains

1581
02:39:19,543 --> 02:39:32,182
are, there is no kind of fixed federation. The peg is operated as a result of, you know,

1582
02:39:32,182 --> 02:39:40,263
random miners participating in a protocol. So it's much more dynamic and anonymous in its operation.

1583
02:39:43,668 --> 02:39:56,668
Thank you very much. My other question is, you mentioned earlier in this conversation how the Ordinal's popularity caused a brief but very noticeable spike in the fee market.

1584
02:39:57,668 --> 02:40:12,668
Do you have a view on the potential for merged mining to be an important and valuable positive factor in Bitcoin's long-term economic security, meaning miner revenue as the subsidy declines?

1585
02:40:12,668 --> 02:40:32,668
I'm not sure. I mean, effectively, there's a supply and demand constraint. So, you know, Lightning enables a lot more transactions, but they're much cheaper transactions.

1586
02:40:32,668 --> 02:40:42,668
And so that, you know, the scalability of a layer to increase the supply, and if the supply gets ahead of demand, then it won't flow much capacity back.

1587
02:40:43,668 --> 02:41:00,668
And, you know, so for example, if for some, you know, there was some technological solution that enabled Bitcoin to have like super large blocks with no side effects, no centralization side effects, the fee market will collapse, right, because there's an excessive capacity.

1588
02:41:00,668 --> 02:41:22,668
And so I think it's a bit of a balance. Like, we don't really know. Like, today, the technology is hard to scale. And so that tends to create fee pressure. But if the scalability got ahead of, you know, transactional demand, the fees could collapse too, right? So I'm not sure.

1589
02:41:23,668 --> 02:41:41,668
But there's the area of the rectangle that matters, like the fee rate, you know, and the supply and demand graph, the price, like dollars per transaction, that may be high or low, but that may minimize, maximize, or something else.

1590
02:41:41,668 --> 02:42:09,668
Well, I mean, I think, yeah, I mean, I think there's, I don't know what the economic term for it is, Paul probably knows, where sort of capacity creates new demand. So if there's capacity to make enormous amounts of super low cost transactions that have the characteristics you're interested in, then other people will get busy and make new applications or use cases that use it up.

1591
02:42:09,668 --> 02:42:19,668
And then, you know, the supply will get used up, and more value will be delivered. So presumably, there'll be a way for people to get paid for it, for miners to get paid.

1592
02:42:21,668 --> 02:42:30,668
Oh, yeah, they may have different, you know, there may be different use cases. A lot of demand at a small or low fee.

1593
02:42:31,668 --> 02:42:32,668
Right.

1594
02:42:32,668 --> 02:42:33,668
Less demand at a higher fee.

1595
02:42:34,668 --> 02:42:47,668
But I mean, do you see what I mean, though? If there is, even if there's a lot of demand, but an infinite supply, then the fees will basically be converged on zero, right? Or near it.

1596
02:42:48,668 --> 02:42:59,668
Right, that'd be a rectangle that is a very, very wide, and not very tall. But you still multiply, you know, because if it's like a million, if it's like 100 billion transactions that pay one fat each.

1597
02:43:00,668 --> 02:43:01,668
Yeah.

1598
02:43:02,668 --> 02:43:16,668
Yeah, it's just kind of a question. It's kind of an unknown. I don't think anyone really knows per se. But yeah, why don't we go? Why don't we let people speak if they want to speak?

1599
02:43:17,668 --> 02:43:23,668
Oflo, thanks for waiting so long to speak. You've had your hand up a long time. Please go ahead with your question or comment.

1600
02:43:24,668 --> 02:43:26,668
Hey, what's going on? Adam, long time.

1601
02:43:28,668 --> 02:43:29,668
Hi.

1602
02:43:29,668 --> 02:43:34,668
So, so I have a few questions, three, three main questions.

1603
02:43:35,668 --> 02:43:57,668
And maybe my first one is for Adam. We touched earlier on simplicity, and I'm actually a proponent of simplicity. I like the concept of it. Obviously, it's very complicated. So a lot of people still don't understand it, myself included. But it seems to make sense above the grade.

1604
02:43:57,668 --> 02:44:16,668
In terms of your perspective, and I've had debates about this, some proponents who do not like the concept of drivechain do not like the concept of simplicity due to the fact that simplicity essentially allows what I've heard people say a free for all when it comes to Bitcoin.

1605
02:44:16,668 --> 02:44:28,668
So my question for you would be, what would be your response to that in terms of simplicity opening the doors to limitless potential of what can be done with Bitcoin?

1606
02:44:28,668 --> 02:44:57,668
Um, yeah, I mean, I guess, you know, either you want to extend sparsity or you don't. But also, simplicity is, you know, it's still in the same model as Bitcoin scripts. So it's still UTXO based, it still doesn't have states, transactions are behaving the mental the same way, you know, if they're valid, once they stay valid.

1607
02:44:58,668 --> 02:45:27,668
And so on, they don't become invalid afterwards. So it preserves Bitcoin script semantics at a high level. And it just makes it easier to implement new opcodes. And of course, you know, opcodes are not going to get used or you won't accidentally receive a transaction that you don't understand because

1608
02:45:28,668 --> 02:45:46,668
you have to generate the address, but you have to have a wallet that is aware of what your, you know, the spend conditions and what you're asking for. So effectively, you might have different use cases with wallets with more capabilities or something. So it's still kind of opt in.

1609
02:45:47,668 --> 02:45:56,668
That was a great question, Oflo. Since we're almost out of time, I wanted to move to the next speaker. But thank you for that really good question. Portland, please go.

1610
02:45:57,668 --> 02:46:21,668
Oh, thank you for having me up, first of all. And my second question is, in terms of finding consensus to get your software to propose software activator, have you considered alternatives, such as maybe just forking Bitcoin or colluding directly with miners, and then having people that want to use your software, they can use it because it's technically a software, right? It's backwards compatible.

1611
02:46:22,668 --> 02:46:47,668
We sort of discussed this to some extent. I mean, we actually said, like, we were kind of talking hypothetically, like, should you just, because at first I thought about it, yes, because I thought I'd been thinking about this for a while, because I just think this is a, this is obviously a good idea that has basically zero risk. And the risk of not doing it is sort of high. And it's the critics seem pointlessly deluded.

1612
02:46:48,668 --> 02:47:00,668
So yeah, I thought about a lot of things, including all those things. And we, the idea of maybe just go to the miners and just ask them to miner activated software, but we kind of, it's in the recording, I think, a little bit of what we discussed.

1613
02:47:01,668 --> 02:47:02,668
Yeah. Thank you.

1614
02:47:02,668 --> 02:47:03,668
Thank you.

1615
02:47:08,668 --> 02:47:24,668
All right. Oflo has another question. So we'll bring it back up. Thank you so much for your second question, Oflo. Go ahead. Oflo, mic is yours for your additional question.

1616
02:47:24,668 --> 02:47:42,668
Awesome. Thank you. My other question is for Paul. And basically, there seems to be an alternative way that drivechains could be activated. It's not really drivechains. There is something called hand crank that was written up by testnet.

1617
02:47:43,668 --> 02:47:56,668
Have you looked into that? And why not create drivechains on something such as liquid or on top of a space chain or hand crank like I was suggesting earlier?

1618
02:47:57,668 --> 02:48:12,668
Well, my view is that what BIP300 does is what drivechain does is significantly different from how space chain and liquid work.

1619
02:48:13,668 --> 02:48:28,668
Liquid is like federated. The whole point of BIP300 is to ditch the federation. So to do it on top, it's kind of like having a car, maybe a little unfair to say, but it's like having a horse pull a Ford Model T or something.

1620
02:48:29,668 --> 02:48:49,668
It's kind of like the whole point of the Model T is not to have the horse. So as for space chains, the space chain was invented after drivechain, and it's basically a reaction to the ability to steal the coins, which is like the so-called ability to steal the coins by just instead preemptively destroying all the coins.

1621
02:48:50,668 --> 02:49:08,668
I think already every drivechain is a space chain. The space chain doesn't have the key feature against, but that would be kind of like instead of having a horse pull a Model T, that would be like instead just getting rid of the horse and the engine so that it's just a cart with wheels.

1622
02:49:08,668 --> 02:49:17,668
But the whole point of going further in the wrong direction, because the whole point is to be able to go back and forth with the coins.

1623
02:49:18,668 --> 02:49:34,668
And then the hand crank thing I think is just a mistranslation or a misunderstanding or something where they say it's not a soft fork, but it's a group of miners agreeing to orphan blocks if something happens, which that actually is a soft fork, if you ask me.

1624
02:49:34,668 --> 02:49:51,668
So what that would be is the equivalent of just going to have the miners and just asking them to unilaterally switch their software over to software that I release and do it like that.

1625
02:49:51,668 --> 02:50:03,668
So it's basically just bad idea plus a lot of lack of self-awareness. So that idea just with less self-awareness, I think is my guess.

1626
02:50:04,668 --> 02:50:15,668
But that's interesting. I think people should look into that so that they themselves understand that if they really wanted to stop drivechains, there's very little they could actually do about it.

1627
02:50:15,668 --> 02:50:24,668
I'm primarily the person holding it up by not doing any of these weird maneuvers that people have done.

1628
02:50:24,668 --> 02:50:31,668
If the miners could just do it, the miners could just do drivechain tomorrow and I would know what would be able to stop them, I think, including me.

1629
02:50:31,668 --> 02:50:34,668
I think all that's a little bit of confusion, convolution.

1630
02:50:34,668 --> 02:50:41,668
I think SuperTestNet agreed with that analysis regarding hand crank as a soft fork, if I'm not mistaken, I think.

1631
02:50:41,668 --> 02:50:46,668
Okay. Do we have a third question or do we have a question from someone else?

1632
02:50:46,668 --> 02:50:51,668
Yeah, Ophal, before your final question, I just want to mention to everyone here that this is a weekly space.

1633
02:50:51,668 --> 02:50:56,668
We meet every week on Friday, sometimes with an invited guest as we have today.

1634
02:50:56,668 --> 02:51:03,668
We're very lucky to have Adam here and sometimes just an open mic the whole time and no invited special speaker.

1635
02:51:03,668 --> 02:51:08,668
So follow LayerTwo Labs here on Twitter for more of these weekly Twitter spaces.

1636
02:51:08,668 --> 02:51:23,668
And you can also find important resources at Layer2Labs.com, including the software you can download to run the drivechain test net and become more familiar with this incredibly powerful technology that can come to Bitcoin.

1637
02:51:23,668 --> 02:51:27,668
Ophal, go ahead with your next question, if you'd like.

1638
02:51:27,668 --> 02:51:35,668
OK, this last question is it it's borderline, more or less the main debating point.

1639
02:51:35,668 --> 02:51:45,668
And you and I have talked about this, Paul, the main debating point that most people have towards drivechains is MEV incentives with the Bitcoin miners.

1640
02:51:45,668 --> 02:51:50,668
Essentially, let's say a big block sidechain.

1641
02:51:50,668 --> 02:52:01,668
Essentially, the miners would have to mine that or run a full node to validate that because of the profitability of that sidechain, if it was highly profitable.

1642
02:52:01,668 --> 02:52:19,668
So my question to you, is it possible that you could at some point even not now if you don't want to, but at some point, could you specifically touch on all the points why MEV will not be an issue specifically for drivechains?

1643
02:52:19,668 --> 02:52:26,668
And if a big block sidechain was highly adopted?

1644
02:52:26,668 --> 02:52:36,668
Yes. So there was like last week or two weeks ago, maybe I'm losing track of which week, but there's like an enormous in-depth discussion about this.

1645
02:52:36,668 --> 02:52:43,668
And, you know, I asked Alex B to write down an example over and over again, and he eventually did not.

1646
02:52:43,668 --> 02:52:48,668
And then he eventually started deleting his tweets and bringing this up.

1647
02:52:48,668 --> 02:52:56,668
And so but because what it ended up meaning is that by MEV, what people just mean is whenever miners earn more money.

1648
02:52:56,668 --> 02:53:00,668
And so the definition has morphed enormously in Ethereum.

1649
02:53:00,668 --> 02:53:07,668
What it meant was miners earning money by by causing mischief with the chain.

1650
02:53:07,668 --> 02:53:09,668
By like swapping the orders of transactions.

1651
02:53:09,668 --> 02:53:15,668
So that is what I presumed people were meant when they first talked about because we had just we have like an Ethereum sidechain.

1652
02:53:15,668 --> 02:53:17,668
So I thought, oh, this is what they're talking about.

1653
02:53:17,668 --> 02:53:27,668
They're saying that whatever the MEV on the Ethereum world would affect layer one, it would affect other people.

1654
02:53:28,668 --> 02:53:35,668
And so then I have to do is point out that, of course, it won't, because what happens is just a transaction like that.

1655
02:53:35,668 --> 02:53:42,668
There may be MEV in the Ethereum land, Ethereum, the altcoin or Ethereum, the sidechain, but that affects them and they sign up for that.

1656
02:53:42,668 --> 02:53:48,668
That's actually what they want. They want to be in Ethereum land and it does not affect layer one at all.

1657
02:53:48,668 --> 02:53:53,668
But what it ended up what ended up meaning is either what either happened was that Alex B.

1658
02:53:53,668 --> 02:53:58,668
And or Bob McAlrath. Probably both.

1659
02:53:58,668 --> 02:54:02,668
But what they would probably happen was they started the argument not knowing how.

1660
02:54:03,668 --> 02:54:08,668
Blindridge mining worked or how to keep track of the differences between L1 and L2.

1661
02:54:09,668 --> 02:54:17,668
And then as the argument went on, they slowly retreated to a completely different argument where they wanted to just say they're trying to, you know.

1662
02:54:17,668 --> 02:54:20,668
Get their way out of the corner they had painted themselves into.

1663
02:54:21,668 --> 02:54:25,668
That was either what they did or either the whole time they have, whatever, whatever the case may be.

1664
02:54:26,668 --> 02:54:30,668
What they mean by miner MEV is just that the miners earn more money.

1665
02:54:31,668 --> 02:54:35,668
And so to them, a transaction on layer one that pays a high transaction fee is MEV.

1666
02:54:36,668 --> 02:54:40,668
And if transaction fees as a whole go up because of ordinals or whatever, that's MEV.

1667
02:54:41,668 --> 02:54:45,668
And if people just decide to write the miners a check, that's MEV.

1668
02:54:46,668 --> 02:54:48,668
If they decide to buy T-shirts from the miners, that's MEV.

1669
02:54:49,668 --> 02:54:54,668
If someone, you know, if the miner gets a job on the side, that's MEV.

1670
02:54:55,668 --> 02:54:58,668
The miner is playing the violin in the subway station.

1671
02:54:59,668 --> 02:55:02,668
So it ends up just not meaning anything.

1672
02:55:03,668 --> 02:55:13,668
But there will be, by that definition, there will be enormous MEV because the whole point of having a whole, every blockchain design is based on collecting transaction fees.

1673
02:55:14,668 --> 02:55:22,668
To deter spam or denial of service or to sort people from being, you know, get out, get rid of Sybil attack and who should make it into a block.

1674
02:55:23,668 --> 02:55:28,668
If there's only a certain amount of space, if the blocks are finite, who should make it in and which transactions have been seen.

1675
02:55:28,668 --> 02:55:30,668
So that all relies crucially on transaction fees.

1676
02:55:31,668 --> 02:55:38,668
And I intend the whole point of this design is for the miners to be making huge amounts of money, ideally from popular sidechains.

1677
02:55:39,668 --> 02:55:40,668
So that's the idea.

1678
02:55:41,668 --> 02:55:46,668
But what I really think happened was they had absolutely no idea what they were talking about whatsoever.

1679
02:55:47,668 --> 02:55:50,668
And they just thought they'd use the MEV word as a critique.

1680
02:55:51,668 --> 02:55:55,668
And so I think that you can look at that and you can go and look into the historical tweets.

1681
02:55:55,668 --> 02:55:59,668
You can look at my Twitter account and you can look at the highlights, I think some of them.

1682
02:56:00,668 --> 02:56:07,668
We'll touch on it or you can just search on Twitter and you can read this enormous debate that all ended up amounting to nothing.

1683
02:56:08,668 --> 02:56:15,668
Because you would, of course, you would think they're talking about MEV on an Ethereum sidechain or Bitcoin having some kind of negative effect on one.

1684
02:56:16,668 --> 02:56:18,668
That's what you would think, which that never happens.

1685
02:56:19,668 --> 02:56:24,668
But it ended up being that they were just talking about literally any time that miners earn more money.

1686
02:56:26,668 --> 02:56:42,668
I wonder if they, I didn't read all of those threads, but I was wondering if they meant that in the hypothetical situation that a sidechain got very centralized, that there would be some kind of secret information, right?

1687
02:56:42,668 --> 02:56:48,668
Like a stream of transactions that was not publicly available.

1688
02:56:49,668 --> 02:56:54,668
So that, you know, a few servers would have an advantage.

1689
02:56:55,668 --> 02:56:57,668
I wouldn't necessarily call it MEV.

1690
02:56:58,668 --> 02:57:03,668
Exactly. This is a situation where the people running the sidechain node have some advantage.

1691
02:57:04,668 --> 02:57:08,668
So you could call that, that would be the MEV on the sidechain that is in the sidechain world.

1692
02:57:09,668 --> 02:57:23,668
But when they try to bid for a blind merge mining, they could get up and then they're going to have to keep bidding because the person next to them knows that whatever the private order flow info or whatever it is, is worth the sidechain blocks fees are worth seven.

1693
02:57:24,668 --> 02:57:33,668
And then there's two extra on top of that, seven Bitcoin plus two Bitcoin of call it EVM EV or something or whatever you want to call it.

1694
02:57:33,668 --> 02:57:51,668
Well, I mean, maybe you can drive the example, but if the case is that, you know, somebody could make a drivechain, right? It's permissionless, they can do crazy stuff.

1695
02:57:51,668 --> 02:58:01,668
So they make a drivechain where the wallet model is you send the transaction to a central server and they collect the fees on it. And if you don't pay them a high fee, they won't process it.

1696
02:58:02,668 --> 02:58:08,248
right? then they choose not to you

1697
02:58:08,248 --> 02:58:09,468
know,

1698
02:58:10,888 --> 02:58:15,388
blind merge mining and you don't know the transactions and they just have an advantage

1699
02:58:15,388 --> 02:58:17,222
in mining or something, right,

1700
02:58:18,242 --> 02:58:20,162
like that, I think might be getting ready.

1701
02:58:20,222 --> 02:58:25,202
You see the problem with that is, of course, can argue on L01, can open Bitcoin Bank,

1702
02:58:25,582 --> 02:58:26,722
Paul's Bitcoin Bank or something.

1703
02:58:26,902 --> 02:58:28,242
It doesn't really have anything to do with Drivechain.

1704
02:58:28,582 --> 02:58:31,776
It's like, in order to reach that point, have to basically assume

1705
02:58:31,816 --> 02:58:33,536
that there's no blockchain anymore anyway.

1706
02:58:33,676 --> 02:58:35,916
So it's kind of like a proof by contradiction in a way.

1707
02:58:36,856 --> 02:58:39,756
I can already say, collect...

1708
02:58:39,756 --> 02:58:41,476
That's kind of similar to what Liquid is.

1709
02:58:41,656 --> 02:58:45,836
like you send to a multisig output on L1, we have our own server,

1710
02:58:47,668 --> 02:58:58,668
We collect the transaction fees on our thing. And then if you don't pay, we don't let you in. And then we don't give you your money back. Whether or not you get your money back is up to us. So.

1711
02:58:59,668 --> 02:59:20,668
I mean, I wasn't thinking about it that way, but kind of. But I mean, I think they're thinking that, you know, because it's supposed to be peer to peer, that then there's a barrier to mining it. But really, I do agree that that's not really a blockchain anymore.

1712
02:59:20,668 --> 02:59:45,668
So I think they're kind of worried about a failure mode, right? And but I mean, I think, as we were discussing earlier, it may not be possible to sort of, you know, provide assurance in a smart contract terms that some of the, you know, aggressive hypothetical centralizations.

1713
02:59:45,661 --> 02:59:50,661
Can be like mathematically prevented, right? You know, if people sign up for stupid stuff.

1714
02:59:50,661 --> 03:00:13,661
Well, this is a very important point about like, like, again, this is opt in situation. So like, I have a story about we never got back to the whole mining pool breaking into the mining pool. So I have a little bit because it's so I have my story about credit is which I've told many times is that I would go to this restaurant over and over again for takeout food because their food was delicious.

1715
03:00:14,661 --> 03:00:27,661
I went all the time. I knew the owner and the owner's son who was the manager. And one time I would always go late at night and wait for the close and I forgot my wallet one time because I hate when they have to wait. They close at 10 and I was getting there at 955.

1716
03:00:28,661 --> 03:00:45,661
I was rushing to get there. And of course, one time I forgot my wallet and obviously they let me leave with the food and they just pinned up the receipt. And then I paid the next time I was there because I come all the time. But the food is fresh and I'm hungry. So the food is going to go bad.

1717
03:00:46,661 --> 03:00:58,661
And I am going to go hungry for no reason. And they're not going to collect the money. Like if you see what I mean, like if they if we wanted to be like what you might call the perfectly cryptographically mathematically secure universe.

1718
03:00:59,661 --> 03:01:08,661
There would be like Peter Todd would be patrolling. He'd shoot anyone who gave anyone anything on credit because it would encourage credit or something.

1719
03:01:09,661 --> 03:01:24,661
But credit is something that helps two people, the buyer and seller, get what they want because of their belief. Credo being Latin for believing, I believe.

1720
03:01:25,661 --> 03:01:43,661
So credit is not like credit has problems, but it by itself is actually just a good thing. And the option to use the large block by chain is a free option. You could say, well, maybe I would prefer I almost never use it. But then I forgot my wallet at my house and now I have no choice. Well, it's better than nothing.

1721
03:01:44,661 --> 03:01:59,661
Which is what it would be if before, if there was no credit possible, that food would have gone to waste. And so, yeah, I think there is too much of an obsession on like the to stamp everything out at all costs mentality.

1722
03:02:00,661 --> 03:02:19,661
And like this is how far like this is the hoops you have to jump through in order to make it totally, completely ironclad is like you need HGLCs that are enforced on L1. You need the L1 fee rate to be so that the lightning dust isn't what it is.

1723
03:02:19,661 --> 03:02:32,661
You need to have something like Ark with lots of moving pieces and you need to have all the liquidity tied up into it that's equal to the total payment volume. You need to do an awful lot just to like get really basically nothing.

1724
03:02:32,661 --> 03:02:50,661
Because at the end of the day, 51% miners can still filter messages out. At the end of the day, you can still get eaten by the L1 fee going up. So that's my one thing. But then the idea of the miner costs, like there's miner costs for the hashers and then there's the pool costs.

1725
03:02:50,661 --> 03:03:06,661
So we cleanly divide it into two. We got to cleanly divide it into two off the bat. We just separate this universe. The pools and the miners. No one has ever cared about the miners costs and nor should there's a node costs, which we should care about.

1726
03:03:06,661 --> 03:03:21,661
But that is only one thing on the list of miners cost. If you care about the miners cost, you should be in favor of removing the upward difficulty adjustments from Bitcoin and you should basically you're removing proof of work from Bitcoin at that point.

1727
03:03:21,661 --> 03:03:41,661
So if you care about how difficult it is to mine a block, you actually maybe you don't know it yet, but you you're against proof of work. And these people are all just be ignored in this. I think the pools is slightly different story where the question is, how can you break into a pool?

1728
03:03:41,661 --> 03:03:59,661
The thing with the pools is they have a certain size that it's just based on statistical variance and really nothing else. Like a pool below a certain size, 1.500, just is never viable. And for the same reason that a solo miner below a third size is never viable.

1729
03:04:00,361 --> 03:04:05,861
Yeah, I mean, think, unfortunately, it's not really viable to have more than, I don't

1730
03:04:05,861 --> 03:04:11,661
know, 20, 30, 40 pools or something because the variance is too bad below that.

1731
03:04:11,721 --> 03:04:12,921
See, this the other

1732
03:04:12,979 --> 03:04:16,959
thing that you say other word, unfortunately, but I think that this is actually a

1733
03:04:16,959 --> 03:04:22,239
mistake that this is like a hot off the press because it was kind of like a newer idea

1734
03:04:22,239 --> 03:04:23,099
of mine or something.

1735
03:04:23,419 --> 03:04:26,139
it's like if people say

1736
03:04:30,661 --> 03:04:53,661
And they each have 1% of the hash rate. 101th person comes. There's no way that they can at least have, they can each have 1% of the hash rate, you know, it has to add up to 100%, basically. So like, you're either hoping that Bitcoin stays so small that everyone involved can mine. Because wouldn't it be great if everyone in Bitcoin could mine?

1737
03:04:53,661 --> 03:05:18,661
But if everyone in Bitcoin mines, like, I mean, I guess you could say maybe there's something else that has different statistical properties that would just be better or something. But the way I'm looking at it is how lucky we are that we have pools at all, because otherwise, there would really only be 20 people who could mine at all. And there would be, it would be like, just down to that list of 20 people.

1738
03:05:18,661 --> 03:05:45,661
So luckily, we have pools. And a pool is now I think of a pool, like you have like the immersion cooling, cooling the miners. The pool is like this weird technology that it de-variances the miner. It makes it so that actually, what would normally be an economy of scale that would kill the miner instantaneously, is now something that they can almost completely ignore.

1739
03:05:45,661 --> 03:05:57,661
But this is besides the point of what I was trying to say, which is that, for better or for worse, that you cannot have a pool that is smaller than like, whatever, 2%, or whatever it is, 145.

1740
03:05:57,661 --> 03:06:10,661
Right.

1741
03:06:27,661 --> 03:06:52,661
So there's like a minimum pool size that is from the variance already. And so if you have 20 pools, and they each have 5%, then what you're really thinking about with this idea of can I break, you mentioned this phrase, like break in and become a new, break into the pool or break into the pool market, you actually need a very intriguing ingredient.

1742
03:06:53,661 --> 03:07:13,661
And that ingredient is you need one of the 20 pools to be making a mistake. If all 20 of them are absolutely perfect, sublime pool, then it's kind of as if you go to a restaurant every day, and they charge you tiny, one cent, and they give you the most amazing meal you've ever had in your entire life, and you have no problems going there.

1743
03:07:14,661 --> 03:07:20,661
So you go to the same restaurant every day, but it's the greatest restaurant in the universe. So it doesn't matter.

1744
03:07:21,661 --> 03:07:35,661
So the reason I bring this up is because if no one is making a mistake, and then you have 20 pools, and they each have 5%, and 5% is like the floor, then nothing bad is actually happening.

1745
03:07:35,661 --> 03:07:55,661
And so the fact that you can't break into the pool market is only because there's no reason for anyone to break in at all. But then as soon as they start, if they start tampering or something, start making a mistake, or they orphan block, or they charge too much money, their fees are too high, then you can break in, then you open your own restaurant.

1746
03:07:55,661 --> 03:08:11,661
You say, I don't like this restaurant anymore. You really need all 20 pools. Every restaurant in this town is doing things, so no one's washing their hands or something. Everyone's doing something wrong, and you offer something.

1747
03:08:11,661 --> 03:08:26,661
But I think something weird has happened with the way that the pools are inherently proportional. Unlike mining, and definitely unlike running a full node, where it's like your own cost of error correcting the chain has nothing to do with what other people do.

1748
03:08:26,661 --> 03:08:44,661
With mining, there's more dependence where you have to mine. In order to be a miner, you have to mine at a rate that is at a cost per hash that is competitive, or you're going out of business. With the pools, it's directly proportional.

1749
03:08:44,661 --> 03:08:57,661
You cannot create a new pool. If you create a new pool, you need to attract 5% of the hash, but they're not going to switch over to you without a reason. That reason has to be the existing pools are making a mistake.

1750
03:08:57,661 --> 03:09:16,661
So I think that this is a big concept, the idea that the costs should be split up into the node costs, which are very important, the actual mining hasher costs, which are totally unimportant and totally incidental, and then the cost of actually starting a new mining pool.

1751
03:09:16,661 --> 03:09:38,661
It inherently depends on whether or not the existing set of pools is making some kind of mistake. It could be possible that we're in nirvana, everything's going perfectly, and it's also impossible to start a new mining pool because the existing set is so perfect. So this is a weird – I've now split it into three different types of pools.

1752
03:09:38,661 --> 03:09:57,661
I think one of the things decentralization provides is censorship resistance, and there have been one or two pools that have talked about selective censorship and then backed off of it.

1753
03:10:08,661 --> 03:10:32,661
So if you have independent mining entities like individuals and small and large companies, then there are pools. So I think Stratum V2 helps quite a lot because it means that a miner can choose their own transactions and still benefit from the variance reduction, whereas today, artificially, the pools are doing the transaction selection. So that's a sort of step forward.

1754
03:10:32,661 --> 03:10:48,661
Well, I mean, think about it like this, because the pool can still – if you use Stratum V2, the pool can still trump the – I can still say we don't do that. So actually, Stratum V2 may make it worse, but I have an even different unification of this.

1755
03:10:48,661 --> 03:11:07,661
See, the reason why it may actually be worse with Stratum V2 is – the real reason is it's actually not affected at all because the mining pool can still veto the block. You know what I mean? Maybe you propose – you do your own job selection, but it's actually the pool that determines which job is picked.

1756
03:11:07,661 --> 03:11:35,661
And if that were the reverse, it would be worse because the mining pool can connect their own – connect themselves as like a tiny miner. This is all besides the point though because – I could get into this weird theory if you want, but the actual – the important thing is that what the miners – the mistakes that the miners can make are censoring transactions that pay a fee.

1757
03:11:35,661 --> 03:11:48,661
So if they discriminate for some reason other than sats per byte or sats per vbyte or whatever, that is the mistake that they would make, and that is the mistake that the mining pool would make that I mentioned before.

1758
03:11:49,661 --> 03:12:00,661
And it's actually – there's already a – we don't even need Stratum V2, although it may help with – Stratum V2 helps with lots of other things such as the encryption and authentication and stuff.

1759
03:12:01,661 --> 03:12:21,661
There's already a way of proving the mistake, and all that is is to broadcast the censored transaction because it by itself is a very small thing that anyone can check, anyone with a full node, can very easily check and see that it is a valid transaction that pays a fee that is high.

1760
03:12:21,661 --> 03:12:35,661
And then all they have to do is wait 20 minutes, and then they will know that for some reason there is a mistake, and they will also know exactly what to do about it, which is to start a new pool themselves that includes this very transaction that is censored.

1761
03:12:35,661 --> 03:12:57,661
So actually if you look at the full node as the cost of error correcting the chain, the cost of detecting errors in the blockchain, you could then look at the mere presence of a censored transaction as part of the error correcting path for fixing the pool situation.

1762
03:12:57,661 --> 03:13:13,661
And so really at the end of the day, there's already a built-in thing that fixes the pool error problem because broadcasting the transaction is exactly what the censored person is going to do and are upping the fee.

1763
03:13:14,661 --> 03:13:38,661
So the entire thing, in my opinion, has just been kind of a complete confusion, and partly basically what happened was the full node cost is very, very important to care about, and I think that the definition is like someone knocked over a glass of water and the ink spilled out into these other areas where it doesn't really belong.

1764
03:13:38,661 --> 03:14:03,661
The cost of running the full node is the important thing, and these other costs are just like some kind of paranoia or something until all the theory was worked out, which is sort of what I'm trying to say is that already the fact that there's a fee-paying transaction, that's obviously not enough to make sure that it gets into a block.

1765
03:14:03,661 --> 03:14:08,661
But if paying the fee is not enough, then what is going to be enough?

1766
03:14:08,661 --> 03:14:22,661
Because already we assume that the mining infrastructure wants to maximize its revenues, and the straightforward way to maximize the fee revenues is to include the fees that pay the most, to sort all the transactions by fee before.

1767
03:14:23,661 --> 03:14:25,661
So this is already what everyone wants to have happen.

1768
03:14:25,661 --> 03:14:36,661
So I think the thing that would be best would be if there were just stuff in place to flag whenever a transaction was censored, but that's what I'm trying to say is there already is that.

1769
03:14:36,661 --> 03:14:41,661
That already is the fact that the people are going to broadcast this censored transaction and say, look, it pays a fee.

1770
03:14:42,661 --> 03:14:48,661
So it's like everything that people would have wanted already kind of exists, or I'm saying the system already works.

1771
03:14:48,661 --> 03:15:04,661
And as you were saying, no one was ever actually really able to get away with even doing any transaction censorship for any length of time because maybe they talked about it once or twice, but nothing actually happened because nothing really can happen.

1772
03:15:11,661 --> 03:15:20,661
Well, I think ultimately what people care about is the censorship-resistant characteristic is reinforced, and there are a few different.

1773
03:15:20,661 --> 03:15:31,661
I don't think there are some more technological directions which could get closer to it with privacy and two-stage commitments and strategy two and different things.

1774
03:15:31,661 --> 03:15:33,661
So the effect of it is working.

1775
03:15:33,661 --> 03:15:40,661
I do think privacy is actually the direction to go exactly, where it's like if you're worried about transaction censorship, then it would actually be a good deterrent.

1776
03:15:40,661 --> 03:15:55,661
If everything was so private that there was really just no basis for anyone censoring, then it really would be irrational to censor because – so yeah, I think that would be the way to go.

1777
03:15:56,661 --> 03:16:04,661
I think just having sidechains just has more block space for people to send coins to themselves, what we call the deniability.

1778
03:16:04,661 --> 03:16:22,661
Or if you use the Zcash sidechain, I actually think the whole thing would be improved in that direction, but I kind of already think that this is like a – the idea of being able to break in as a new mining pool, it doesn't necessarily – let's say there's only one pool, but it's absolutely perfect.

1779
03:16:23,661 --> 03:16:28,661
If there's one pool, but it's perfect, then maybe you can never break in because you're like, how can I break in?

1780
03:16:28,661 --> 03:16:32,661
But on the other hand, it's a perfect pool and never censors.

1781
03:16:32,661 --> 03:16:34,661
It always charges zero fees or something.

1782
03:16:34,661 --> 03:16:38,661
It's just evolved to be the optimal thing, so it's kind of like –

1783
03:16:38,661 --> 03:16:44,661
I mean hypothetically, I mean with the current technology, that would be a bit of a central point of failure there, right?

1784
03:16:44,661 --> 03:16:49,661
I mean if it misbehaved, I guess somebody would start another pool pretty quickly and people would switch to it.

1785
03:16:49,661 --> 03:16:51,661
Yeah, well, that's the point that I'm saying.

1786
03:16:51,661 --> 03:16:53,661
It's because of the misbehavior that the new pools appear.

1787
03:16:53,661 --> 03:16:58,661
So it's not as though you always need – it's not as though you always need the ability to start a new pool.

1788
03:16:58,661 --> 03:17:01,661
You only need it in a situation where the pool is not perfect.

1789
03:17:01,661 --> 03:17:13,661
The pool is perfect that it actually doesn't matter if you can start another pool and compete because they partly compete on their brand and stuff, which they must because the brand means they have something to lose.

1790
03:17:13,661 --> 03:17:16,661
They charge a fee that covaries with their reputation.

1791
03:17:16,661 --> 03:17:32,661
The origin of the word brand, which is a fun little fact that a lot of people know but probably some people in the audience don't know, is the cattle brand because the cattle would go off into the wilderness to some extent, and they might mix.

1792
03:17:32,661 --> 03:17:38,661
So it's hard to tell which farmers are feeding their cattle well or the brand.

1793
03:17:38,661 --> 03:17:43,661
They would literally brand the cattle, and then it was against the law to use someone else's brand.

1794
03:17:43,661 --> 03:17:54,661
So you had to have the – enforce this suspension of freedom of speech in a very narrow sense, which is that you could not brand the cattle using someone else's brand.

1795
03:17:54,661 --> 03:17:58,661
But when you have a brand, it means you have something to lose.

1796
03:17:58,661 --> 03:18:09,661
And so that's – you charge fees, and that is what allows good pools or good players to distinguish themselves from bad people.

1797
03:18:09,661 --> 03:18:15,661
So when you go into CVS or something, you see Crest versus Colgate, and then you see something that you've never heard of before.

1798
03:18:15,661 --> 03:18:24,661
You think, well, I don't really know what that is, so I will ignore that, and I will buy Crest because they're all basically the same cost.

1799
03:18:24,661 --> 03:18:29,661
But then maybe as time goes on, you get a little bit of a reputation or something.

1800
03:18:29,661 --> 03:18:32,661
Adam and Paul, we've been going for over three hours.

1801
03:18:32,661 --> 03:18:36,661
Do you keep on going or do you want to –

1802
03:18:36,661 --> 03:18:39,661
Maybe we should pause and pick it up for next time.

1803
03:18:39,661 --> 03:18:43,661
That sounds good. This has been an outstanding conversation.

1804
03:18:43,661 --> 03:18:44,661
Okay, cool.

1805
03:18:44,661 --> 03:18:46,661
Thank you, Adam, so much for being here.

1806
03:18:46,661 --> 03:18:48,661
Yeah, thanks a lot.

1807
03:18:48,661 --> 03:18:50,661
Thanks for the discussion.

1808
03:18:50,661 --> 03:18:54,661
Yeah, we really appreciate you sharing your thoughts today on drivechain and related matters.

1809
03:18:54,661 --> 03:18:59,661
This has been a – Austin, did you have something to say as well?

1810
03:18:59,661 --> 03:19:04,661
No, I'm just going to thank Adam back for joining.

1811
03:19:04,661 --> 03:19:06,661
Yeah, this has been a great conversation.

1812
03:19:06,661 --> 03:19:12,661
We've had on stage, just for those who've joined in the middle, Austin Alexander, one of the co-founders of LayerTwo Labs,

1813
03:19:12,661 --> 03:19:23,661
Paul Sztorc, one of the co-founders of LayerTwo Labs, and CryptAxe, one of the co-authors of the BIPS 300 and 301, the drivechain BIPS.

1814
03:19:23,661 --> 03:19:25,661
So thank you, Dr. Beck.

1815
03:19:25,661 --> 03:19:27,661
There's a co-founder.

1816
03:19:27,661 --> 03:19:29,661
There's four co-founders.

1817
03:19:30,661 --> 03:19:34,661
There's four co-founders and two co-authors.

1818
03:19:34,661 --> 03:19:40,661
Thank you, Dr. Beck, for being here today, and we'd love to have you back again soon.

1819
03:19:40,661 --> 03:19:43,661
So once again, these are weekly Friday Twitter spaces.

1820
03:19:43,661 --> 03:19:47,661
Sometimes we go for six hours here on Twitter.

1821
03:19:47,661 --> 03:19:51,661
And you can find more information at layer2labs.com.

1822
03:19:51,661 --> 03:19:58,661
Sorry, Vogue Blocker, Vogue Blackheart, we didn't get time for your questions, but I really appreciate you coming up.

1823
03:19:58,661 --> 03:20:00,661
Please come next time.

1824
03:20:00,661 --> 03:20:04,661
And you can go to layer2labs.com to download the drivechain software.

1825
03:20:04,661 --> 03:20:08,661
Everyone can do that and get more involved in this important movement.

1826
03:20:08,661 --> 03:20:17,661
We have over 700 members in our Telegram group, where we discuss drivechain all the time,

1827
03:20:17,661 --> 03:20:21,661
and you can find a link to the Telegram group at layer2labs.com.

1828
03:20:21,661 --> 03:20:28,661
There are extensive essays and analyses at drivechain.info, if you're interested.

1829
03:20:28,661 --> 03:20:39,661
You can read Paul's blog at truthcoin.info, where he's written for many years about Bitcoin in extensive advanced essays.

1830
03:20:39,661 --> 03:20:42,661
So, yeah, see you all next week for our next conversation here.

1831
03:20:42,661 --> 03:20:47,661
Perhaps next week will be an open mic where anyone can come up instead of an invited guest,

1832
03:20:47,661 --> 03:20:50,661
although we could have an invited guest, of course.

1833
03:20:50,661 --> 03:20:54,661
Thanks again, Adam Back, for being here, the CEO and co-founder of Blockstream.

1834
03:20:54,661 --> 03:20:56,661
It was really great to have you here today.

1835
03:20:56,661 --> 03:20:59,661
Thank you again.

1836
03:20:59,661 --> 03:21:00,661
Yeah, thank you.

1837
03:21:00,661 --> 03:21:02,661
I'll see you guys next time, too.

1838
03:21:02,661 --> 03:21:03,661
Peace.

1839
03:21:03,661 --> 03:21:05,661
Thanks, everyone.