1
00:00:00,000 --> 00:00:13,640
So, for those of you who watched my TabConf talk a few years ago, it was very
light-hearted.

2
00:00:13,640 --> 00:00:17,880
I'm going to try to keep this as light-hearted, and it should get very fun.

3
00:00:17,880 --> 00:00:23,680
I'm going to try to talk to you for as little time as I can, and then we'll make
this as

4
00:00:23,680 --> 00:00:29,080
interactive as possible, do Q&A, but I guess it'll be a debate format or something,
this

5
00:00:29,080 --> 00:00:30,080
is what I've been told.

6
00:00:30,080 --> 00:00:35,000
So, I'm going to try to keep this light, hopefully it stays light, I don't know how
well I'll

7
00:00:35,000 --> 00:00:38,240
be able to manage because this topic is a little more dense, but we're going to do
our

8
00:00:38,240 --> 00:00:39,240
best.

9
00:00:39,240 --> 00:00:45,560
Yeah, so, Mev and Mewvel, and apparently some cats.

10
00:00:45,560 --> 00:00:52,160
So, probably useful to do a little bit of background on Mev.

11
00:00:52,160 --> 00:00:57,360
Mev is a concept that comes to us from Ethereum, boo, I know, how terrible, but
they've been

12
00:00:57,360 --> 00:01:01,200
struggling with this for a long time, and we should probably learn something from
them

13
00:01:01,200 --> 00:01:04,120
because they've spent a lot of time on it.

14
00:01:04,120 --> 00:01:05,120
So what is Mev?

15
00:01:05,120 --> 00:01:09,600
Mev is literally just all of the money that the miners make, every single penny.

16
00:01:09,600 --> 00:01:14,160
It is a super broad concept, it includes all the transaction fees, it includes all
the

17
00:01:14,160 --> 00:01:19,100
money they can make from out-of-band fees, so things like Mempool Accelerator, it
includes

18
00:01:19,100 --> 00:01:23,000
money that they can make doing super advanced attacks, which we'll get into in a
minute,

19
00:01:23,000 --> 00:01:26,960
it includes money that they can make by reorging the blockchain and stealing
people's money.

20
00:01:26,960 --> 00:01:31,720
It's really, really broad, and in fact, so broad that it's basically a useless term,

21
00:01:31,720 --> 00:01:34,400
so we should try to avoid using it.

22
00:01:34,400 --> 00:01:40,320
Instead, I tried to coin this term called Mevel, that is how you pronounce it,
Mevel,

23
00:01:40,320 --> 00:01:46,200
not Mevel, I don't know why everyone decided it's Mevel, it's Mevel, it's evil, and
it's

24
00:01:46,200 --> 00:01:49,320
just Mev that's evil, so decentralizing Mev.

25
00:01:49,320 --> 00:01:53,040
So the problem with Mev, from the Bitcoin perspective and kind of from the network's

26
00:01:53,040 --> 00:01:57,840
perspective, not necessarily from the user's perspective, is that Mev can create
strong

27
00:01:57,840 --> 00:01:59,000
centralization.

28
00:01:59,000 --> 00:02:04,200
So Mev in Ethereum is actually super complicated to capture, like to optimize the
amount of

29
00:02:04,200 --> 00:02:09,700
Mev you get, the amount of money you get, the amount of money you get for your
blocks,

30
00:02:09,700 --> 00:02:14,680
you have to hire the best engineers, so you can't just hire, you know, fine
engineers

31
00:02:14,680 --> 00:02:20,080
and pay them fine salaries, no, you have to hire the best engineers and attract the
absolute

32
00:02:20,080 --> 00:02:25,400
best talent, and spend millions of dollars a year, or tens of millions of dollars a
year,

33
00:02:25,400 --> 00:02:28,600
no matter how big a miner or a staker you are, even if you're a very small one, in
order

34
00:02:28,600 --> 00:02:34,540
to get the maximal revenue, you have to invest the same amount of money as the big
players

35
00:02:34,540 --> 00:02:38,480
in order to build the best block template.

36
00:02:38,480 --> 00:02:43,560
So this is bad for centralization, all of a sudden your costs as a miner don't scale

37
00:02:43,560 --> 00:02:47,320
with the size of your mining operation, so if you're a small miner, or a small
staker

38
00:02:47,320 --> 00:02:51,600
in Ethereum, you have to spend the same amount of money as the big guys, and that's
really

39
00:02:51,600 --> 00:02:56,800
bad, it means the big guys have a big advantage, and creates strong centralization
pressure.

40
00:02:56,800 --> 00:03:01,640
So in Bitcoin, we would call this evil, or Mevil.

41
00:03:01,640 --> 00:03:06,000
And I know, you know, hiring top-notch engineers looks easy, because all the Bitcoin
engineers

42
00:03:06,000 --> 00:03:10,980
are these great guys, and like, they're all easily recruited, but I promise you it's
actually

43
00:03:10,980 --> 00:03:15,160
not easy, and it's also because Bitcoin is fascinating, and working for a for-profit

44
00:03:15,480 --> 00:03:19,120
company attacking Bitcoin is like, less exciting, and so it's actually really hard
to hire those

45
00:03:19,120 --> 00:03:20,120
engineers.

46
00:03:20,120 --> 00:03:23,040
Okay, so what does this actually look like in practice?

47
00:03:23,040 --> 00:03:27,400
This very abstract, high-level nonsense, like alright, there's like this
centralizing pressure

48
00:03:27,400 --> 00:03:29,680
nonsense, like what does this actually mean?

49
00:03:29,680 --> 00:03:34,720
In practice, not a majority, but a plurality of Mev and Ethereum is just what's
called

50
00:03:34,720 --> 00:03:36,500
sex-dex arbitrage.

51
00:03:36,500 --> 00:03:40,920
So you have these centralized exchanges, Coinbase, Binance, whatever, you have
decentralized

52
00:03:40,920 --> 00:03:45,320
exchanges, Uniswap, there's a handful of them, doesn't really matter what they are,

53
00:03:45,320 --> 00:03:49,960
and you're arbitraging, just straight-up arbitrage, classic price arbitrage, you're
building a

54
00:03:49,960 --> 00:03:55,320
block template, you want to take the price on the dex and trade it against the price

55
00:03:55,320 --> 00:04:00,960
on the sex, and just, you make money on the margin.

56
00:04:00,960 --> 00:04:05,880
Super easy to understand, not super complicated, actually really hard to do right,
and to do

57
00:04:05,880 --> 00:04:10,200
the best right, not just right, but the best, because you actually have to be a
high-frequency

58
00:04:10,200 --> 00:04:11,200
trading firm.

59
00:04:11,200 --> 00:04:14,800
Because in Ethereum, you have fixed time windows for your blocks, you know exactly
when the

60
00:04:14,800 --> 00:04:19,760
time window for the block is up, down to the millisecond, so if you put your
block-producing

61
00:04:19,760 --> 00:04:24,480
machine in the right data center, you're going to have a slight marginal advantage,
and it

62
00:04:24,480 --> 00:04:28,680
has literally turned into, you have to be a high-frequency trader in order to make
blocks

63
00:04:28,680 --> 00:04:29,680
in Ethereum.

64
00:04:29,680 --> 00:04:34,400
Great, that's not how we want it to be, but hey, that's how it is.

65
00:04:34,400 --> 00:04:38,920
It's again maybe useful to point out that Mev here, again, is a useless term, Mev in

66
00:04:38,920 --> 00:04:44,120
Ethereum, we use to refer, they use to refer to everyone getting harmed, so if a
user gets

67
00:04:44,120 --> 00:04:52,840
harmed and they have price slippage, because this Mev extraction process thing
happens

68
00:04:52,840 --> 00:04:57,560
and they get a worse price, this is a part of Mev, but it doesn't impact the system,

69
00:04:57,560 --> 00:04:58,560
right?

70
00:04:58,560 --> 00:05:02,600
So when we're talking about centralization, where we're talking about Meval, we're
saying

71
00:05:02,600 --> 00:05:05,680
the thing that matters is the centralization pressure from that.

72
00:05:05,680 --> 00:05:09,400
If you have to be a high-frequency trader in order to be a miner, this is bad for
mining

73
00:05:09,400 --> 00:05:14,320
and Bitcoin the system, but we don't necessarily care about like, oh, user's got a
worse price

74
00:05:14,320 --> 00:05:17,680
execution on this decentralized exchange, that's not our problem.

75
00:05:17,680 --> 00:05:20,280
That's the decentralized exchange's problem, that's the user's problem, that's not
our

76
00:05:20,280 --> 00:05:21,280
problem.

77
00:05:21,280 --> 00:05:28,400
Okay, so that's the easy-to-understand Mev, that's the classic example, it's
mid-double-digit

78
00:05:28,400 --> 00:05:33,040
percentage of the total Mev extraction in Ethereum is just sex-to-sex arbitrage, and

79
00:05:33,040 --> 00:05:39,040
specifically sex-to-sex arbitrage in the Ethereum-USDC pair or Ethereum-USDT pair.

80
00:05:39,040 --> 00:05:43,040
But there's also a ton of other things, right, that's less than 50% or maybe right
around

81
00:05:43,040 --> 00:05:44,040
50%.

82
00:05:44,040 --> 00:05:47,520
There's all kinds of fancy things you can do, you have all these crazy smart
contracts,

83
00:05:47,520 --> 00:05:50,720
you can do all kinds of attacks, right, if you can exploit a smart contract and
steal

84
00:05:50,720 --> 00:05:55,080
a bunch of money from it, maybe you use a flash loan, so you use a loan in Ethereum

85
00:05:55,080 --> 00:06:01,240
that gives you infinite money, but then you have to repay that money within the same
block,

86
00:06:01,240 --> 00:06:04,000
and you can use that infinite money to make the price move over here, and then you
can

87
00:06:04,000 --> 00:06:08,160
go steal some money over here, and then, yeah, they do all kinds of crazy crap.

88
00:06:08,160 --> 00:06:11,560
So that's all Mev, right, that's all stuff, if you're a miner and you want to get
the

89
00:06:11,560 --> 00:06:16,840
most money, you have to come up with all this crazy crap, and you have to be the
best engineers

90
00:06:16,840 --> 00:06:21,640
coming up with the craziest way to exploit all the stupid broken crap in order to be

91
00:06:21,640 --> 00:06:24,280
competitive.

92
00:06:24,280 --> 00:06:28,160
And without going to jail, which has become a problem recently, but they didn't
think

93
00:06:28,160 --> 00:06:29,160
it was going to be a problem.

94
00:06:29,160 --> 00:06:31,160
Weird and funny how that works.

95
00:06:31,160 --> 00:06:37,760
Okay, so whatever, that's what Mev is, that's what Mevil is, and that's the problems
they

96
00:06:37,760 --> 00:06:38,760
have with it.

97
00:06:38,760 --> 00:06:40,080
What have they done about this, right?

98
00:06:40,080 --> 00:06:44,120
So the Ethereum people aren't dumb, I know people like to think that they have no
brain

99
00:06:44,120 --> 00:06:47,360
power over there, but actually they have a lot of smart people over there, they're
trying

100
00:06:47,360 --> 00:06:48,360
to solve this stuff.

101
00:06:48,360 --> 00:06:49,360
What have they come up with?

102
00:06:49,360 --> 00:06:52,200
What have they actually built?

103
00:06:52,200 --> 00:06:56,440
It's not great.

104
00:06:56,440 --> 00:07:01,880
So they recognize this problem, a long time ago in fact, that this centralized block
building,

105
00:07:01,880 --> 00:07:07,560
like just running a node and building an efficient block template is not an easy
problem, and

106
00:07:07,560 --> 00:07:08,560
it doesn't scale.

107
00:07:08,560 --> 00:07:11,160
It doesn't scale down to small miners, small stakers.

108
00:07:11,160 --> 00:07:12,160
So what do we do?

109
00:07:12,160 --> 00:07:13,800
They disaggregated that market, right?

110
00:07:13,800 --> 00:07:17,040
So they said, okay, you can be a staker, but then someone else is going to actually
select

111
00:07:17,040 --> 00:07:20,040
the transactions that go in your block, and we're going to build this marketplace
for

112
00:07:20,040 --> 00:07:21,360
block templates.

113
00:07:21,360 --> 00:07:24,960
Sadly, in Ethereum, the marketplace is for full block templates, right?

114
00:07:24,960 --> 00:07:29,160
So it's not just like, here's some transactions I want you to include in your block,
no, it's

115
00:07:29,160 --> 00:07:33,120
here's an entire block, here's all the transactions that are in it, and all the
transactions that

116
00:07:33,120 --> 00:07:36,440
aren't in it, and I'm going to give you those transactions, and you must include
exactly

117
00:07:36,440 --> 00:07:40,200
those transactions and nothing more, and that's your block.

118
00:07:40,200 --> 00:07:44,280
This is kind of in part because Ethereum has this global state, right?

119
00:07:44,280 --> 00:07:48,080
They don't have UTXOs, they have addresses, and any smart contract can impact the
global

120
00:07:48,080 --> 00:07:52,440
state, and this can cause any other random smart contract to fail, and you can't do
analysis

121
00:07:52,440 --> 00:07:53,440
of this, right?

122
00:07:53,440 --> 00:07:56,880
This is just a global state, and you might run one transaction, you know, include
one

123
00:07:56,880 --> 00:08:00,480
transaction, and it might cause some other transaction to fail, and you can't, you
can't

124
00:08:00,480 --> 00:08:04,520
in a kind of theoretical computer science way, determine this ad hoc.

125
00:08:04,520 --> 00:08:08,640
You have to actually simulate it, you have to do a ton of actually advanced computer

126
00:08:08,640 --> 00:08:11,640
science to figure out which transactions are going to interact with each other, and
how

127
00:08:11,640 --> 00:08:15,240
are they going to interact, and what is this going to cause my ability to include
different

128
00:08:15,240 --> 00:08:18,440
transactions.

129
00:08:18,440 --> 00:08:23,160
So okay, they have this whole thing built around full templates, that kind of sucks,

130
00:08:23,160 --> 00:08:27,320
but these templates are worth a lot more than naive block building, and in fact,
more

131
00:08:27,320 --> 00:08:32,080
so every day, because this system exists, and because this has become a part of
Ethereum,

132
00:08:32,080 --> 00:08:36,400
it has now gotten so bad that many systems in Ethereum that are being built today
don't

133
00:08:36,400 --> 00:08:39,720
even use the public mempool, there's barely even a public mempool anymore, it just
doesn't

134
00:08:39,720 --> 00:08:43,440
exist, there's just these private mempools, you relay your transaction directly to
one

135
00:08:43,440 --> 00:08:47,280
of these people participating in this marketplace, one of these literal high
frequency traders,

136
00:08:47,280 --> 00:08:51,640
they include your transaction for you, it's great, it's super permissioned, and you
just,

137
00:08:51,640 --> 00:08:56,080
you know, contact your friendly high frequency trader, and you send them all your
transactions,

138
00:08:56,080 --> 00:09:01,360
and they include it in your blocks, and it's great, yeah, you don't need any
decentralization

139
00:09:01,360 --> 00:09:02,920
here.

140
00:09:02,920 --> 00:09:09,240
So anyway, sadly, so that's not the worst of it, right, there's in fact only a few
of

141
00:09:09,240 --> 00:09:12,760
these marketplaces on top, right, so there's not just a few of these entities that
can

142
00:09:12,760 --> 00:09:15,960
include the transactions, but there's a few of these marketplaces, there's only like
two,

143
00:09:15,960 --> 00:09:23,280
three I think of them, and they do affect censorship on top, right, so in order to

144
00:09:23,280 --> 00:09:26,240
get a good return, you know, if you're a staker and you want to get a good return on
your

145
00:09:26,240 --> 00:09:31,760
block template, you have to participate in one of these three, four marketplaces,
and

146
00:09:31,760 --> 00:09:34,640
those marketplaces are censored, and by the way, if those marketplaces decide they
don't

147
00:09:34,640 --> 00:09:37,960
like you, suddenly you can't build good block templates, and you're not going to
make very

148
00:09:37,960 --> 00:09:42,560
much money being a staker, how terrible.

149
00:09:42,560 --> 00:09:47,000
And then yeah, I also talked about, it's not just these marketplaces that are
centralized,

150
00:09:47,000 --> 00:09:50,760
it's just the players, you know, all this theoretical MEV stuff where it's going to

151
00:09:50,760 --> 00:09:55,280
become centralizing, no, it actually is very centralizing, there are only like a
small

152
00:09:55,280 --> 00:10:00,280
handful, you know, less than five firms that can competitively build a block
template in

153
00:10:00,280 --> 00:10:04,840
Ethereum, right, if you want to get the most money for your block in Ethereum, you
have

154
00:10:04,840 --> 00:10:10,080
to buy a block template, or you have to sell your block template to one of these
three

155
00:10:10,080 --> 00:10:13,640
or four companies, and that's it, anybody else trying to do it, their block
templates

156
00:10:13,640 --> 00:10:19,160
are going to be worth a little less, not always a lot less, but a little less, so
this job

157
00:10:19,160 --> 00:10:24,280
of hiring the best, and not just good, but the best engineers, turns out to be very
centralizing,

158
00:10:24,280 --> 00:10:28,080
and we can see this in practice, and a few people can actually do it.

159
00:10:28,080 --> 00:10:33,680
Okay, so that's Ethereum, Ethereum is terrible, we all hate Ethereum, great, they
screwed

160
00:10:33,680 --> 00:10:37,720
everything up, it's terrible, and what does this actually mean for us, you know, who
cares

161
00:10:37,720 --> 00:10:49,760
about Ethereum? Okay, so it turns out it's actually worse for us, okay, I'll get to
that,

162
00:10:49,760 --> 00:10:55,060
but first I just want to highlight, you know, centralization is why we're here, I
hope everyone

163
00:10:55,060 --> 00:10:59,960
kind of understands that intuitively, right, that if we end up in that same spot
where

164
00:10:59,960 --> 00:11:04,280
there's three or four companies that select all the transactions that go in the
chain,

165
00:11:04,280 --> 00:11:08,200
and three or four companies on top of that, that select all the people who are
allowed

166
00:11:08,200 --> 00:11:13,120
to mine competitively, then like, this is kind of useless, we should just kind of
like,

167
00:11:13,120 --> 00:11:17,400
why are we using Bitcoin, we should probably use Fedcoin, because like, if the US
government

168
00:11:17,400 --> 00:11:21,520
can decide who gets to put their transactions on the blockchain, that's, you know,
what

169
00:11:21,520 --> 00:11:27,920
was the point? So just, hopefully that's kind of intuitive to everyone, but okay,
yeah,

170
00:11:27,920 --> 00:11:32,840
so it's actually worse for us, so in Ethereum, they have this whole staking thing,
right,

171
00:11:32,840 --> 00:11:39,200
so if you're a staker, you put some Ethereum up, you have no operational costs, no
OPEX,

172
00:11:39,200 --> 00:11:45,160
which has other problems, but you have no OPEX as a staker, right, so if you are a
staker

173
00:11:45,160 --> 00:11:50,140
and you mine these like, less than super profitable block templates, they're like,
okay, mediocre

174
00:11:50,140 --> 00:11:53,280
block templates, maybe you don't make very much money from them, but whatever,
that's

175
00:11:53,280 --> 00:11:56,960
okay, like, you're just getting less money, you know, you're getting a lower
interest

176
00:11:56,960 --> 00:12:01,080
rate on your Ethereum, but like, you're not really losing money. In proof of work,
no,

177
00:12:01,080 --> 00:12:04,560
you're screwed, you're going out of business, right, because in proof of work, you
have

178
00:12:04,560 --> 00:12:08,680
the same OPEX as everybody else, roughly, you know, you've got to pay your power
bill

179
00:12:08,680 --> 00:12:14,080
at the end of the month. If you are making 10, 20, 30, 40% less than your
competition,

180
00:12:14,080 --> 00:12:18,320
you're going out of business. There's no, there's no like, oh, you can keep going,
and

181
00:12:18,320 --> 00:12:21,840
like, some people can still keep going, like in Ethereum, we see this as a long tail
of

182
00:12:21,840 --> 00:12:26,880
small stakers that do build their own block templates. You just can't have that in a
proof

183
00:12:26,880 --> 00:12:33,200
of work system, it won't work. Okay, so that's terrible, but maybe, maybe, maybe we
can solve

184
00:12:33,200 --> 00:12:36,520
this another way, right, so Ethereum has this like, this global state I talked
about, it's

185
00:12:36,520 --> 00:12:40,920
terrible, you know, they have to sell whole block templates, but Bitcoin's better,
right,

186
00:12:40,920 --> 00:12:45,200
we have this UTXO system, we have, we can look at a transaction, we can say, okay,
it

187
00:12:45,200 --> 00:12:49,680
spends these UTXOs, and this one is unrelated, and they don't interact, and we can
build

188
00:12:49,680 --> 00:12:54,320
a different market. So I think a useful way to conceptualize some of this MEV stuff
is

189
00:12:54,320 --> 00:12:59,560
it's really the marketplace for block space, right, so it is, how are we structuring
the

190
00:12:59,560 --> 00:13:03,920
marketplace for block space, and Ethereum has these decentralized marketplaces that

191
00:13:03,920 --> 00:13:08,120
only sell whole block templates, Bitcoin still has this public mempool thing,
there's some

192
00:13:08,120 --> 00:13:13,000
private mempools starting to crop up, they're not, they don't really have enough
money in

193
00:13:13,000 --> 00:13:17,000
them, like there's not enough demand for them to be super relevant, but maybe they
will

194
00:13:17,000 --> 00:13:21,240
be in the future, maybe some of this MEV stuff will come to pass, but what is it,
what should

195
00:13:21,240 --> 00:13:27,040
this marketplace look like? And so you can, in Bitcoin, because we don't have this
global

196
00:13:27,040 --> 00:13:31,600
state, you can really still visualize it as individual order books per UTXO, right,
so

197
00:13:31,600 --> 00:13:36,360
I have a transaction that spends this UTXO, and I'm going to bid a certain amount
for

198
00:13:36,360 --> 00:13:39,600
that, and we could actually have different order books, we could say I'm going to
outbid

199
00:13:39,600 --> 00:13:44,360
you to spend that UTXO, but, you know, this other transaction spending some
unrelated

200
00:13:44,360 --> 00:13:48,920
UTXOs, bid separately, right, so we could have different order books, so we could
actually,

201
00:13:48,920 --> 00:13:52,600
we could build something that's much better, we could say, okay, the miners are
going to

202
00:13:52,600 --> 00:13:58,200
actually clear this order book, the miners are going to, they're going to still
include

203
00:13:58,200 --> 00:14:01,280
their own transactions, both from the public mempools and some private mempools and
some

204
00:14:01,280 --> 00:14:06,440
whatever else, and we'll still have some decentralization, we'll still have
censorship resistance, right,

205
00:14:06,440 --> 00:14:11,240
because they can pull from the public mempool, and we can have these private
mempools that

206
00:14:11,240 --> 00:14:18,520
have different properties. So that's great, you know, if Bitcoin is all about this
per

207
00:14:18,520 --> 00:14:24,880
UTXO model, and all the state is encapsulated in this UTXO model, then we can have a
better

208
00:14:24,880 --> 00:14:28,280
marketplace than Ethereum. We can still have this crazy MEV marketplace thing, but
we can

209
00:14:28,280 --> 00:14:34,200
have a better one that's not, doesn't introduce the censorship failures that we see
on Ethereum.

210
00:14:34,200 --> 00:14:39,400
We still have some centralization concerns, but maybe, maybe we can live with that,
I

211
00:14:39,400 --> 00:14:46,440
don't know. Okay, well, sadly, yeah, client-side validation kind of fucks us, and
no, I don't

212
00:14:46,440 --> 00:14:50,960
mean the Jonas, Nick, and Robin Linus thing they were just talking about. I mean,
okay,

213
00:14:50,960 --> 00:14:56,760
I do mean that, I mean exactly that, but not their version. So you can imagine a
client-side

214
00:14:56,760 --> 00:15:01,680
validation system where, or roll-up system or whatever, where you have all of the
same

215
00:15:01,680 --> 00:15:06,680
global state, like you copy and paste the EVM, and all of a sudden we have global
state

216
00:15:06,680 --> 00:15:11,440
on the Bitcoin blockchain, and just by, you know, using data, we're going to like
parse

217
00:15:11,440 --> 00:15:14,240
the data out of the Bitcoin blockchain, and we're going to have the EVM, and now
we're

218
00:15:14,240 --> 00:15:19,280
going to have global state, and now we're fucked. Obviously, like the Jonas, Nick,
and

219
00:15:19,280 --> 00:15:24,560
the Robin Linus, the client-side shielded CSV thing, that doesn't have this problem.

220
00:15:24,560 --> 00:15:28,000
It's not designed for expressivity, right? We don't have this case where we're
trying

221
00:15:28,000 --> 00:15:33,560
to build, you know, DEXs and all kinds of crazy stuff that might introduce MEV, so
it's

222
00:15:33,560 --> 00:15:38,680
just a transaction layer, it's probably not going to introduce this crazy, crazy MEV
crap.

223
00:15:38,680 --> 00:15:43,640
But that same concept, right, if we expand it, then suddenly we have literally EVM
just

224
00:15:43,640 --> 00:15:48,520
on the Bitcoin blockchain, and now we have the same problems. Maybe there's a world
where

225
00:15:48,520 --> 00:15:53,480
we can have this marketplace, you know, I talked about this is a marketplace, right?

226
00:15:53,480 --> 00:15:57,240
We can structure the order book for our marketplace however we want. Maybe we can
come up with

227
00:15:57,240 --> 00:16:02,680
a way to structure the orders so that you don't have quite this problem. Maybe we
can

228
00:16:02,680 --> 00:16:06,760
come up with a way to say, like, okay, you know, the order book is actually aware of

229
00:16:06,760 --> 00:16:13,360
this underlying CSV system that exists. Maybe we can do that, I don't know. This is
all

230
00:16:13,360 --> 00:16:19,120
very speculative, obviously. We really don't know where any of this stuff is going.
And

231
00:16:19,120 --> 00:16:23,440
it's important to, you know, I think people will always point out, and it is
important

232
00:16:23,440 --> 00:16:28,080
to highlight that MEV will always exist, like some amount of it. The question is
just how

233
00:16:28,080 --> 00:16:33,040
much of it, how much is a problem, right? We have it today, you know, people do,
mempool

234
00:16:33,040 --> 00:16:38,760
accelerator exists, almost no one uses it, but the mempool accelerator exists. We
have

235
00:16:38,920 --> 00:16:44,360
slipstream, raise your hand in the back. You know, we've got slipstream, you know,
it exists.

236
00:16:45,080 --> 00:16:49,080
Again, it's not very, it's not a big marketplace. Like, there's just not a lot of
extra revenue

237
00:16:49,080 --> 00:16:55,560
from that. So much, sufficiently little so that, you know, miners care about it,
like

238
00:16:55,560 --> 00:16:58,760
they kind of want it, but it's not like, oh, yeah, we got to have that or we're
going to

239
00:16:58,760 --> 00:17:01,800
go out of business next month because we're not going to make our power bill because
our

240
00:17:01,800 --> 00:17:05,560
competitors are making more money than us. We're not even close to that point,
right?

241
00:17:05,560 --> 00:17:10,600
So there is a continuum here between, like, something that exists and, like, isn't
really

242
00:17:10,600 --> 00:17:14,440
relevant to all of a sudden everything is super centralized and we just, like, start

243
00:17:14,440 --> 00:17:17,480
throwing away the public mempool because, like, it's not really worth having
anymore,

244
00:17:17,480 --> 00:17:22,680
it's complicated, and fuck it, which is where Ethereum has ended up. So there's some
continuum,

245
00:17:22,680 --> 00:17:27,160
and I don't have an answer for, like, where we are or where we should be, but it's
important

246
00:17:27,160 --> 00:17:32,280
to recognize that there's a continuum here. All right, last question, and then we'll
get

247
00:17:32,280 --> 00:17:38,280
into the whole debate and Q&A thing, or last slide. So what does this actually mean
for Bitcoin,

248
00:17:38,280 --> 00:17:42,040
especially as we look towards the future? You know, we talked about the concerns,

249
00:17:42,040 --> 00:17:47,720
but what does this mean? So, you know, expressivity always runs some risk,

250
00:17:48,440 --> 00:17:54,200
but it carries some value, right? If we add super deep covenants, people might build
these crazy

251
00:17:54,200 --> 00:17:58,840
DEXs, which might introduce, you know, sex arbitrage, it might introduce, you know,

252
00:17:58,840 --> 00:18:05,640
people might start building these crazy flashloan stuff and all kinds of crazy smart
contracts that

253
00:18:05,640 --> 00:18:15,800
get exploited. Like, that might happen, and those are useful, but it might also
cause evil, right?

254
00:18:15,800 --> 00:18:21,400
So there's a trade-off here, like, it might have a lot of value or it might end up
screwing us.

255
00:18:22,040 --> 00:18:29,640
So evaluating these proposals thus becomes very quickly a second order effect thing,
right?

256
00:18:29,640 --> 00:18:33,800
What kinds of systems will people actually build? Because most of the things people
are talking

257
00:18:33,800 --> 00:18:37,800
about building actually don't have MEVL problems. They have MEV problems, right?
People are talking

258
00:18:37,800 --> 00:18:44,280
about building roll-ups, where you have all of the MEV goes to a centralized
counterparty who does

259
00:18:44,280 --> 00:18:49,640
the sequencing. Well, that's not a MEVL problem. There's no concerns for
centralizing mining.

260
00:18:49,640 --> 00:18:54,040
You already have a centralized party. They're taking all of the miner extractable
value of that

261
00:18:54,040 --> 00:18:59,720
roll-up, and they're keeping it for themselves, and that's great. For Bitcoin, it's,
you know,

262
00:19:00,280 --> 00:19:03,720
fine for the users of that system, I guess. It has some other great properties. So
people are

263
00:19:03,720 --> 00:19:08,120
talking about building those things. So then you might say, well, okay, you know,
people are just

264
00:19:08,120 --> 00:19:11,640
going to build that, then none of this is going to come to pass, and then we don't
have to worry

265
00:19:11,640 --> 00:19:16,200
about it. But of course, it is also a big risk, right? If we end up there, if we end
up where

266
00:19:16,200 --> 00:19:23,000
Ethereum is, kind of what was the point? Why are we doing this Bitcoin thing? So we
end up very

267
00:19:23,000 --> 00:19:27,320
quickly getting into this, like, evaluating what people are actually going to build,
not just what

268
00:19:27,320 --> 00:19:30,360
they're talking about, but what people are going to build in 10 years, and how does
this impact

269
00:19:30,360 --> 00:19:36,040
Bitcoin, and this gets real fuzzy, and real terrible, and disgusting, and does not
turn into

270
00:19:36,040 --> 00:19:42,920
a fun conversation, but hey. Also, you know, I talked about fancy MEV marketplaces.
It's probably

271
00:19:42,920 --> 00:19:45,960
something we should be investigating more, probably something we should be looking
into,

272
00:19:45,960 --> 00:19:51,400
something we should probably start building. You know, these don't exist at all.
There's a few

273
00:19:51,400 --> 00:19:57,560
people trying to copy and paste Ethereum MEV extraction marketplaces onto Bitcoin,
which would

274
00:19:57,560 --> 00:20:02,760
be horrendous, for all the reasons I talked about. Like, we don't want that. So
probably it's the

275
00:20:02,760 --> 00:20:08,280
case that we should start thinking about how to build better MEV extraction
marketplaces,

276
00:20:08,840 --> 00:20:14,840
and better private mempools, and better order book for the more flexible and
expressive order

277
00:20:14,840 --> 00:20:20,840
book for the Bitcoin block space. I have some ideas. I don't like the ideas, but it
might be

278
00:20:20,840 --> 00:20:25,560
the best we can do, and it's probably something that should be researched more in
the community.

279
00:20:26,440 --> 00:20:33,080
Okay, so that's all the discussions and background I wanted to give. Now we're going
to do, I don't

280
00:20:33,080 --> 00:20:40,120
know, half hour Q&A? I don't know how long we have, but a while. We have a while,
and yeah,

281
00:20:40,120 --> 00:20:46,760
and I think Tidwell wants to just go right into debate. Yeah, so first, can I just
get a quick

282
00:20:46,760 --> 00:20:57,480
round of applause for Matt here? So this is the second half, or the next, you know,
two-thirds of

283
00:20:57,480 --> 00:21:04,040
the conversation, and let me give you a little bit of context here. Literally
everything that I

284
00:21:04,040 --> 00:21:09,480
learned about in Bitcoin, all these changes, roughly half of them is like, oh, we
can't do

285
00:21:09,480 --> 00:21:15,720
that because MEV, or we can't do this because of MEV, or even like, we're talking
about symphony

286
00:21:15,720 --> 00:21:21,320
and simplicity earlier, like, oh, this may introduce MEV, and I was really, I'm
really excited about

287
00:21:21,320 --> 00:21:25,560
Drivechains, for instance, but every time I speak about it publicly, everyone's
like, wait, MEV.

288
00:21:25,560 --> 00:21:31,160
I'm like, well, shit, okay, this MEV thing sounds very scary. Who would be really
good at it giving

289
00:21:31,160 --> 00:21:37,000
a really good primer on it? I trust Matt to do a good job explaining that, giving us
a primer to

290
00:21:37,000 --> 00:21:44,040
work off of, but I don't like to just trust, I like to verify. So what I wanted to
do here is,

291
00:21:44,040 --> 00:21:49,160
if anyone wanted to have a semi-long-form discussion with Matt, it doesn't
necessarily

292
00:21:49,160 --> 00:21:54,360
have to be a debate, it could be Q&A as well, I would invite you to actually walk
forward and sit

293
00:21:54,360 --> 00:21:59,320
in this front row so you can come on stage, and if that makes you uncomfortable, you
can also take

294
00:21:59,320 --> 00:22:03,320
questions from the audience, but I would ask, if you do want to ask questions from
the audience,

295
00:22:03,320 --> 00:22:08,200
keep it to only one question and no follow-up questions. I really want this to be
interactive

296
00:22:08,200 --> 00:22:14,280
because I'm really tired of people talking about MEV, disagreeing, and silence. I
want this to be

297
00:22:14,280 --> 00:22:19,960
up here, and I want us to get the conversations and discussions out in the open. If
that makes

298
00:22:19,960 --> 00:22:26,280
you feel comfortable, I invite you on stage, and Matt is very reasonable, we can
agree to disagree

299
00:22:26,280 --> 00:22:31,240
at the end of the day, but I want us to at least have the conversation. So I
appreciate if you want

300
00:22:31,240 --> 00:22:37,960
to be involved, and we'll take, we have chairs up here for people to sit, and Matt,
this is the next

301
00:22:37,960 --> 00:22:44,360
half, so if we already have a few people, you can come up on stage, these three,
yeah, you can come

302
00:22:44,360 --> 00:22:49,240
up on stage, but I will ask you, after you've made some points and discussions,
we'll rotate people.

303
00:22:50,360 --> 00:22:55,480
Is that fair? Yeah, let's do it. And I don't think I need to moderate here, but

304
00:22:55,480 --> 00:22:59,880
there are mics, I might moderate if need be, so mics next to your chair should work,
you have to turn it on.

305
00:23:03,000 --> 00:23:12,760
Testing, testing. Hello, hello, is this on? Okay, test, test, test. Okay, all right,
do you want to just go down the line?

306
00:23:14,520 --> 00:23:18,040
Go ahead. Yeah, no need for introductions, let's just keep it on topic here.

307
00:23:18,200 --> 00:23:24,600
What about the fact that currently in Bitcoin, there are transactions that can be
broadcast on

308
00:23:24,600 --> 00:23:32,040
the broader network that can create problems, such as DOS blocks? I think that
miners being

309
00:23:32,040 --> 00:23:36,440
more centralized in that aspect actually protects the network somewhat, because
their incentives

310
00:23:36,440 --> 00:23:42,200
align better for the success of Bitcoin than, for example, a highly decentralized
network where you

311
00:23:42,200 --> 00:23:48,200
could have three to five percent of the network causing substantial issues until
these consensus

312
00:23:48,200 --> 00:23:52,920
problems have been resolved. Yeah, I mean, I think that's this total issue. I mean,
that's the reality

313
00:23:52,920 --> 00:23:57,160
we have today. I hope that's not the reality we end up with. I hope that's not the
end state, right?

314
00:23:57,160 --> 00:24:02,200
Absolutely. That's a problem, that's the reality we have today, but also

315
00:24:02,920 --> 00:24:08,120
what we should be working on is fixing that and making it so that you can't cause
those problems,

316
00:24:08,600 --> 00:24:13,960
an individual miner can't cause those problems, an individual transactor can't cause
those

317
00:24:13,960 --> 00:24:20,040
problems, and then hopefully we can not worry about that in a decade or whenever we
get there.

318
00:24:20,920 --> 00:24:26,600
Thank you. It's hard to know where to begin, because it seems like the entire thing
is

319
00:24:26,600 --> 00:24:33,480
misconceived, but I think that, first of all, mining obviously is something where we
have

320
00:24:33,480 --> 00:24:37,400
evolution across time, because the difficulty adjustment is firing the
underperformers,

321
00:24:37,400 --> 00:24:41,320
so it's not like everyone should be able to mine. I mean, I hope you would agree
with that,

322
00:24:41,320 --> 00:24:46,360
because the whole point of mining is that it's only the people who have done the
proof of work

323
00:24:47,400 --> 00:24:53,560
who get to find blocks, and so the idea that just anyone can find blocks is the
exact opposite of

324
00:24:53,560 --> 00:24:57,507
proof of work in my opinion, so I guess that's like my first question, if you agree
with that.

325
00:24:57,507 --> 00:25:10,507
Yeah, so obviously there's... the structure of the proof of work market, obviously
you're totally right, like, if you have less efficient hardware and a higher power
cost, then you get fired, that's how the network works.

326
00:25:10,507 --> 00:25:26,507
But the structure of that is, as much as possible, in order for decentralization to
exist, as much as possible, it needs to be the case that your CapEx and OpEx, in
fact, scale linearly with your size.

327
00:25:26,507 --> 00:25:52,507
Well, I don't know about scale linearly with size, but I think that at least this is
the relevant dimension, at least. But now it seems to me that MEV casts a
distinction between hardware and software, that it seems to be arbitrary, first of
all, and then the other thing is the software part is very easy to solve, in my
opinion, so I have good news for everyone who's worried about MEV, if they want to
listen, we can get to that part in a second.

328
00:25:53,507 --> 00:26:11,507
Don't you think it's kind of arbitrary? I mean, what if someone said we have a new,
we have ASIC boost, or we have a new type of cooling tower or something, and in fact
you cannot spend, like, you can't spend like $4.50 on mining, right, because there's
a minimum viable ASIC size, so there's all these various scale economies.

329
00:26:11,507 --> 00:26:40,507
Sure, but I think once you get past a fairly low margin, today, it doesn't scale
that well, right, if you, there are economies of disk scale as well as economies of
scaling mining, right, so there are, you can have super cheap, very small scale
power, or you can have a very large mining farm that has other economies of scale on
cooling, on, you know, cheaper hardware, but also then you might pay a higher power
bill because you don't, you're not using unused power or excess power or something
like that.

330
00:26:42,507 --> 00:26:47,507
But didn't you say just a moment ago about this, it has to be a perfectly linear
scale, which is like, basically not going to happen.

331
00:26:47,507 --> 00:27:05,507
No, of course it doesn't have to be perfect, it has to be as close as we can, in
order for decentralization to be, to exist in a large way, it has to be as much as
we can get it, and ideally it's in fact not linear, it's sub-linear, right, ideally
as you get bigger, it costs more, now we can't actually, that's not real, and that
can't happen, but like...

332
00:27:05,507 --> 00:27:11,507
Partly linear, partly super linear parts of, you know, everything's just a big mix,
in the real world everything's like a big mix, of course.

333
00:27:11,507 --> 00:27:18,507
Right, but the point is that it has to be kind of sort of, as close as, you know, as
low as we can get it, as close to linear as we can get it.

334
00:27:18,507 --> 00:27:46,507
And the, you're right, obviously, there's nothing inherently unique about MEV, it's
like a CapEx, right, an OpEx, it's just another cost, but I think it's a very
substantial other cost, and we've seen in Ethereum that there are only, I mean
there's really only like three companies that are able to do it competitively, and
this is like a huge market, I mean they make a good bit of money, there's a lot of
people competing.

335
00:27:46,507 --> 00:28:15,507
I think Ethereum handles it very bad. Okay, I'm going to wrap up my third thing, and
then I will let other people get a word in, or whatever, but as you probably know,
since we discussed, we've had this conversation before, in October 2016, you and I,
with Adam back, and this led to the publication in January 2017 when I published
Blind Merged Mining, which, although Ethereum can't do this on their L1, I mean they
probably can do it, it doesn't really make any difference, it's a sort of...

336
00:28:15,507 --> 00:28:37,507
it doesn't really make any difference, it's so easy to make layers, but the whole
Blind Merged Mining idea is basically that anything that's software, you can
outsource the entire project to someone else who makes the block, they pay whatever
the costs are, if they have to hire engineers, whatever, so either this project is
losing money, or it's making money.

337
00:28:37,507 --> 00:29:00,507
If it's losing money, then nothing matters, because people will just stop doing it
eventually, and miners certainly don't want to buy into losing money, but if they
make money, then since anyone can run this software, then you have just like a
bidding market where people bid to put the L2, so the things that you mentioned
about like roll-ups and things, they seem to be going way worse than...

338
00:29:01,507 --> 00:29:11,507
So I don't know if you disagree that that is a solution to MEV, because the Blind
merge mining transforms all of the L2's block value, both the transaction fees and
anything else that anyone can cash in.

339
00:29:12,507 --> 00:29:25,507
Blind Merged Mining is very, very similar to the marketplace that Ethereum has, where
you're a staker, someone else is going to tell you what to put in your block, and
they're going to pay you the optimal amount, and hopefully you get a good return on
that.

340
00:29:26,507 --> 00:29:41,507
I think the point of looking at Ethereum here is that they've built that, and it
works great. It does, from an efficient market perspective, it actually works fairly
well. There are only a few companies that can do it, but they are very competitive,
but that results in strong centralization pressure.

341
00:29:42,507 --> 00:29:44,507
Okay, I'm almost done, I promise. I'm sorry.

342
00:29:44,507 --> 00:30:06,507
With Blind Merged Mining, it is anyone can run the software, it captures all the
transaction fee and the other benefits, and it's a case where the miner still builds
the L1 block, they just say, I'm not looking at the L2 at all, I don't care, it's
the blind part, and they just see someone will pay $80,000, someone will pay
$81,000, someone will pay $81,000.

343
00:30:06,507 --> 00:30:12,507
It's precisely that first part of the claim. It's that anyone can run the software,
and they can capture all the benefits.

344
00:30:12,507 --> 00:30:15,507
That's BIP301.

345
00:30:15,507 --> 00:30:18,507
That's the question. Can they capture all of the benefits?

346
00:30:18,507 --> 00:30:20,507
So anyone can do...

347
00:30:20,507 --> 00:30:32,507
The claim is that, well, you can design a system where anyone can just run a piece
of open source software, and they can capture all of the meth, they can capture all
the benefits, they can get all of the value of creating the block template, right?

348
00:30:32,507 --> 00:30:39,507
And I think that, in my view, Ethereum has shown us that if you have a sufficiently
complicated system, now not true today in Bitcoin, right?

349
00:30:39,507 --> 00:30:46,507
Basically, anyone can run Bitcoin Core, build a block template that's like 99.999%
as good in capturing all the benefits.

350
00:30:46,507 --> 00:30:56,507
But in Ethereum, you have to invest a lot of money on high quality engineers, on the
best engineers, in order to actually capture all the benefits.

351
00:30:56,507 --> 00:31:14,507
I guess one thing that I was confused about when you were talking with Paul is that
I guess I understood blind emerge mining as basically being the only problem that I
guess would be like potentially sufficiently complex, would ultimately just be like
how much somebody is willing to pay to have it included.

352
00:31:14,507 --> 00:31:16,507
So it doesn't seem like too complex.

353
00:31:16,507 --> 00:31:19,507
Blind Merged Mining is very straightforward.

354
00:31:19,507 --> 00:31:30,507
And if that other system, if that Drivechain is just kind of a transaction
Drivechain, it doesn't have all of these like Ethereum smart contract deep complex
thing that creates a lot of meth.

355
00:31:30,507 --> 00:31:31,507
But why does that make any difference?

356
00:31:31,507 --> 00:31:32,507
What? Sorry?

357
00:31:32,507 --> 00:31:33,507
But why does it make any difference?

358
00:31:33,507 --> 00:31:36,507
Well, because that creates a lot of meth, right?

359
00:31:36,507 --> 00:31:48,507
No, because on L1, the miners just see it as – L1, all the miners see is a new L1
transaction coming in that says I'll pay $80,000 to have this hash included in the
Coinbase.

360
00:31:48,507 --> 00:31:51,507
I'll pay $81,000. I'll pay $81,007.

361
00:31:52,507 --> 00:32:05,507
In principle, I agree with you that if you have a Drivechain-like system, if you've
split that between L1 and L2, it's probably not super centralizing for the L1.

362
00:32:05,507 --> 00:32:09,507
It's certainly less centralizing for the L1 than if you did these things on the L1.

363
00:32:09,507 --> 00:32:11,507
So I think to some extent we agree there.

364
00:32:11,507 --> 00:32:13,507
It's just kind of like a weird inscription.

365
00:32:13,507 --> 00:32:16,507
I'd like to change the topic. Can we talk about block time?

366
00:32:17,507 --> 00:32:22,507
Real quick, Paul. We're going to – you have to have other people fight your fights
at this point.

367
00:32:22,507 --> 00:32:27,507
We're going to bring you offstage. I love you, man.

368
00:32:27,507 --> 00:32:29,507
He pulled me onstage.

369
00:32:29,507 --> 00:32:34,507
Hey, I did say I agree. Now Paul can be happy and we can –

370
00:32:34,507 --> 00:32:41,507
So one thing I didn't hear you address in your talk was the difference in block
times between Ethereum and Bitcoin.

371
00:32:41,507 --> 00:32:42,507
That's a great question.

372
00:32:42,507 --> 00:32:45,507
The last time I checked, Ethereum had 12-second block times.

373
00:32:45,507 --> 00:32:51,507
So isn't that 500 more times as much opportunity for MEV in each block?

374
00:32:51,507 --> 00:32:54,507
I think it's just different MEV.

375
00:32:54,507 --> 00:32:57,507
So I think this comes up all the time. I think this is a great topic.

376
00:32:57,507 --> 00:33:02,507
So I think to some extent – so I think Ethereum – it's not just the block time.

377
00:33:02,507 --> 00:33:09,507
It's also the block – the Bitcoin – when a block is found as random, in Ethereum
it's very fixed.

378
00:33:09,507 --> 00:33:12,507
So that changes the MEV marketplace very dramatically.

379
00:33:12,507 --> 00:33:16,507
In Ethereum, it does create that fixed window.

380
00:33:16,507 --> 00:33:27,507
Yes, the block time, but also the fixed window, the fact that like, okay, it is on
this millisecond that I have to have my block at all of the other nodes and no
later.

381
00:33:27,507 --> 00:33:34,507
That makes the high-frequency trading dynamic of sex arbitrage much worse.

382
00:33:35,507 --> 00:33:39,507
So you can – if you – you were literally a high-frequency trader at that point.

383
00:33:39,507 --> 00:33:45,507
You have an exact fixed window, and you have to figure out when you have to meet
that window buy and whatever.

384
00:33:45,507 --> 00:33:54,507
So that makes it worse in principle in Ethereum, but I don't think that means that
proof of work has no problems.

385
00:33:54,507 --> 00:33:56,507
I think that – you're right.

386
00:33:56,507 --> 00:34:09,507
I think that the general argument that I've heard many times – well, just one of the
variants of the arguments that I've heard is that, well, people won't use a DEX on
Bitcoin because the block times are so wide.

387
00:34:09,507 --> 00:34:11,507
People just won't use it.

388
00:34:11,507 --> 00:34:14,507
That it's – your slippage is going to be too high.

389
00:34:14,507 --> 00:34:17,507
You're going to get such bad execution of your orders.

390
00:34:17,507 --> 00:34:21,507
Why would you use a DEX on Bitcoin?

391
00:34:21,507 --> 00:34:32,507
And as a corollary to that, if there exists a sequenced rollup – so if you have one
of these rollups that's sequenced or basically it's like a pseudo.

392
00:34:32,507 --> 00:34:34,507
You can think of it like a separate blockchain, but only kind of.

393
00:34:34,507 --> 00:34:35,507
It's CSV.

394
00:34:35,507 --> 00:34:38,507
But it has some kind of central one miner.

395
00:34:38,507 --> 00:34:41,507
Like that miner gets to decide all the transactions in the block.

396
00:34:41,507 --> 00:34:48,507
That means that those – that other system can kind of have a virtual block time much
faster.

397
00:34:49,507 --> 00:34:58,507
If that exists and there's a DEX on that, and then there's also a DEX on Bitcoin,
your execution is just going to be way better on this centralized thing.

398
00:34:58,507 --> 00:35:00,507
And your execution on Bitcoin is going to really suck.

399
00:35:00,507 --> 00:35:02,507
So people are just not going to use the one on Bitcoin.

400
00:35:02,507 --> 00:35:05,507
They're going to use the one on the centralized thing, and you're going to have
great execution.

401
00:35:05,507 --> 00:35:07,507
I think that's probably true.

402
00:35:07,507 --> 00:35:17,507
And I think this gets back to kind of my point about evaluating these things very
quickly gets into what will people use and what's like – what are people going to
build and what are people going to use?

403
00:35:17,507 --> 00:35:26,507
I think clearly if we had EVM on Bitcoin, if we had unlimited flexibility on
Bitcoin, people would build a DEX on Bitcoin.

404
00:35:26,507 --> 00:35:27,507
It would exist.

405
00:35:27,507 --> 00:35:30,507
There's open question as to whether people would use it.

406
00:35:30,507 --> 00:35:34,507
Would people prefer to use that over some rollup DEX or whatever?

407
00:35:34,507 --> 00:35:36,507
And I don't know the answer.

408
00:35:36,507 --> 00:35:39,507
I don't think it's as simple as saying your execution is going to be better.

409
00:35:39,507 --> 00:35:40,507
Here, people can use that.

410
00:35:40,507 --> 00:35:42,507
People go where there's liquidity.

411
00:35:42,507 --> 00:35:51,507
Yeah, some people will prefer to use that, but maybe it ends up having some other
censorship concerns that causes people to want to use the decentralized one.

412
00:35:51,507 --> 00:35:52,507
I don't know.

413
00:35:52,507 --> 00:35:55,507
Really quick, you said an alternative time passer.

414
00:35:55,507 --> 00:35:56,507
What was that?

415
00:35:56,507 --> 00:35:57,507
A sequencer.

416
00:35:58,507 --> 00:36:06,507
Yeah, in rollup lingo, the entity that's selecting the transactions that goes in the
block is the sequencer.

417
00:36:06,507 --> 00:36:08,507
And you can have centralized ones.

418
00:36:08,507 --> 00:36:09,507
You can have multi-sig ones.

419
00:36:09,507 --> 00:36:14,507
You could even have no sequencer or like a virtual sequencer that's just the L1
miners.

420
00:36:14,507 --> 00:36:18,507
And then they're involved in the marketplace.

421
00:36:18,507 --> 00:36:21,507
So they can take the MEV, right?

422
00:36:21,507 --> 00:36:22,507
They can capture that MEV directly.

423
00:36:22,507 --> 00:36:29,507
They can say, okay, I'm selecting the transactions that go in the rollup, and
there's all this MEV in the rollup, and I'm going to claim it for myself.

424
00:36:29,507 --> 00:36:32,507
And then none of it kind of leaks to the L1.

425
00:36:32,507 --> 00:36:34,507
And this is kind of what Paul was talking about a little bit.

426
00:36:34,507 --> 00:36:42,507
Like you have this other chain, and there's some entity, there's some process by
which someone else gets that MEV, and then it doesn't leak.

427
00:36:42,507 --> 00:36:49,507
And in Paul's design, it does leak, but it doesn't cause centralization necessarily.

428
00:36:50,507 --> 00:36:52,507
I'm interested in a very interesting debate.

429
00:36:52,507 --> 00:36:54,507
I want to simplify the assumptions.

430
00:36:54,507 --> 00:37:03,507
So if we look at Ethereum, which is very illustrative comparing to Bitcoin, I want
to identify what are the key differences that make MEV worse or better on each
system.

431
00:37:03,507 --> 00:37:05,507
So it seems like there's a couple of things we're rounding on.

432
00:37:05,507 --> 00:37:07,507
One of them is fixed-time blocks.

433
00:37:07,507 --> 00:37:09,507
Like if you know when the block lands, that's really bad.

434
00:37:09,507 --> 00:37:13,507
Another is global state, which we don't have on Bitcoin.

435
00:37:13,507 --> 00:37:14,507
That makes it really bad.

436
00:37:14,507 --> 00:37:18,507
Are there any other like fundamental things you can think of?

437
00:37:18,507 --> 00:37:19,507
I think it's that.

438
00:37:19,507 --> 00:37:21,507
But I think to a large extent, it ends up being cultural.

439
00:37:21,507 --> 00:37:24,507
It ends up being what do people use Ethereum for.

440
00:37:24,507 --> 00:37:30,507
And it makes this debate so abstract and so frustrating.

441
00:37:30,507 --> 00:37:33,507
And I don't kind of like having that debate.

442
00:37:33,507 --> 00:37:37,507
But that's why I want to narrow it down and get to some concrete stuff.

443
00:37:37,507 --> 00:37:39,507
I think that really is a big part of it.

444
00:37:39,507 --> 00:37:46,507
It's like how do people use the system, and how do people choose to use the system,
and not just what abstract features does it have.

445
00:37:46,507 --> 00:37:47,507
I think it matters a lot too.

446
00:37:47,507 --> 00:37:52,507
Like if we had arbitrary things on Bitcoin, would people –

447
00:37:52,507 --> 00:37:55,507
I think there's an open question of like, okay, we're going to build all this stuff
on Bitcoin.

448
00:37:55,507 --> 00:38:00,507
And like will people actually use it versus using Ethereum?

449
00:38:00,507 --> 00:38:09,507
Like they might not want to use – they might want to continue to use Bitcoin to
trade and use Bitcoin as a payment network and a savings network and store value and
whatever.

450
00:38:09,507 --> 00:38:13,507
And they might actually not care that much about having a DEX on Bitcoin.

451
00:38:13,507 --> 00:38:14,507
I don't know.

452
00:38:14,507 --> 00:38:20,507
So your question was that there – what is the issue with the global state and fixed
block size that –

453
00:38:20,507 --> 00:38:30,507
Well, specifically I want to narrow down the design features of a cryptocurrency
that cause MEV to blow up and become big problems.

454
00:38:30,507 --> 00:38:33,507
And I think there's two, and I want to know what the others are.

455
00:38:33,507 --> 00:38:36,507
I mean that's pretty much the gist of it.

456
00:38:36,507 --> 00:38:45,507
Like Matt was saying, the constraint of like the long block intervals and small
block size is inherently like a natural defense mechanism against that for Bitcoin.

457
00:38:45,507 --> 00:38:51,507
Like that's exactly my concern with Paul's proposal and two-way pegs in general.

458
00:38:51,507 --> 00:39:02,507
I think like that problem bleeds very deeply into the base layer depending on what
people build on second layers if you have an open peg and an open sequence system.

459
00:39:03,507 --> 00:39:09,507
Like the big concern I have with Blind Merged Mining is this idea that miners don't
have to care at all.

460
00:39:09,507 --> 00:39:25,507
But that's inherently assuming that whoever is operating this thing is making
profit, has profit left over after what they're paying to miners, which is implying
there's an extra profit margin for miners to capture if they just engage in this
themselves directly.

461
00:39:25,507 --> 00:39:32,507
And given the nature of a peg like that, there is zero way to constrain the type of
system that it's pegging into.

462
00:39:32,507 --> 00:39:39,507
Like once you implement that peg, the nature of the system that it's going to link
into is just whatever you want it to be.

463
00:39:39,507 --> 00:39:50,507
And then depending on how you architect or design that, like you take the base
layer, you have an Ethereum sidechain, then you have just open sequencer Ethereum
rollups.

464
00:39:50,507 --> 00:39:54,507
And that can just domino the whole way down to base layer miners.

465
00:39:54,507 --> 00:40:03,507
Like you don't have a closed sequencer on a rollup defending the MEB, which leaks
into the chain on the second layer, which then leaks down onto the first layer.

466
00:40:03,507 --> 00:40:11,507
So I think – yeah, so I think to answer your question more concretely, I think the
other – the reason I point out kind of how people use it is like, okay, flexibility
is one.

467
00:40:11,507 --> 00:40:19,507
It's like the overarching, the biggest top line difference between Bitcoin and
Ethereum is how much complexity can you fit on the chain.

468
00:40:20,507 --> 00:40:32,507
That's the big thing, and it is possible that we will end up there with just – with
client-side validation stuff, with the BRC20 people are working on some stuff in
this direction.

469
00:40:32,507 --> 00:40:37,507
Like it's possible that someone will assign client-side validation, and now we have
all the complexity, and now we're stuck.

470
00:40:37,507 --> 00:40:40,507
You just joined us. Go ahead.

471
00:40:40,507 --> 00:40:42,507
I had a couple questions in different directions.

472
00:40:42,507 --> 00:40:51,507
I mean, to that point, something like 50% of the network merge mines rootstock, a
large percentage of the network also merge mines fractal.

473
00:40:51,507 --> 00:41:00,507
You can – from a Bitcoin L1 perspective, you could imagine that being a rollup or
like a CSV type system where you have a different set of consensus rules.

474
00:41:00,507 --> 00:41:20,507
If L1 miners are ultimately able to sequence transactions in those other blockchains
and choose the canonical ordering of the transactions to then merge mine into
Bitcoin, then miners who are running the arbitrarily complex software to validate
those blocks can earn more revenue per unit of hash than miners who are not.

475
00:41:20,507 --> 00:41:28,507
So it sounds like the conditions for creating Meeble kind of already exist today
with those kinds of systems.

476
00:41:28,507 --> 00:41:36,507
Oh, absolutely, yeah. Those could absolutely cause Meeble. I think Paul would say
you should do blind merge mine. You should – and I think this is right here, right?

477
00:41:36,507 --> 00:41:55,507
I think that you can – as long as it is the second-layer system, like, yeah, you
could build kind of the Ethereum marketplace where like the rootstock blocks are bid
up by some third-party capturing the MEV, and then the L1 miners don't necessarily
get caught up in it as much.

478
00:41:55,507 --> 00:42:14,507
It still is – there's still centralization questions and like how that marketplace
is structured and yada, yada, but I think you're right that like it could end up
there or that separation might be big enough that we can fix it with bidding.

479
00:42:14,507 --> 00:42:22,507
And like, okay, that would result in those systems being censored and being OFAC,
whatever, but like maybe the L1 will survive.

480
00:42:22,507 --> 00:42:29,507
I mean, maybe, but also if it lets the L1 miners deploy CapEx at a higher rate, then
they accrue more hash power and they get bigger.

481
00:42:29,507 --> 00:42:46,507
The other question I had for you is I think a dynamic that we're seeing play out is
people trying to build these systems either on Bitcoin or on other blockchains are
moving towards having some kind of permission sequencer or permission coordinator
role.

482
00:42:46,507 --> 00:43:07,507
You mentioned BRC20 stuff earlier, like in the Ordinal's marketplaces, I think the
marketplaces are starting to learn that because we have 10-minute block times and
because you can do things like RBF transactions where if you have a sequence of
causally related transactions and then somebody pays a higher enough fee, you end up
disrupting the trading experience for everybody who's downstream from that.

483
00:43:08,507 --> 00:43:14,507
So marketplaces are starting to require that they co-sign it or that they somehow
are involved in the transaction.

484
00:43:14,507 --> 00:43:21,507
On the more roll-up-y side, what you're seeing is that it's very capital intensive
to build a roll-up or an L2.

485
00:43:21,507 --> 00:43:29,507
You've got to do wallet development, you've got to build all these different layers,
and those companies don't want to leak that enterprise value to L1 miners.

486
00:43:29,507 --> 00:43:31,507
They want to capture it all for themselves.

487
00:43:32,507 --> 00:43:48,507
So I feel like the arc that we're seeing these meta protocols or these L2s I put in
big air quotes on top of Bitcoin going is towards permission sequencing or
authoritative sequencing while at the same time we have merge mining of arbitrary
sidechains happening.

488
00:43:48,507 --> 00:43:57,507
So I feel like we're already in a place where the worst possible thing could happen
but isn't and the direction that everybody wants to be going is actually better.

489
00:43:57,507 --> 00:43:59,507
But no one uses Rootstock.

490
00:43:59,507 --> 00:44:00,507
Fair.

491
00:44:01,507 --> 00:44:04,507
To your point, it matters how much people actually use these.

492
00:44:04,507 --> 00:44:16,507
So I think you're right. We have this worst case possible right now but also you're
totally right that the direction of people building roll-ups and stuff is towards
centralized sequencers.

493
00:44:16,507 --> 00:44:17,507
It is on Ethereum, it is here.

494
00:44:17,507 --> 00:44:18,507
Yep.

495
00:44:18,507 --> 00:44:25,507
And so maybe it's going to be totally fine and this is just totally not worth
worrying about because everything is going to be centralized and this is going to be
fine.

496
00:44:26,507 --> 00:44:32,507
I don't know if I want to bet the farm on that one but I think it might be.

497
00:44:32,507 --> 00:44:34,507
So we got one last –

498
00:44:34,507 --> 00:44:36,507
If you want to make a final question or comment then –

499
00:44:37,507 --> 00:44:54,507
When you were describing marketplaces that exist, I guess the one thing that it
seemed to me was that if you select the MEV which you are going to have a
marketplace for, you're kind of picking winners in that sense.

500
00:44:54,507 --> 00:45:02,507
If people have a specialty of like, oh, I know a lot about this type of MEV but
there's no marketplace built out for it.

501
00:45:03,507 --> 00:45:26,507
So I guess the solution that I see then is having highly modularized ways of
marketing MEV which I see that as basically sidechains and having a secondary market
on the coins themselves dependent on their likelihood of being able to peg back in
at the original rate.

502
00:45:27,507 --> 00:45:31,507
Yeah, certainly it's the case that the marketplace needs to be sufficiently
flexible.

503
00:45:31,507 --> 00:45:37,507
You're totally right that the marketplace has to be sufficiently flexible to capture
the different kinds of MEV extraction that people want to do.

504
00:45:37,507 --> 00:45:47,507
I think my point about raising marketplace design here is that we really want to
build something that allows for separated order books.

505
00:45:47,507 --> 00:46:00,507
So I think the Ethereum MEV marketplace problem in part is that they just have like,
you bid – the only marketplace is for a full block and you must bid for a full block
and that's the only type of order you can put or you can submit.

506
00:46:00,507 --> 00:46:10,507
But you should be able to submit an order that is just like, I know how to do MEV
extraction for this smart contract and I'm going to submit orders that only impact
that smart contract and don't impact anything else.

507
00:46:10,507 --> 00:46:15,507
And the miner should be able to take my order and somebody else who specializes in
something else should be able to specialize in that.

508
00:46:15,507 --> 00:46:17,507
And that would be great.

509
00:46:17,507 --> 00:46:22,507
And I think with UTXOs we're potentially a little bit better able to do this.

510
00:46:22,507 --> 00:46:31,507
And yeah, I mean if it's like sidechains or Drivechains or something where you can
have different types of things on different chains, that's also a possible outcome.

511
00:46:31,507 --> 00:46:38,507
But yeah, I think your point is well made that we have to have sufficiently
complicated order types.

512
00:46:39,507 --> 00:46:41,507
So we're on time.

513
00:46:41,507 --> 00:46:49,507
But if everyone's okay with it and you're okay and everyone – we can go another 10,
15 if that's okay with everyone?

514
00:46:49,507 --> 00:46:50,507
Okay.

515
00:46:50,507 --> 00:46:51,507
All right.

516
00:46:51,507 --> 00:46:53,507
So you have your panelists up here?

517
00:46:53,507 --> 00:46:54,507
Yeah.

518
00:46:54,507 --> 00:46:57,507
We also have – I am allowing people to come back on stage.

519
00:46:57,507 --> 00:46:59,507
Like who would believe it?

520
00:46:59,507 --> 00:47:01,507
Paul in Portland if they want.

521
00:47:01,507 --> 00:47:05,507
But I just want to let you know if you are done, you can come up on stage on your
own.

522
00:47:05,507 --> 00:47:08,507
And we'll probably cap it at this in terms of panelists.

523
00:47:08,507 --> 00:47:09,507
Everyone's queued up now.

524
00:47:09,507 --> 00:47:10,507
Yeah, sounds good.

525
00:47:10,507 --> 00:47:12,507
So make your final points.

526
00:47:12,507 --> 00:47:13,507
Yeah.

527
00:47:13,507 --> 00:47:14,507
So we had two new folks come up.

528
00:47:14,507 --> 00:47:15,507
Yeah.

529
00:47:15,507 --> 00:47:16,507
I have like one question.

530
00:47:16,507 --> 00:47:20,507
It might be a new – I might be opening a can of worms.

531
00:47:20,507 --> 00:47:22,507
So feel free to stop me.

532
00:47:22,507 --> 00:47:24,507
But say I have an opcode proposal.

533
00:47:24,507 --> 00:47:29,507
Like a lot of these layer 2s depend on some opcode, maybe OP_CAT or whatever.

534
00:47:29,507 --> 00:47:34,507
How can I do the due diligence to make sure that I don't introduce me evil?

535
00:47:34,507 --> 00:47:36,507
Or like how can I measure it?

536
00:47:36,507 --> 00:47:38,507
You said it's like a social kind of question, right?

537
00:47:38,507 --> 00:47:42,507
Like how do people use it, not what can they in theory use?

538
00:47:42,507 --> 00:47:46,507
So is it even possible that I do due diligence?

539
00:47:46,507 --> 00:47:48,507
How can I do it?

540
00:47:48,507 --> 00:47:49,507
Yeah.

541
00:47:49,507 --> 00:47:50,507
I mean, you're right.

542
00:47:50,507 --> 00:47:51,507
It opens a can of worms.

543
00:47:51,507 --> 00:47:53,507
But I think the answer is there is no answer.

544
00:47:53,507 --> 00:47:58,507
I think there's just – it is to some extent undecidable.

545
00:47:58,507 --> 00:48:05,507
And it ends up falling to the Bitcoin community of like do we want to – to use the
phrase

546
00:48:05,507 --> 00:48:09,507
from the last comment was do we want to bet the farm on this?

547
00:48:09,507 --> 00:48:13,507
Like there's a lot of value in having recursive covenants and being able to build
all this

548
00:48:13,507 --> 00:48:15,507
cool stuff that everyone wants to build.

549
00:48:15,507 --> 00:48:21,507
And there's also a lot of risk to it depending on how you think people are going to
use it.

550
00:48:21,507 --> 00:48:25,507
And it becomes a question – I mean, it's certainly not a question that I get to
decide.

551
00:48:25,507 --> 00:48:27,507
This is a question for the Bitcoin community, right?

552
00:48:27,507 --> 00:48:28,507
Right.

553
00:48:28,507 --> 00:48:32,507
So it's about being honest, about like looking for an evil.

554
00:48:32,507 --> 00:48:37,507
If someone finds something, then trying to address that or changing a proposal or
like

555
00:48:37,507 --> 00:48:41,507
pointing out this is actually not an attack or something.

556
00:48:41,507 --> 00:48:42,507
Yeah.

557
00:48:42,507 --> 00:48:43,507
We shut it down.

558
00:48:43,507 --> 00:48:45,507
But yeah, I don't like – I don't know.

559
00:48:45,507 --> 00:48:47,507
I mean, I think certainly – yeah.

560
00:48:47,507 --> 00:48:48,507
Yeah.

561
00:48:48,507 --> 00:48:52,507
I mean like we were talking about kind of like the thing that these opcodes change
is

562
00:48:52,507 --> 00:48:57,507
not necessarily that we go from being able to have – from being at risk of having
evil

563
00:48:57,507 --> 00:49:01,507
to not – or from not having any risk to having risk, right?

564
00:49:01,507 --> 00:49:02,507
There's some risk today.

565
00:49:02,507 --> 00:49:06,507
The things that exist today are basically unused or at least not used in any
sufficient

566
00:49:06,507 --> 00:49:08,507
size that it matters.

567
00:49:08,507 --> 00:49:10,507
Sorry.

568
00:49:11,507 --> 00:49:18,507
But – and these other things that people are talking about adding opcodes that are

569
00:49:18,507 --> 00:49:23,507
going to enable people to build more stuff and then like, okay, is that going to –
some

570
00:49:23,507 --> 00:49:25,507
people are going to build stuff with that and hopefully, that will get used.

571
00:49:25,507 --> 00:49:27,507
Otherwise, why are we adding the opcode?

572
00:49:27,507 --> 00:49:29,507
We should certainly not add an opcode if no one is going to use it.

573
00:49:29,507 --> 00:49:30,507
Hopefully, it will get used.

574
00:49:30,507 --> 00:49:36,507
But then what kinds of things are people going to build and that it – I don't have
an

575
00:49:36,507 --> 00:49:37,507
answer.

576
00:49:37,507 --> 00:49:38,507
It's scary.

577
00:49:38,507 --> 00:49:40,507
There are certain patterns that enable Meevil, right?

578
00:49:40,507 --> 00:49:45,507
Like certain script patterns, certain introspections that are just very susceptible
to Meevil and

579
00:49:45,507 --> 00:49:47,507
you can try to avoid them.

580
00:49:47,507 --> 00:49:51,507
And maybe even if you have those patterns, the people don't use it to create Meevil.

581
00:49:51,507 --> 00:49:52,507
Yeah.

582
00:49:52,507 --> 00:49:56,507
I think the other thing too is we can – we can also attack the other – you're
totally

583
00:49:56,507 --> 00:49:57,507
right.

584
00:49:58,135 --> 00:50:06,135
It's scary. I don't have an answer. I think the only other thing is that we should
also look at the problem for the other end and we should build the marketplace,
right?

585
00:50:06,135 --> 00:50:19,135
We should say this exists in some form. It exists today. It's very – it's tiny. So
we should build the marketplace and see and then at least if we end up having a
problem,

586
00:50:19,135 --> 00:50:33,135
whether it's from client-side validation where there's no soft forks or whatever,
whether we add full simplicity to Bitcoin and people build like a DEX and now we
have all kinds of trading and all kinds of sex tax arbitrage and whatever,

587
00:50:33,135 --> 00:50:45,135
at least we have a thing, some marketplace that creates some centralization within
that ecosystem but maybe it's not as bad for the rest of us.

588
00:50:45,135 --> 00:50:59,135
Maybe there's – the best ideas I have are centralized marketplaces like if you want
to be a competitive miner, you have to be a part of some centralized marketplace but
at least it's not exclusive like you have in Ethereum where you have to just be a
part of that centralized marketplace

589
00:50:59,135 --> 00:51:16,135
and you can't take that block template and add your own transactions to remove the
filtering that they might have done. That's the best ideas I have there but at least
that's something we should probably analyze and maybe build and then maybe we're in
a better spot. I don't know.

590
00:51:16,135 --> 00:51:44,135
Well, what's your thoughts on how the trust model of a system impacts whether it
attracts users in the first place? Like you look at something like Rootstock, it's
just a straight up federation and people have been very hesitant to opt in and
really view that as something they would trust and then on the flip side of that,
like client-side systems without a peg that's a trust model users would want to
interact with,

591
00:51:44,135 --> 00:51:59,135
then that is either a trusted system like Rootstock or something that's
free-floating in value. So like the degree of harm that those types of systems
versus a truly trustless peg could create in terms of the amoebal value that would
generate.

592
00:51:59,135 --> 00:52:27,135
Yeah, somebody brought up earlier kind of like what is the difference in Ethereum
and Bitcoin and I tried to make the point kind of poorly that I think the ecosystem
is different and the culture is different and I think that kind of gets to this
point is that like people will build different things and I think people might build
the same set of things on Bitcoin and Ethereum but the things that actually end up
getting adopted might be different because the culture is different and that it
might be the case and really might.

593
00:52:27,135 --> 00:52:48,135
I'm speculating, I'm pretty far out on a ledge here but it might be the case that if
you have a DEX in Bitcoin that people will prefer to use that DEX over a DEX on a
roll-up because the DEX on Bitcoin is more decentralized and less trustful and
whatever versus the one on the roll-up might be, you know, super, it's super
trustful, super, you know, you're trusting some sequencer or whatever.

594
00:52:48,135 --> 00:53:04,135
So that might be a big thing here that we have to think about or it might not matter
at all. It might just be the economics is like it's a little better and like all the
liquidity is going to go there because your execution is just a little better and no
one's ever going to care.

595
00:53:04,135 --> 00:53:31,135
Okay. My biggest criticism to your argument is that you're not making an economic
argument. If you were making an economic argument you would say MEV is bad, it is
something we can control over on the software side even though there are many minor
competitive things that we cannot control and it's going to be this big

596
00:53:31,135 --> 00:53:45,135
and the opportunity cost of saying we can't have any MEV, we need to shut down all
these possibilities is we're going to lose these other possibilities. Would you
characterize as the Ethereum way of life?

597
00:53:45,135 --> 00:54:14,135
I would like to take you back to 2012 when Mike Hearn made I think the first what
Bitcoin can be talk and it was really this era where, you know, we knew what Bitcoin
should be and it needed to be the best, most capable money which included
decentralization because if we lost the decentralization then we would become a
political money.

598
00:54:14,135 --> 00:54:33,135
But within that all of the other aspects of money, all of the other economic value
that people bring to money was critical and we knew that and we knew that's why the
price was going to go up and why we were going to put our own time and energy into
it.

599
00:54:33,135 --> 00:54:51,135
And when you aren't making an economic argument about MEV is this big, the
opportunity cost is this big and we're going to trade this for this, then you sound
like you just don't want the bad things to happen and you don't want change.

600
00:54:51,135 --> 00:55:11,135
And yet in your argument you allow that there is change that you cannot stop, the L2
stuff, the client side stuff and this doesn't seem to bother you and that makes your
argument weaker again. It's not that it doesn't bother you, it's that you recognize
you don't have control over it. That's a better statement.

601
00:55:11,135 --> 00:55:34,135
So that's the thing I want to focus on. If this was going to make sense and we were
going to trade what Bitcoin can and should be for this problem, then we would need
to know how big is this problem and how big is the thing we're giving up on.

602
00:55:34,135 --> 00:56:01,135
And I submit to you that we actually have a market price for the value that we've
walked away from and it is the market price of all the shit coins. When we give up
on things that we could do better, we push people into worse solutions and we push
them towards custodial solutions.

603
00:56:01,135 --> 00:56:11,135
People are heading towards the ETFs and Coinbase because we're not doing a better
job with software and better custody design options.

604
00:56:11,135 --> 00:56:31,135
Yeah, so I think that's a good point. I think a few things. So you're totally right
that we, the Bitcoin community, and I don't – I think maybe you're – my argument
comes off as we must not do a CAD or covenants or whatever.

605
00:56:31,135 --> 00:56:43,135
It's not an argument I want to make. I want to have a discussion about the problems
here because I think that that discussion was basically completely nonexistent a
year ago.

606
00:56:43,135 --> 00:57:06,135
That risks of introducing MEV and now what I would call MEVL was nonexistent a year
ago, and I think that that argument has to be had because you're right. We do need
to – we, not me, the Bitcoin community – needs to make a decision about tradeoffs of
what – about these tradeoffs.

607
00:57:06,135 --> 00:57:22,135
And I think you kind of – you gave it away a little bit. You quoted Mike Hearn and
you said Mike Hearn talked about we have to have all these things that money needs
to have and also decentralization.

608
00:57:22,135 --> 00:57:44,135
And I think my point is that if we do not have – if we end up with lots of MEVL,
whether it's because of some soft fork or whether it's because of client-side
validation, it doesn't matter. You're right. It could be any reason at all. If that
happens, we lose decentralization, and I think that's basically true.

609
00:57:44,135 --> 00:58:07,135
We already lost it. We're not one CPU, one vote. We were supposed to be. So there's
stuff we can control and stuff we cannot control. And we need to use a serenity
prayer and worry about the things we can control, and we need to not worry about
stuff we can't control. And there's a lot of MEV that we cannot control. And you're
saying we can control this much MEV and it's going to make or break us, and I
disagree.

610
00:58:07,135 --> 00:58:35,135
I think we, the Bitcoin community, can absolutely control – we can decide what we
build. We can decide what – we can't control what some people build and choose to
use. But the reality of what the feature set of the chain are does absolutely
dictate to a large extent what ends up getting adoption and what ends up getting
built.

611
00:58:35,135 --> 00:58:44,135
Client-side validation is an example of something that can exist. I mean, Peter Todd
wrote about client-side validation, what, 10 years or more than a decade ago?

612
00:58:44,135 --> 00:59:04,135
But the – it hasn't been – it hasn't gotten any use because the practical tradeoffs
of it – I mean you can only do a one-way peg, right? Like you can maybe build –
maybe people are starting to figure out how to build a two-way peg? Unclear.

613
00:59:05,135 --> 00:59:23,135
But the practical tradeoffs of it are just very high. And if we add a lot of
features, maybe those tradeoffs go away and that does change things. And I think
that is an analysis – you're right, I did not do a thorough analysis of that. That
is absolutely analysis that needs to happen as we look to making changes.

614
00:59:23,135 --> 00:59:46,135
We need to do that level of analysis, and I think you're totally right. My argument
is woefully incomplete and is not a compelling argument that we shouldn't make
changes to Bitcoin. But I think it is hopefully a compelling argument that we should
do the type of analysis that you're talking about prior to deciding to make changes
to Bitcoin because I think there's a lot of risk there.

615
00:59:46,135 --> 00:59:47,135
I think we get stuck.

616
00:59:47,135 --> 01:00:13,135
Pause for a moment. Just for a moment. We're over time. We're actually going into
the next talk. I do think it would be respectful and interesting to just get ten
seconds closing thoughts, but no more questions. This is not like a back and forth.
You can give your closing thoughts, and then you're done. And I really appreciate
you all. So after closing thoughts, we can do round of applause. Please hurry up and
get off the stage.

617
01:00:13,135 --> 01:00:27,135
I think that was my closing thought. I think this was a great point. I think I
hopefully did not make an argument that we shouldn't make changes to Bitcoin. Some
people have interpreted it that way. I'm sure that was not my intent. We need to do
more analysis.

618
01:00:28,135 --> 01:00:45,135
I have a closing thought, which is that, Matt, you are yourself the greatest source
of evil, are you not? Because you are a centralized thing that we cannot – there's
no competition for you. You're a monopoly on yourself. Apparently, miners have to go
to you if they're mining high.

619
01:00:45,135 --> 01:00:46,135
Ten seconds.

620
01:00:46,135 --> 01:00:49,135
And you can bless or reject what they do. So that's my thought.

621
01:00:50,135 --> 01:00:53,135
All right. More later. Go, go, go.

622
01:00:53,135 --> 01:01:06,135
Okay. Ultimately, I believe miners will act rationally, collecting the highest
amount of fees they possibly can in every single Bitcoin block. If that option is
given to them in any way, even if it's through methods of Meeble, it will be
utilized.

623
01:01:07,135 --> 01:01:24,135
I think Matt's right. We just need to really analyze this problem a lot more deeply
and really see, to the best degree we can, if improving the trustlessness of MEV
generating systems is what's going to draw users or not, and whether that's an
important factor.

624
01:01:25,135 --> 01:01:40,135
It also seems to me that there's a lot of potential Meeble that we can get without
changing Bitcoin, with client-side validation and things we don't know about. So we
should do more research into this.

625
01:01:40,135 --> 01:01:44,135
Everybody, please, round of applause for everyone that participated.

626
01:01:46,135 --> 01:01:57,135
Matt, I really appreciate you putting that energy in. That means a lot to me because
I think this is one of the more important topics that we silently disagree on, and
thank you so much. I really do mean that.