1
00:00:00,000 --> 00:00:18,200
Paul, while you're getting mic'd up, maybe Phil would have a question or two just to

2
00:00:18,200 --> 00:00:22,920
get started on the boiling the oceans question.

3
00:00:22,920 --> 00:00:28,280
And I think it's kind of a special opportunity that we shouldn't pass up.

4
00:00:28,280 --> 00:00:33,080
I don't know, I can introduce basically, because Phil said this to me over the
phone,

5
00:00:33,080 --> 00:00:38,840
that the energy efficiency of Bitcoin is often in the news.

6
00:00:38,840 --> 00:00:42,120
And go ahead, you can add to that.

7
00:00:42,120 --> 00:00:43,120
Yeah.

8
00:00:43,120 --> 00:00:45,920
I need to go home soon, so I can't stay for your whole thing.

9
00:00:45,920 --> 00:00:51,400
But if you front load it with the anti-ocean boiling ideas, that would be nice.

10
00:00:51,720 --> 00:00:57,160
Well, I wrote an essay called, Nothing is cheaper than proof of work.

11
00:00:57,160 --> 00:01:03,360
And so the long story short is, I think you should actually, to be an apples to
apples

12
00:01:03,360 --> 00:01:08,000
comparison, you should really look at all of the energy wasted in ATMs and point of

13
00:01:08,000 --> 00:01:09,040
sale machines.

14
00:01:09,040 --> 00:01:15,240
And actually, people have done this, and it's several orders of magnitude lower than
is

15
00:01:15,240 --> 00:01:15,880
currently used.

16
00:01:15,880 --> 00:01:20,720
The other thing that is often not explained well in Bitcoin is that it's a fixed
cost

17
00:01:20,720 --> 00:01:22,320
for a two week period.

18
00:01:22,320 --> 00:01:25,640
So it doesn't matter how many transactions you end up fitting in there.

19
00:01:25,640 --> 00:01:31,080
So even though a tremendous amount of energy is used, as the technology improves and
more

20
00:01:31,080 --> 00:01:36,000
and more transactions can be fit, and theoretically, an infinite number of
transactions, not infinite,

21
00:01:36,000 --> 00:01:40,240
but a very high number of transactions can be fit into each block.

22
00:01:40,240 --> 00:01:44,440
And there are many blocks in each two week period.

23
00:01:44,440 --> 00:01:50,200
The energy used per transaction can actually fall even more, many orders of
magnitude.

24
00:01:50,200 --> 00:01:56,480
But yeah, the issue with these other systems is that so far, no one has designed a
system

25
00:01:56,480 --> 00:02:02,000
that has yet to regress to consuming resources in some form or another.

26
00:02:02,000 --> 00:02:05,880
But I wrote kind of a giant essay about that, which I'd be happy to email.

27
00:02:05,880 --> 00:02:10,320
I think you may find this talk, the first 10 minutes of this talk, more interesting

28
00:02:10,320 --> 00:02:12,800
though than the essay that you could read.

29
00:02:12,800 --> 00:02:14,800
But it's, nothing is cheaper than proof of work.

30
00:02:14,800 --> 00:02:22,000
And if you just Google that phrase, you'll find it because it's right at the top.

31
00:02:22,000 --> 00:02:24,360
Or at least it was the last time I checked.

32
00:02:24,360 --> 00:02:27,280
So, I mean, I wish we had time to talk about so many things.

33
00:02:27,280 --> 00:02:30,200
But should we just get right into this thing here?

34
00:02:30,200 --> 00:02:31,480
Because, oh, yeah, where's my note?

35
00:02:31,480 --> 00:02:33,280
Let's give Fred an applause here.

36
00:02:33,280 --> 00:02:39,000
Okay.

37
00:02:39,000 --> 00:02:44,480
So I keep, I constantly refine this talk and I've been messing with it for the
first, yes.

38
00:02:44,800 --> 00:02:45,800
Sure, no problem.

39
00:02:45,800 --> 00:02:48,800
The first few, you know, we just met each other.

40
00:02:48,800 --> 00:02:49,800
Okay.

41
00:02:49,800 --> 00:02:50,800
Okay.

42
00:02:50,800 --> 00:02:51,800
Okay, thanks.

43
00:02:51,800 --> 00:02:59,800
So yeah, I made this presentation in January and I've changed it around so many
times.

44
00:02:59,800 --> 00:03:03,040
So you have to just bear with me as I keep refining it.

45
00:03:03,040 --> 00:03:08,040
But basically, okay, here we go.

46
00:03:08,040 --> 00:03:13,520
So this, I wrote a project about blockchain prediction markets and I wrote a paper
behind

47
00:03:13,560 --> 00:03:19,560
it and the code, but I don't have anywhere near enough time to talk about the whole
thing.

48
00:03:19,560 --> 00:03:23,760
So I'm just going to talk about my favorite thing that motivated me to create the
project

49
00:03:23,760 --> 00:03:24,760
in the first place.

50
00:03:24,760 --> 00:03:25,760
So here we go.

51
00:03:25,760 --> 00:03:28,200
So then I, this is part of the presentation I changed, I have to read it.

52
00:03:28,200 --> 00:03:34,000
So in modern times, a vast portion of our wealth and the welfare of billions of
lives

53
00:03:34,000 --> 00:03:39,360
depend crucially on the outcome of institutions that we call elections.

54
00:03:39,360 --> 00:03:47,400
So okay, yes.

55
00:03:47,400 --> 00:03:50,720
And bad public policy can ruin just about everything.

56
00:03:50,720 --> 00:03:58,200
So unnecessary wars, poverty, you know, people, miserable social lives as people
obsess over

57
00:03:58,200 --> 00:04:01,840
conspiracy theories, right?

58
00:04:01,840 --> 00:04:06,400
Slower economic growth, meaning that lives in the future will be much worse than
they

59
00:04:06,400 --> 00:04:07,400
otherwise would.

60
00:04:07,440 --> 00:04:11,600
And then since most people who will live will live in the future, assuming that the
species

61
00:04:11,600 --> 00:04:18,040
doesn't go extinct, which is generous maybe, but most people will live in the
future.

62
00:04:18,040 --> 00:04:21,040
And so we should care a little bit about them.

63
00:04:21,040 --> 00:04:26,640
Obviously though, we're talking, today we're talking percentage points of global GDP
growth

64
00:04:26,640 --> 00:04:29,480
and millions of people who are alive today.

65
00:04:29,480 --> 00:04:31,480
So that's just part that I haven't tried.

66
00:04:31,480 --> 00:04:32,480
Okay, yes.

67
00:04:32,560 --> 00:04:38,600
So during the Enlightenment, people came up with this basic kind of structure here.

68
00:04:38,600 --> 00:04:43,920
And you know, political thinkers used what they had learned from the Roman Empire
over

69
00:04:43,920 --> 00:04:48,640
years of tinkering and from Athenian democracy, and they came up with this thing
here.

70
00:04:48,640 --> 00:04:52,400
And a lot of it is pretty cool and worth talking about.

71
00:04:52,400 --> 00:04:56,960
But the most important thing is that we have these secret ballots.

72
00:04:57,040 --> 00:05:03,160
And when their errors inevitably show up, we fire someone using the secret ballot
and

73
00:05:03,160 --> 00:05:08,880
we say, I didn't like this, so it's time for you to get lost.

74
00:05:08,880 --> 00:05:16,040
And so a lot of things, basically all this stuff hinges on this thing here.

75
00:05:16,040 --> 00:05:23,320
Now when it was designed, or what I mean by that is like in the 1800s when the US
Constitution

76
00:05:23,320 --> 00:05:27,320
was written, things were a little different and it sort of worked, I think, a little
bit

77
00:05:27,320 --> 00:05:28,640
better than it works today.

78
00:05:28,640 --> 00:05:35,440
But now it does require, we'll get to that in a second, but it does require the
voter

79
00:05:35,440 --> 00:05:37,200
to do a lot of counterfactual thinking.

80
00:05:37,200 --> 00:05:41,520
So like, would this have been better under this other person versus, you know, is
what

81
00:05:41,520 --> 00:05:43,720
we have just the best that we can do?

82
00:05:43,720 --> 00:05:45,840
And maybe a new person will make it worse.

83
00:05:45,840 --> 00:05:48,480
And they have to do a lot of thinking, basically.

84
00:05:48,520 --> 00:05:53,520
You also have to do some game theoretic reasoning, thanks to something called
electability that

85
00:05:53,520 --> 00:06:03,800
I'm going to explain in tremendous detail that ruins elections to a tremendous
degree.

86
00:06:03,800 --> 00:06:09,640
So the thing that I was mentioning before is that while it was sort of working, I
think,

87
00:06:09,640 --> 00:06:13,280
much better in the 1800s, now it doesn't work as well, I think.

88
00:06:13,280 --> 00:06:20,120
And so some of the impacts here are that as society has progressed, there are more

89
00:06:20,120 --> 00:06:24,640
people living here, and that just means you have a smaller impact per vote.

90
00:06:24,640 --> 00:06:29,800
So if you have three people voting, it's virtually guaranteed that, you know, your
vote will

91
00:06:29,800 --> 00:06:32,020
really count.

92
00:06:32,020 --> 00:06:37,640
But if you make a simple model, which I have done, and you just assume that you
don't know

93
00:06:37,640 --> 00:06:41,520
other people are voting with a coin flip on whether or not they vote up or down or
just

94
00:06:41,520 --> 00:06:47,480
abstain, then at around eight other voters, your probability of affecting the
outcome

95
00:06:47,480 --> 00:06:49,400
at all drops below 50%.

96
00:06:49,400 --> 00:06:52,920
And at 41 other voters, it drops basically to 0%.

97
00:06:52,920 --> 00:06:58,400
So in the United States, like 115 million people vote, so it's kind of a lost cause.

98
00:06:58,400 --> 00:07:03,920
But the other thing is this point about media ecology, which I'd love to make in
more detail,

99
00:07:03,920 --> 00:07:05,960
but when I do, it just drags the presentation on.

100
00:07:05,960 --> 00:07:09,560
But everyone should read this book, Neil Postman's Amusing Ourselves to Death.

101
00:07:09,560 --> 00:07:11,560
It's a really wonderful book.

102
00:07:11,560 --> 00:07:17,760
And in the past, 95% of Americans, this is a very American-centered part of the
talk,

103
00:07:17,760 --> 00:07:23,320
I'm sorry about that, but when Thomas Jefferson aligned all this stuff, almost
everyone was

104
00:07:23,320 --> 00:07:29,560
a literate farmer, and so they had a lot of free time to read, and you had an over
95%

105
00:07:29,560 --> 00:07:34,840
of the population of the United States at that time was people who were literate
farmers.

106
00:07:35,160 --> 00:07:39,920
Almost all the media that was consumed was typography, it wasn't even handwriting,
it

107
00:07:39,920 --> 00:07:41,560
was the printed word.

108
00:07:41,560 --> 00:07:48,000
So it had these qualities of being very abstract and even anonymous, because it
would look

109
00:07:48,000 --> 00:07:54,760
the same no matter who wrote it, and you had Thomas Paine and these other famously
anonymous

110
00:07:54,760 --> 00:07:56,360
pamphlets that were circulated.

111
00:07:56,360 --> 00:08:00,400
So again, I could talk about this point for a very, very long time, but the short
version

112
00:08:00,400 --> 00:08:07,520
is that today, we're bombarded with, another interesting point, two more, is that
people

113
00:08:07,520 --> 00:08:12,200
were very famous, people like Charles Dickens who wrote these giant, giant
long-winded books,

114
00:08:12,200 --> 00:08:16,840
but they were as famous as the Beatles or movie stars or something, and people
really

115
00:08:16,840 --> 00:08:20,280
liked this stuff, this long-form content.

116
00:08:20,280 --> 00:08:27,240
And one of the Lincoln-Douglas debates in the middle of the 1800s between President

117
00:08:27,240 --> 00:08:32,120
Abraham, well, then candidate Abraham Lincoln and the guy he was running against,
one of

118
00:08:32,120 --> 00:08:36,160
those debates, they closed the town for the whole day.

119
00:08:36,160 --> 00:08:41,640
The first person spoke for three uninterrupted hours, just speaking, no visual aids.

120
00:08:41,640 --> 00:08:46,000
Then they took intermission, everyone went home and had dinner with their families
and

121
00:08:46,000 --> 00:08:50,560
came back for the rebuttal, which was four hours of uninterrupted speaking.

122
00:08:50,560 --> 00:08:56,720
And that's like, no, today, this is unthinkable, because we're bombarded constantly
with information,

123
00:08:56,720 --> 00:08:59,920
and some of that information is just entertaining, so there's not enough.

124
00:08:59,920 --> 00:09:04,640
In the past, there was kind of this free bandwidth for people to just think about
what they should

125
00:09:04,640 --> 00:09:09,480
do or who they should elect or how to solve their collective problems, but today,
there's

126
00:09:09,480 --> 00:09:15,000
just this attention economy, and you can't even get people to fill out surveys to
try

127
00:09:15,000 --> 00:09:18,640
and improve their hotel experience or whatever.

128
00:09:18,640 --> 00:09:24,560
And then my third point here is that society is a lot more complicated now.

129
00:09:24,560 --> 00:09:31,000
In the past, I think society was pretty complicated, but it was like, how
complicated can it really

130
00:09:31,000 --> 00:09:36,440
be in an era where, so now, stuff can happen very, very quickly, and you have whole
teams

131
00:09:36,440 --> 00:09:41,120
of specialists, people go to school, they train their entire lives to be what, like

132
00:09:41,120 --> 00:09:46,320
a jet fighter, and they know everything about how very quickly you could take off
and bomb

133
00:09:46,320 --> 00:09:51,200
some facility 800 miles away or something, and there's no way, something like Brexit

134
00:09:51,240 --> 00:09:56,160
or say, like, this is just like outside of the scope of, you know, should we build
more

135
00:09:56,160 --> 00:10:01,600
walls or something, so it's like, you know, a more simple thing, okay.

136
00:10:01,600 --> 00:10:05,600
And again, I've done this presentation a lot, and there's so many things I want to
talk

137
00:10:05,600 --> 00:10:10,400
a lot about, but I'll just touch a little bit briefly on this one here.

138
00:10:10,400 --> 00:10:13,240
You have to know a lot of things if you want to vote well.

139
00:10:13,240 --> 00:10:18,160
One of them is how what your, not only what your existing option is, but what the,
who

140
00:10:18,160 --> 00:10:21,800
the challenger is, and that's, that seems like a low bar, but we'll see that at the

141
00:10:21,800 --> 00:10:27,320
bottom, or over here, we're not even at that stage one, because 71% of Americans
cannot

142
00:10:27,320 --> 00:10:32,360
even name the one congressional representative that is their own congressional
representative,

143
00:10:32,360 --> 00:10:37,000
but you also need to know the challenger and both of their plans, and also whether
or not

144
00:10:37,000 --> 00:10:41,720
they're lying about what their plans really are, and even if they're earnest, you
need

145
00:10:41,720 --> 00:10:45,040
to then figure out what would happen if they actually introduced those plans, and if
it

146
00:10:45,040 --> 00:10:47,120
would be maybe a giant disaster.

147
00:10:47,120 --> 00:10:52,720
So you need to know quite a bit, and usually only a few specialists know that for
some

148
00:10:52,720 --> 00:10:54,840
crucial decisions.

149
00:10:54,840 --> 00:10:59,200
And as I just mentioned, basically no one even knows who they, who it is that they
elected

150
00:10:59,200 --> 00:11:05,560
in the United States, let alone the challenger or any of the policies, like you can
go down

151
00:11:05,560 --> 00:11:06,560
the line.

152
00:11:06,560 --> 00:11:10,400
Sometimes I, for fun, I ask people to name their congressperson, or I name the, who
they

153
00:11:10,400 --> 00:11:11,640
ran against.

154
00:11:11,640 --> 00:11:12,640
No one can remember anything.

155
00:11:12,640 --> 00:11:16,440
I bet, sometimes I do already, people can't even remember like who Hillary Clinton's
running

156
00:11:16,440 --> 00:11:17,440
mate was and things.

157
00:11:17,440 --> 00:11:21,320
That was like just a very short time ago.

158
00:11:21,320 --> 00:11:25,720
And I have a funny story about these lobbyists who are on a plane, but I think I
probably

159
00:11:25,720 --> 00:11:26,720
won't tell it unless we have time.

160
00:11:26,720 --> 00:11:27,720
I was looking at that on the slider.

161
00:11:27,720 --> 00:11:28,720
Do you want it?

162
00:11:28,720 --> 00:11:29,720
I'll tell it to you.

163
00:11:29,720 --> 00:11:30,720
What is that?

164
00:11:30,720 --> 00:11:38,760
Okay, so I happen to know someone, I don't want to out this person, but this person
was

165
00:11:38,760 --> 00:11:45,800
a congresswoman, and she gave this story, which is very funny, and she was on, she
is

166
00:11:45,800 --> 00:11:53,520
from, she is a member of the House of Representatives from State A, and she was in
State B flying

167
00:11:53,520 --> 00:11:55,880
to Washington, D.C.

168
00:11:55,880 --> 00:12:01,320
And lo and behold, next to her on the plane are a bunch of lobbyists, and they were
flying

169
00:12:01,320 --> 00:12:07,280
also from State B to Washington, D.C., so that they could lobby someone.

170
00:12:07,280 --> 00:12:15,300
And the person that they were lobbying was from a different state, State C, and the
lobbyists,

171
00:12:15,300 --> 00:12:18,580
they had literally studied this person, I think it's not even a joke, they had
literally

172
00:12:18,580 --> 00:12:23,580
studied them down to their shoe sizes, that's a phrase in English, but I think they
literally

173
00:12:23,580 --> 00:12:26,940
were like buying this person like some shoes or something crazy like that, or they
had

174
00:12:26,940 --> 00:12:28,260
like done some of this stuff.

175
00:12:28,260 --> 00:12:36,140
So they had all this pages and pages on this one person, the target of their
lobbying,

176
00:12:36,140 --> 00:12:42,620
and then she asked them if they knew who their own representative was in State B,
and they

177
00:12:42,620 --> 00:12:43,620
didn't know.

178
00:12:43,620 --> 00:12:48,260
So they're professional lobbyists, it's their job to influence the congresspeople,
but they

179
00:12:48,260 --> 00:12:53,580
only know the one that they're paid to know, even if they work in the space, they
can't

180
00:12:53,580 --> 00:12:55,420
be bothered to look it up.

181
00:12:55,420 --> 00:13:00,020
And this is a shock, if you pair it with something else, I just think this is
unbelievable, and

182
00:13:00,020 --> 00:13:04,820
there's no possible way that this is what people had in mind in ancient Athens.

183
00:13:04,820 --> 00:13:09,500
I just think that like there's no possible way that these are the re-election rates,

184
00:13:09,500 --> 00:13:15,060
and this is zero, and this is 100% for the United States Senate and the House of
Representatives,

185
00:13:15,060 --> 00:13:20,820
and almost always the same people get re-elected, even though no one really knows
who they are,

186
00:13:20,820 --> 00:13:26,140
and the congressional approval rate is always very low, and for the last 20 years or
so,

187
00:13:26,140 --> 00:13:34,100
I think it's even been in the single digits, the September 11th, 2001 era,
notwithstanding,

188
00:13:34,100 --> 00:13:39,740
that it's just been like below 40% forever, and then just falling, and then below
10%.

189
00:13:39,740 --> 00:13:46,100
So definitely this is this rational ignorance component here that people just, they
can't

190
00:13:46,100 --> 00:13:51,300
be bothered to look into this, like who they should vote for, it's just too much
work.

191
00:13:51,300 --> 00:13:56,220
And now I have this thing about electability, so there's this joke, and I actually,
I researched

192
00:13:56,220 --> 00:14:00,060
this on a plane, and I'm wondering if anyone knows what I'm talking about here,
because

193
00:14:00,060 --> 00:14:05,020
I looked this up, and there's a very famous episode of the American television
comedy

194
00:14:05,020 --> 00:14:09,100
South Park, and I looked up, and I was like, does anyone in Amsterdam watch this,
and apparently

195
00:14:09,100 --> 00:14:15,060
it is very popular in Amsterdam, and this is from a bunch of years ago, and someone

196
00:14:15,060 --> 00:14:19,300
said they know what this is, and so this is a joke, and you can see I censored some
of

197
00:14:19,300 --> 00:14:23,100
the letters here, but do you want to like just shout out, or explain what's going on

198
00:14:23,100 --> 00:14:27,620
in this, I guess you don't have the microphone, but yeah, exactly.

199
00:14:27,620 --> 00:14:32,140
So this joke is that, it's done very well, as only South Park can do with these
funny

200
00:14:32,140 --> 00:14:38,780
cartoons, so that it's just cleverly hidden enough, the parallels to the real world,
but

201
00:14:38,780 --> 00:14:43,620
basically this joke kind of spirals out of control, and everyone at this fourth
grade

202
00:14:43,620 --> 00:14:49,740
class is forced to vote between two candidates for mascot, and one of them is a
giant douche,

203
00:14:49,740 --> 00:14:53,900
and the other one is a turd sandwich, and then the child goes home to complain about

204
00:14:53,900 --> 00:14:59,740
how inane this is, and his parents tell him that in every election, it's always a
choice

205
00:14:59,740 --> 00:15:06,980
between a giant douche and a turd sandwich, so that's a very funny thing, so I'll be
back

206
00:15:06,980 --> 00:15:12,060
with more about rational ignorance, and I'm going to talk more about electability
right

207
00:15:12,060 --> 00:15:17,620
after this, but both of the solutions are really, can be solved using the same, or I

208
00:15:17,620 --> 00:15:22,660
think they, my guess is that they, my theory is that they can be solved with these
things

209
00:15:22,660 --> 00:15:28,860
called conditional prediction markets, so I hope that you have an open mind to that.

210
00:15:28,860 --> 00:15:32,540
Now I'm going to talk about this electability thing for a second, because it's
amazing how

211
00:15:32,540 --> 00:15:38,660
much this one concept ruins not only elections, but lots of other things, so why
don't we

212
00:15:38,660 --> 00:15:44,460
just, if we don't like Congress or something, in the United States, again I
apologize that

213
00:15:44,460 --> 00:15:50,780
I didn't want to look up, I did look up a bunch of things about the government here

214
00:15:50,900 --> 00:15:54,540
in Amsterdam, but I just thought it would be too much work and too contrived, too
much

215
00:15:54,540 --> 00:15:59,900
of a put-up job to kind of pretend that I know anything about what's going on here
politically,

216
00:15:59,900 --> 00:16:03,700
but if you don't like the people in charge, and it's a very common thing to complain
about,

217
00:16:03,700 --> 00:16:07,500
I actually have other slides that I took out about like the mayor from the Simpsons,
because

218
00:16:07,500 --> 00:16:13,020
I wasn't sure if anyone here would get that, but it's like every time politicians
are portrayed

219
00:16:13,020 --> 00:16:17,900
in popular media, like everyone kind of doesn't like them, and so it's like, well,
why not,

220
00:16:17,900 --> 00:16:21,980
because we can just get rid of them at any time, and there is an answer to this
question,

221
00:16:21,980 --> 00:16:28,860
which is that why is it always between two unlikable candidates, and that is what
I'm

222
00:16:28,860 --> 00:16:34,380
about to explain, something called multi-factor competition, so competition is
great, and

223
00:16:34,380 --> 00:16:38,860
it works really, really well, but it doesn't work, you have to compete on one thing
at

224
00:16:38,860 --> 00:16:44,660
a time, so my example here is going to involve one person choosing where he eats for
dinner

225
00:16:44,660 --> 00:16:51,540
alone, but then he's going to choose where he eats with a group of people, and the
group

226
00:16:51,540 --> 00:16:56,340
is going to ruin everything, even though everyone involved is individually rational
and all

227
00:16:56,340 --> 00:17:02,860
that, so you want to spend up to $20 on dinner, you're choosing and dining alone,
and so then

228
00:17:02,860 --> 00:17:07,180
all you have to do is compare restaurant A, restaurant B, restaurant C, and what
your

229
00:17:07,180 --> 00:17:12,060
expected satisfaction is, and then you just decide, well, I think my satisfaction
will

230
00:17:12,060 --> 00:17:16,740
be highest with restaurant B, and then you pick restaurant B, and it's very easy for

231
00:17:16,740 --> 00:17:21,980
you, you get the highest satisfaction possible when you have competition on one
dimension,

232
00:17:21,980 --> 00:17:26,940
always, but the secondary effect is that the restaurants are induced to compete on
this

233
00:17:26,940 --> 00:17:32,300
metric here of satisfaction per dollar, so there's a natural corrective force here,
you

234
00:17:32,300 --> 00:17:38,220
can veto people and find the place that you like the most, and the restaurant that's
vetoed

235
00:17:38,220 --> 00:17:43,180
too much goes out of business, and their capital is freed up for other things, which

236
00:17:43,180 --> 00:17:44,180
is desirable.

237
00:17:44,180 --> 00:17:49,220
Okay, so now, this is the part where everything is ruined and why we always have
unlikable

238
00:17:49,220 --> 00:17:53,360
candidates. In the United States, it also, there's a different thing, you guys have
proportional

239
00:17:53,360 --> 00:17:58,700
representation here, which is, I think, worse, and you should read, Karl Popper
wrote an

240
00:17:58,700 --> 00:18:03,260
essay about this in The Economist, that you should read, called The Open Society and
Its

241
00:18:03,260 --> 00:18:09,980
Enemies. So, it's even worse, but the solution I'm going to describe actually works
for both

242
00:18:09,980 --> 00:18:14,780
situations, but you guys have different problems, you have a dictatorship of the
third largest

243
00:18:14,780 --> 00:18:21,300
party or weird party coalitions and other things that I'd also be happy to talk
about,

244
00:18:21,300 --> 00:18:22,300
but it's a much stronger...

245
00:18:22,300 --> 00:18:23,300
Yeah.

246
00:18:23,300 --> 00:18:24,300
Yeah.

247
00:18:24,300 --> 00:18:32,220
Well, okay, we'll see, but it doesn't matter, the solution I'm going to pitch will
work

248
00:18:32,220 --> 00:18:37,380
for both, just as a kind of happy coincidence, maybe. But anyway, here's what's
going on

249
00:18:37,380 --> 00:18:42,340
here, is that a lot of people are choosing to go somewhere and they want to have the

250
00:18:42,340 --> 00:18:48,100
best meal possible, but they also want to show up at the same place that everyone
else

251
00:18:48,100 --> 00:18:54,860
is showing up. And now, your decision depends not only on what you prefer, this is,
excuse

252
00:18:54,860 --> 00:19:00,260
me, sorry, this is the little graphic from the previous slide, depends not only on
what

253
00:19:00,260 --> 00:19:08,100
you prefer, but it also depends on where other people might go, and that is in
itself

254
00:19:08,100 --> 00:19:12,700
a function of what they might prefer. Now, I'm going to be very nice here and assume

255
00:19:12,700 --> 00:19:16,500
that they have the exact same preferences as you, although Kenneth Arrow won a Nobel

256
00:19:16,500 --> 00:19:21,820
Prize for describing that, if these are different at all, we're all doomed and we'll
never be

257
00:19:21,820 --> 00:19:26,300
able to figure out which restaurant to go to, using math. But I'm going to say, even

258
00:19:26,300 --> 00:19:34,260
though they're the same, because your decision depends on where you think they will
go, but

259
00:19:34,260 --> 00:19:38,740
that's a function of what they, what you think they prefer, but also where you think
they

260
00:19:38,740 --> 00:19:44,420
might go. And they, if you think, and so this actually is an infinite, like,
recursing thing

261
00:19:44,420 --> 00:19:49,080
here, it just recurses forever, so I could have just put a bunch of things, but
instead

262
00:19:49,080 --> 00:19:54,620
I just say, well, that's a function of everyone's collective best guess on where
most people

263
00:19:54,620 --> 00:19:58,380
currently plan on going, which I'm going to call the status quo, even though that's

264
00:19:58,380 --> 00:20:06,380
a very kind of free use of that term. But then also, since time and attention and
communication

265
00:20:06,380 --> 00:20:11,540
and just information processing in general is not free, especially in the human
brain,

266
00:20:11,540 --> 00:20:17,620
it also depends on how open to communication and negotiation everyone is. So, what
I'm

267
00:20:17,620 --> 00:20:21,980
saying is, everyone hates restaurant C, but they think everyone else is going there,
so

268
00:20:21,980 --> 00:20:26,380
this is kind of like a family reunion dinner or something where everyone just, they
don't

269
00:20:26,380 --> 00:20:31,180
want to, they don't know each other kind of well enough to say that they hate
restaurant

270
00:20:31,180 --> 00:20:37,020
C and it might be misconstrued as impolitely saying that you don't like the family
gathering

271
00:20:37,020 --> 00:20:42,020
or something. So, this crazy family just ends up going to restaurant C year after
year after

272
00:20:42,020 --> 00:20:47,580
year until someone finally reveals that they actually, they all hate restaurant,
they all

273
00:20:47,580 --> 00:20:54,420
hate restaurant C. So, you have four criteria here, but you have to make one
selection and

274
00:20:54,420 --> 00:20:58,140
you end up choosing based on where you think other people might go because that's
more

275
00:20:58,140 --> 00:21:03,460
important to you than some other things, which basically the analog for this is
that, you

276
00:21:03,460 --> 00:21:08,620
know, no one wants anarchy, so we put up with this dumb process that gives us
unlikable

277
00:21:08,620 --> 00:21:23,220
candidates, but. So, I have a little bit more of a, yes. So, there's, excuse me,
yes.

278
00:21:23,220 --> 00:21:29,020
How far away the restaurant is, is that what you're saying? Yeah, yeah. Yes, well
that

279
00:21:29,020 --> 00:21:33,700
is, that would help in this case because what I'm saying is that there's some
information

280
00:21:33,700 --> 00:21:38,980
that you'd have to process, but if there's something everyone should be able to, if
everyone's

281
00:21:38,980 --> 00:21:42,260
in a group and they're walking around and they have yet to determine the restaurant,

282
00:21:42,260 --> 00:21:49,500
they would rely on commonly available signals like what everyone can see right now
and,

283
00:21:49,500 --> 00:21:53,220
you know, the fact that everyone prefers to walk. Everyone prefers less walking to
more

284
00:21:53,220 --> 00:21:56,980
walking. So, yeah, that would help, but yes.

285
00:21:56,980 --> 00:22:10,420
Okay. So, there's a self-fulfilling prophecy on this last one here, which is that
there

286
00:22:10,420 --> 00:22:16,020
are some people will become cynical, kind of Tyrian Lannister types, and they will
decide

287
00:22:16,020 --> 00:22:19,420
that negotiating to pick a better restaurant is a lost cause, and so there's no
point to

288
00:22:19,420 --> 00:22:23,860
voting or no point to participating, and then you have all the smartest people
dropping

289
00:22:23,860 --> 00:22:30,180
out at which point it becomes even more horrible. So, and this becomes worse when
there's lots

290
00:22:30,180 --> 00:22:34,860
of people and many alternatives. So, in the United States, there's 250 million
adults,

291
00:22:34,860 --> 00:22:41,820
the near infinite number of policies. So, instead, you, you have, you have this
issue

292
00:22:41,820 --> 00:22:45,220
where since you prefer to meet at the same place as everyone else, this is the
so-called

293
00:22:45,220 --> 00:22:49,820
wasted votes problem where you, if you vote for some, the third party candidate, you
just

294
00:22:49,820 --> 00:22:55,500
throw your vote away, and as a result, everyone has to think strategically. That was
that

295
00:22:55,500 --> 00:23:01,700
second thing about electability, the red part, is that in, when you're doing the
primaries

296
00:23:01,700 --> 00:23:08,980
in the United States and in other, this, a version of this tick crops up in every
form

297
00:23:08,980 --> 00:23:14,220
of democracy, but when you are whittling down the choices that will be on the ballot
in

298
00:23:14,220 --> 00:23:19,420
like the pre-ballot phase, you have to know that your...

299
00:23:19,905 --> 00:23:28,905
you, you may, you may end up hurting your own side by voting for it, because you
need to have, you need to pool all of your votes.

300
00:23:28,905 --> 00:23:35,905
It depends a little bit on the voting scheme that's used, but you need to pool all
of your votes so that you have enough to actually win the election.

301
00:23:35,905 --> 00:23:43,905
And this has happened numerous times in election history where someone has split the
vote and caused the side to lose.

302
00:23:44,905 --> 00:23:52,905
I think I explained this already, that you want to stay in sync with others, so
that's kind of what's driving all of this here,

303
00:23:52,905 --> 00:24:00,905
and then you can compare what it would be like if you wanted to start up a new
barbershop versus if you wanted to start up a new Facebook,

304
00:24:00,905 --> 00:24:03,905
and I think I put some blue text up here, yeah, okay.

305
00:24:03,905 --> 00:24:11,905
So, a lot of people don't like Facebook, and they find that it drives them, makes
them miserable and forces them, compels them to pretend that their life is better
than it is

306
00:24:11,905 --> 00:24:19,905
and post weird pictures, and they kind of don't understand what's happening to
themselves, and they don't like the fact that some weird corporation has all
their...

307
00:24:19,905 --> 00:24:25,905
You know, a lot of people really don't like Facebook, if you ask them, but they put
up with it anyway,

308
00:24:25,905 --> 00:24:34,905
because the alternative is not having any Facebook at all and having basically
social isolation in this one narrow sense,

309
00:24:35,905 --> 00:24:40,905
but the alternative is this problem that I've been describing, which is endless and
ultimately futile,

310
00:24:40,905 --> 00:24:48,905
because you will lose negotiation with unmotivated friends and family to switch to
one particular alternative among many,

311
00:24:48,905 --> 00:24:56,905
so I don't say we should switch to Freebook or whatever, but you've got to get all
of your friends and family to switch over,

312
00:24:56,905 --> 00:25:03,905
and they know, when you're talking to them about switching over, they know that they
need to convince their own friends,

313
00:25:03,905 --> 00:25:07,905
who in turn need to convince their own friends and so on and so forth,

314
00:25:07,905 --> 00:25:14,905
since everyone knows that just a little bit of intractability is enough to stall
this whole process out,

315
00:25:14,905 --> 00:25:19,905
there's the self-fulfilling prophecy that I mentioned before, and as I say at the
bottom part of this text,

316
00:25:19,905 --> 00:25:24,905
that it's the same way that we put up with governments, even though it's a very
commonly believed,

317
00:25:24,905 --> 00:25:29,905
and I think pretty accurately believed, that they are less than optimally competent,

318
00:25:30,905 --> 00:25:37,905
which is because the alternatives are no government at all, or some weird, giant,
violent revolution that may not work,

319
00:25:37,905 --> 00:25:38,905
but I think we have a question.

320
00:25:38,905 --> 00:25:42,905
Just quickly on the Facebook thing, I don't use Facebook anymore.

321
00:25:42,905 --> 00:25:43,905
No, I haven't used it in five years.

322
00:25:43,905 --> 00:25:44,905
Exactly.

323
00:25:44,905 --> 00:25:46,905
But we're like the weird guys.

324
00:25:46,905 --> 00:25:51,905
And in fact, I was literally planning on bringing it, I'm literally planning to go
back,

325
00:25:51,905 --> 00:25:55,905
because now that everyone knows how bad it is, I'm a little bit more okay with
bringing it back to them.

326
00:25:55,905 --> 00:26:01,905
So here's the thing, is that you may be tempted to use it because you're a social
friend and family,

327
00:26:01,905 --> 00:26:04,905
but if you don't use it, they have less incentive to use it.

328
00:26:04,905 --> 00:26:05,905
Correct, yes.

329
00:26:05,905 --> 00:26:07,905
No, but that's my point, is that it is...

330
00:26:07,905 --> 00:26:10,905
It becomes a self-fulfilling prophecy if everyone stops using it.

331
00:26:10,905 --> 00:26:12,905
Right, if everyone stopped using it, it would make it easier,

332
00:26:12,905 --> 00:26:15,905
but that's my point, is that there is a kind of gravity to this,

333
00:26:15,905 --> 00:26:19,905
where if the Earth were somewhere else, there would be more gravity keeping it
there.

334
00:26:19,905 --> 00:26:21,905
But it's the same with governments.

335
00:26:21,905 --> 00:26:22,905
Yeah.

336
00:26:22,905 --> 00:26:23,905
No, it is the same, yes.

337
00:26:23,905 --> 00:26:29,905
If we get to a new peak, or a new something, a new local optimum.

338
00:26:29,905 --> 00:26:31,905
Yes, it is the same, I think, yes.

339
00:26:31,905 --> 00:26:34,905
The problem is getting there, that's what I'm trying to explain.

340
00:26:35,905 --> 00:26:36,905
Okay.

341
00:26:36,905 --> 00:26:40,905
So I hinted before, Ken Arrow won the Nobel Prize for this.

342
00:26:40,905 --> 00:26:43,905
If you have some rules, then basically, mathematically,

343
00:26:43,905 --> 00:26:47,905
it's impossible for preferences to be aggregated,

344
00:26:47,905 --> 00:26:51,905
and some of them are just so simple that you almost feel like you can't even

345
00:26:51,905 --> 00:26:55,905
state them like having more than one person involved,

346
00:26:55,905 --> 00:26:56,905
so that there's at least two people,

347
00:26:56,905 --> 00:27:00,905
and some deterministic way of fusing those preferences,

348
00:27:00,905 --> 00:27:04,905
then you can only have one of these two things, mathematically,

349
00:27:04,905 --> 00:27:07,905
which is that you can't hurt an option when everyone ranks it higher.

350
00:27:07,905 --> 00:27:10,905
So when people are persuaded of something,

351
00:27:10,905 --> 00:27:14,905
you can't have the group be persuaded in the same direction,

352
00:27:14,905 --> 00:27:17,905
which is, again, these seem like they should all be no-brainers,

353
00:27:17,905 --> 00:27:20,905
but mathematically, you have to omit one,

354
00:27:20,905 --> 00:27:24,905
and the second one is irrelevant alternatives are really irrelevant.

355
00:27:24,905 --> 00:27:29,905
So if someone, you had a preference for A or B,

356
00:27:29,905 --> 00:27:31,905
I shouldn't say that, because that's no good.

357
00:27:31,905 --> 00:27:36,905
I should say something like, let's say, Churchill versus, I don't know, FDR.

358
00:27:36,905 --> 00:27:39,905
So let's say you're electing Churchill, and you prefer Churchill to FDR,

359
00:27:39,905 --> 00:27:41,905
and then you have Stalin and Hitler.

360
00:27:41,905 --> 00:27:46,905
If you prefer Churchill to FDR, and then society runs this algorithm,

361
00:27:46,905 --> 00:27:48,905
and it says, okay, Churchill wins,

362
00:27:48,905 --> 00:27:51,905
and then if you reran it again with everyone swapping

363
00:27:51,905 --> 00:27:55,905
how they voted for Stalin versus Hitler in place three and four,

364
00:27:55,905 --> 00:28:01,905
it shouldn't then give you the answer of FDR instead of Churchill,

365
00:28:01,905 --> 00:28:04,905
because the ones at the top didn't change.

366
00:28:04,905 --> 00:28:07,905
But unfortunately, you have to break one of these,

367
00:28:07,905 --> 00:28:09,905
because for math reasons.

368
00:28:09,905 --> 00:28:12,905
So it's an unsolvable problem.

369
00:28:12,905 --> 00:28:16,905
That's why in the United States,

370
00:28:16,905 --> 00:28:18,905
we have this problem with splitting the vote,

371
00:28:18,905 --> 00:28:22,905
and where you have this thing where votes cast for a lost cause

372
00:28:22,905 --> 00:28:24,905
do not contribute at all.

373
00:28:24,905 --> 00:28:26,905
But you guys have proportional representation,

374
00:28:26,905 --> 00:28:28,905
which is a slightly different conversation,

375
00:28:28,905 --> 00:28:31,905
but there is still a version of this theorem

376
00:28:31,905 --> 00:28:33,905
that applies to proportional representation,

377
00:28:33,905 --> 00:28:37,905
because it applies to all preference aggregation.

378
00:28:37,905 --> 00:28:42,905
So that is part of why we always have these two people,

379
00:28:42,905 --> 00:28:45,905
because you need to vote for someone that you think other people will vote for.

380
00:28:45,905 --> 00:28:46,905
Yes?

381
00:28:46,905 --> 00:28:50,905
There are several voting algorithms, like instant runoff.

382
00:28:50,905 --> 00:28:51,905
Yes.

383
00:28:51,905 --> 00:28:54,905
That go along with that problem.

384
00:28:54,905 --> 00:28:57,905
No, they go some way, but they do not actually go.

385
00:28:57,905 --> 00:29:00,905
Yeah, the question was about instant runoff and these other things.

386
00:29:00,905 --> 00:29:02,905
You rank a bunch of people.

387
00:29:02,905 --> 00:29:07,905
But there are still incentives to put the electable people

388
00:29:07,905 --> 00:29:09,905
to misrepresent your own vote.

389
00:29:09,905 --> 00:29:14,905
So even though it does help you with the votes ranked

390
00:29:14,905 --> 00:29:16,905
like two, three, four, five, six,

391
00:29:16,905 --> 00:29:19,905
it does not actually help with the person that you put number one,

392
00:29:19,905 --> 00:29:22,905
because if the other people break with the scheme,

393
00:29:22,905 --> 00:29:25,905
or the intended scheme, and lie and put all theirs number one,

394
00:29:25,905 --> 00:29:27,905
then the other people, the other team,

395
00:29:27,905 --> 00:29:32,905
will still win by breaking the scheme.

396
00:29:32,905 --> 00:29:38,905
So actually it is unsolvable in this direction, I think.

397
00:29:38,905 --> 00:29:39,905
Okay.

398
00:29:39,905 --> 00:29:43,905
Do you have to go to decidability before you can have a fair election?

399
00:29:43,905 --> 00:29:44,905
Yeah.

400
00:29:44,905 --> 00:29:50,905
The issue is that we want to make it one-dimensional,

401
00:29:50,905 --> 00:29:52,905
one dimension of competition,

402
00:29:52,905 --> 00:29:56,905
so that it is more like the restaurant case before.

403
00:29:56,905 --> 00:30:00,905
So we are going to need a way to get all this information out

404
00:30:00,905 --> 00:30:02,905
because of rational ignorance,

405
00:30:02,905 --> 00:30:07,905
and then we are going to need to get rid of this electability problem

406
00:30:07,905 --> 00:30:09,905
that someone can split the vote

407
00:30:09,905 --> 00:30:15,905
and screw with the preference aggregation process of the ballot system itself,

408
00:30:15,905 --> 00:30:19,905
and then we need a way to stop this from being captured

409
00:30:19,905 --> 00:30:23,905
by anyone who is not a voter.

410
00:30:23,905 --> 00:30:27,905
So we want only the voters to be able to influence this process.

411
00:30:27,905 --> 00:30:31,905
And that third thing is going to be the blockchain and Bitcoin

412
00:30:31,905 --> 00:30:33,905
and all that other stuff.

413
00:30:33,905 --> 00:30:36,905
So what this brings me to what the solution is,

414
00:30:36,905 --> 00:30:39,905
and it is these conditional prediction markets.

415
00:30:39,905 --> 00:30:40,905
This is Hal Finney.

416
00:30:40,905 --> 00:30:45,905
He wrote this on Robin Hanson's blog, overcomingbias.com.

417
00:30:45,905 --> 00:30:48,905
And this is Ralph Merkle,

418
00:30:48,905 --> 00:30:51,905
and he said there are only a few people alive today worth listening to.

419
00:30:51,905 --> 00:30:52,905
Robin Hanson is one of them.

420
00:30:52,905 --> 00:30:53,905
So this is Robin Hanson's blog,

421
00:30:53,905 --> 00:30:59,905
and this is Hal Finney posting in July 29, 2008 on this topic.

422
00:30:59,905 --> 00:31:03,905
So that's just some social proof and appeal to authority

423
00:31:04,905 --> 00:31:06,905
to just kind of maybe liven the mood.

424
00:31:06,905 --> 00:31:08,905
It's kind of a long talk.

425
00:31:08,905 --> 00:31:09,905
You have to ignore that.

426
00:31:09,905 --> 00:31:14,905
Okay, so I'm going to explain what a normal prediction market is,

427
00:31:14,905 --> 00:31:17,905
and then I'm going to explain what these conditional prediction markets are.

428
00:31:17,905 --> 00:31:20,905
So a normal prediction market is this thing.

429
00:31:20,905 --> 00:31:22,905
It's often called an event derivative.

430
00:31:22,905 --> 00:31:25,905
So it pays some amount if the event happens,

431
00:31:25,905 --> 00:31:27,905
and it pays nothing otherwise.

432
00:31:27,905 --> 00:31:32,905
And this particular example, which I got from Intrade before I was closed down,

433
00:31:32,905 --> 00:31:38,905
is about whether or not 2012 will be the warmest year on record.

434
00:31:38,905 --> 00:31:42,905
And when this was created, it was January 2011.

435
00:31:42,905 --> 00:31:44,905
So this was a long time ago.

436
00:31:44,905 --> 00:31:46,905
This slide is from the past.

437
00:31:46,905 --> 00:31:49,905
And the market was created in January 11,

438
00:31:49,905 --> 00:31:54,905
and then it traded at these prices until the end of 2012, of course,

439
00:31:54,905 --> 00:31:58,905
when it settled in January 13, which is the end there.

440
00:31:59,905 --> 00:32:04,905
It was worth $100 if 2012 was the warmest year on record.

441
00:32:04,905 --> 00:32:07,905
So you could bet on this before 2012 even started,

442
00:32:07,905 --> 00:32:10,905
which is this first half here on the horizontal axis.

443
00:32:10,905 --> 00:32:15,905
And then in 2012, it became slowly clear that it was not in the cards,

444
00:32:15,905 --> 00:32:19,905
and that even though 2014 was the warmest year on record at the time,

445
00:32:19,905 --> 00:32:22,905
2012 was not going to be.

446
00:32:22,905 --> 00:32:24,905
And so it ended up being worth zero,

447
00:32:24,905 --> 00:32:26,905
because it's worth zero if it doesn't happen,

448
00:32:26,905 --> 00:32:28,905
and it's worth $100 if it does happen.

449
00:32:28,905 --> 00:32:33,905
Now the issue here is that you can reference the current price

450
00:32:33,905 --> 00:32:35,905
as a kind of probability,

451
00:32:35,905 --> 00:32:40,905
a society's aggregated knowledge about the likelihood of the event.

452
00:32:40,905 --> 00:32:45,905
So it traded from this initial price, which is about $40,

453
00:32:45,905 --> 00:32:48,905
for something that's worth $100 if the thing happens.

454
00:32:48,905 --> 00:32:53,905
And then the price went up to $50, and then it went all the way down.

455
00:32:54,905 --> 00:32:59,905
To $0 when it ultimately did not happen.

456
00:33:02,905 --> 00:33:04,905
A lot of people take issue with this.

457
00:33:04,905 --> 00:33:07,905
Some people have some hang-ups when it comes to markets.

458
00:33:09,905 --> 00:33:13,905
Some free market people say that free markets not only make people rich,

459
00:33:13,905 --> 00:33:16,905
but they're also really good for society,

460
00:33:16,905 --> 00:33:18,905
and people have a hang-up about that.

461
00:33:18,905 --> 00:33:20,905
And now I'm kind of also saying that they're also smart,

462
00:33:20,905 --> 00:33:23,905
which is almost too much for some people.

463
00:33:23,905 --> 00:33:28,905
But the issue here is that it's a kind of proof by contradiction,

464
00:33:28,905 --> 00:33:33,905
which is that if this price were not really truly representative

465
00:33:33,905 --> 00:33:36,905
of the objective knowledge that was existing in society,

466
00:33:36,905 --> 00:33:40,905
then someone or some fund or something,

467
00:33:40,905 --> 00:33:43,905
some group of people or some kind of co-op or something,

468
00:33:43,905 --> 00:33:47,905
there's some feedback process in place to correct that error.

469
00:33:48,905 --> 00:33:51,905
If this market were on something that everyone thought

470
00:33:51,905 --> 00:33:54,905
was definitely going to happen in 2012,

471
00:33:54,905 --> 00:33:56,905
then the prices would not have started at $40,

472
00:33:56,905 --> 00:33:59,905
and then they would not have declined to $0.

473
00:33:59,905 --> 00:34:02,905
They would have started at $70 or $80,

474
00:34:02,905 --> 00:34:04,905
and they would have gone up to $100.

475
00:34:04,905 --> 00:34:07,905
So nothing is perfect, and nothing can see the future,

476
00:34:07,905 --> 00:34:11,905
which is another hang-up I think people have on prediction markets.

477
00:34:12,905 --> 00:34:16,905
They're like, you know, what kind of magic is this

478
00:34:16,905 --> 00:34:18,905
that you're telling me you know everything

479
00:34:18,905 --> 00:34:20,905
about society's knowledge or something?

480
00:34:20,905 --> 00:34:23,905
So if you did a coin flip,

481
00:34:23,905 --> 00:34:27,905
the prices would have to be 50-50 until the coin landed

482
00:34:27,905 --> 00:34:30,905
because these only allow you to aggregate knowledge

483
00:34:30,905 --> 00:34:33,905
that actually exists in society.

484
00:34:33,905 --> 00:34:36,905
And so they're not like some magical perfect thing.

485
00:34:36,905 --> 00:34:39,905
Another last hang-up before I go on to the next slide

486
00:34:39,905 --> 00:34:42,905
is that there's a lot of ignorance of probability.

487
00:34:42,905 --> 00:34:47,905
So something is trading at $0.25 on the dollar.

488
00:34:47,905 --> 00:34:50,905
It should happen exactly one out of every four times,

489
00:34:50,905 --> 00:34:53,905
but I have a million friends.

490
00:34:53,905 --> 00:34:56,905
Something is trading at $0.25 on the dollar

491
00:34:56,905 --> 00:34:58,905
in a prediction market world,

492
00:34:58,905 --> 00:35:01,905
and I say, well, it's probably not going to happen.

493
00:35:01,905 --> 00:35:03,905
It's only got a 25% chance of happening.

494
00:35:03,905 --> 00:35:06,905
And then most of the time, three out of four times,

495
00:35:06,905 --> 00:35:08,905
it doesn't happen, but then when the fourth time happens,

496
00:35:08,905 --> 00:35:10,905
you get a bunch of e-mails from people telling you

497
00:35:10,905 --> 00:35:12,905
that the market was wrong or something,

498
00:35:12,905 --> 00:35:16,905
when in reality it was 100% perfectly calibrated.

499
00:35:16,905 --> 00:35:18,905
But this is kind of the pitch,

500
00:35:18,905 --> 00:35:22,905
that there's a tight connection here

501
00:35:22,905 --> 00:35:26,905
between the price and the likelihood of the event,

502
00:35:26,905 --> 00:35:28,905
and if you can soup this up,

503
00:35:28,905 --> 00:35:30,905
if you can soup this up,

504
00:35:30,905 --> 00:35:32,905
then by adding some more anonymity

505
00:35:32,905 --> 00:35:34,905
and making sure the costs are very low

506
00:35:34,905 --> 00:35:36,905
and making sure there's not a lot of friction,

507
00:35:36,905 --> 00:35:41,905
you can try to get a greater fine-tuning of this feedback

508
00:35:42,905 --> 00:35:44,905
so that the connection is more and more tight.

509
00:35:44,905 --> 00:35:47,905
And unlike the voting I mentioned before,

510
00:35:47,905 --> 00:35:50,905
the scaling is very good

511
00:35:50,905 --> 00:35:52,905
because the more people who know about a market,

512
00:35:52,905 --> 00:35:54,905
the more people will kind of glance at it

513
00:35:54,905 --> 00:35:56,905
and see if it looks totally wrong to them,

514
00:35:56,905 --> 00:35:58,905
and the more liquidity will be in the market,

515
00:35:58,905 --> 00:36:03,905
and so the transaction frictions will be even lower.

516
00:36:03,905 --> 00:36:05,905
And so, I mean, I already explained this. It's okay.

517
00:36:05,905 --> 00:36:07,905
Now, the interesting thing about this is that

518
00:36:07,905 --> 00:36:09,905
not only is this information very accurate,

519
00:36:09,905 --> 00:36:11,905
but it's common,

520
00:36:11,905 --> 00:36:14,905
so not only everyone gets the same information,

521
00:36:14,905 --> 00:36:16,905
but everyone knows that everyone else got it,

522
00:36:16,905 --> 00:36:18,905
so it's very much unlike the example

523
00:36:18,905 --> 00:36:20,905
where everyone hated the restaurant

524
00:36:20,905 --> 00:36:23,905
but didn't know that everyone hated the restaurant.

525
00:36:23,905 --> 00:36:26,905
With this, everyone knows the exact same information,

526
00:36:26,905 --> 00:36:28,905
and so you have this, it's what's called

527
00:36:28,905 --> 00:36:31,905
common knowledge in the field of game theory.

528
00:36:31,905 --> 00:36:34,905
It's like private knowledge is something that only you know,

529
00:36:34,905 --> 00:36:37,905
mutual knowledge is like a very dramatic kind of knowledge

530
00:36:37,905 --> 00:36:39,905
that two people know something,

531
00:36:39,905 --> 00:36:41,905
but they don't know that the other knows it,

532
00:36:41,905 --> 00:36:44,905
so like if you oversaw someone murder someone,

533
00:36:44,905 --> 00:36:47,905
and they know they murdered the person, and you know,

534
00:36:47,905 --> 00:36:49,905
but you saw from far away,

535
00:36:49,905 --> 00:36:51,905
and then you see them later getting coffee or something,

536
00:36:51,905 --> 00:36:53,905
and it's a very dramatic kind of,

537
00:36:53,905 --> 00:36:55,905
because you're just both, you're pretending,

538
00:36:55,905 --> 00:36:57,905
and he doesn't know that you saw, and all this other stuff.

539
00:36:57,905 --> 00:36:59,905
But common knowledge is when everyone knows.

540
00:36:59,905 --> 00:37:01,905
It's like this presentation right here

541
00:37:01,905 --> 00:37:03,905
where you are all watching, and you all know

542
00:37:03,905 --> 00:37:05,905
that everyone else can see it.

543
00:37:05,905 --> 00:37:08,905
So these kind of do all the research,

544
00:37:08,905 --> 00:37:10,905
and they do all the persuading,

545
00:37:10,905 --> 00:37:12,905
which is my, you know,

546
00:37:12,905 --> 00:37:15,905
it's kind of a tough sell to say this one little thing

547
00:37:15,905 --> 00:37:18,905
just gets you up on these little estimates for things.

548
00:37:18,905 --> 00:37:21,905
Something very hotly debated in the United States

549
00:37:21,905 --> 00:37:24,905
is global warming, and a lot of people don't,

550
00:37:24,905 --> 00:37:27,905
you know, they're not ready to

551
00:37:27,905 --> 00:37:29,905
leave whatever trench they're in,

552
00:37:29,905 --> 00:37:31,905
of which there are many, but

553
00:37:31,905 --> 00:37:34,905
this is kind of the pitch for this part of it.

554
00:37:34,905 --> 00:37:37,905
But as I said, there's more to it,

555
00:37:37,905 --> 00:37:40,905
the conditional aspect that I'll get to later.

556
00:37:40,905 --> 00:37:42,905
But yeah, I think, okay,

557
00:37:42,905 --> 00:37:44,905
the other thing, I'll mention this,

558
00:37:44,905 --> 00:37:46,905
this forces a clear definition,

559
00:37:46,905 --> 00:37:49,905
so actually the example that I used over here is,

560
00:37:49,905 --> 00:37:51,905
there was this, you could click into like a little thing

561
00:37:51,905 --> 00:37:53,905
and see what the fine print was, and it said,

562
00:37:53,905 --> 00:37:55,905
okay, we're going to look at this NASA satellite,

563
00:37:55,905 --> 00:37:58,905
and it's going to do global surface temperature anomaly,

564
00:37:58,905 --> 00:38:02,905
and NASA is going to publish this on this URL or whatever.

565
00:38:02,905 --> 00:38:04,905
So that's a lot better,

566
00:38:04,905 --> 00:38:06,905
because a lot of people use disagreements

567
00:38:06,905 --> 00:38:08,905
just for social reasons, just to posture,

568
00:38:08,905 --> 00:38:10,905
and just to kind of like have a lot of fun,

569
00:38:10,905 --> 00:38:12,905
and they don't, they're not,

570
00:38:12,905 --> 00:38:15,905
caring about the issue a lot is not really the point.

571
00:38:15,905 --> 00:38:18,905
It's just like, you know, being loyal to your,

572
00:38:18,905 --> 00:38:21,905
you know, tribe or party or whatever,

573
00:38:21,905 --> 00:38:23,905
and so definitely in the United States,

574
00:38:23,905 --> 00:38:25,905
global warming is just kind of a flag

575
00:38:25,905 --> 00:38:28,905
for people to join their respective camps,

576
00:38:28,905 --> 00:38:31,905
and no one has really even given any thought

577
00:38:31,905 --> 00:38:33,905
to like defining exactly what would it mean,

578
00:38:33,905 --> 00:38:35,905
which is in this case, specific satellite,

579
00:38:35,905 --> 00:38:37,905
specific website, specific everything,

580
00:38:37,905 --> 00:38:40,905
specific metric, global surface temperature anomaly,

581
00:38:40,905 --> 00:38:43,905
and a very specific thing here,

582
00:38:43,905 --> 00:38:45,905
2012 being the warmest year on record,

583
00:38:45,905 --> 00:38:47,905
and so you can argue about

584
00:38:47,905 --> 00:38:49,905
whether or not that represents global warming,

585
00:38:49,905 --> 00:38:51,905
which is a good conversation to have,

586
00:38:51,905 --> 00:38:53,905
but then you sidestep a lot of this stuff

587
00:38:53,905 --> 00:38:55,905
where people are just talking on and on and on

588
00:38:55,905 --> 00:38:57,905
about abstractions, you know,

589
00:38:57,905 --> 00:38:59,905
abstractions like global warming

590
00:38:59,905 --> 00:39:01,905
or gun control or whatever,

591
00:39:01,905 --> 00:39:03,905
stuff that's very vague,

592
00:39:03,905 --> 00:39:05,905
which you don't want at all.

593
00:39:05,905 --> 00:39:07,905
Okay, so yes, I say at all times

594
00:39:07,905 --> 00:39:09,905
everyone agrees with the price,

595
00:39:09,905 --> 00:39:11,905
which is again a very extreme claim,

596
00:39:11,905 --> 00:39:13,905
but the proof is by contradiction

597
00:39:13,905 --> 00:39:15,905
because you can make expected profits

598
00:39:15,905 --> 00:39:17,905
if you don't agree with it.

599
00:39:17,905 --> 00:39:19,905
Now there's a lot of, you know, risk aversion,

600
00:39:19,905 --> 00:39:21,905
and if the transaction costs are low,

601
00:39:21,905 --> 00:39:23,905
and if people really trust that this software works,

602
00:39:23,905 --> 00:39:25,905
and blah, blah, blah, blah, blah.

603
00:39:25,905 --> 00:39:27,905
But I think those problems are very fixable over time.

604
00:39:27,905 --> 00:39:29,905
You can accumulate trust in the institution

605
00:39:29,905 --> 00:39:31,905
and make it more user-friendly

606
00:39:31,905 --> 00:39:33,905
and all these other things.

607
00:39:33,905 --> 00:39:35,905
So this is the, you know, kind of a claim

608
00:39:35,905 --> 00:39:37,905
that this price is society's aggregated knowledge,

609
00:39:37,905 --> 00:39:39,905
and so that's kind of a neat thing here.

610
00:39:39,905 --> 00:39:41,905
Okay, so now I have to talk about

611
00:39:41,905 --> 00:39:43,905
how does this end up actually solving any problems

612
00:39:43,905 --> 00:39:45,905
we're trying to solve,

613
00:39:45,905 --> 00:39:47,905
and how does this end up

614
00:39:47,905 --> 00:39:49,905
because so far we've only been able to do

615
00:39:49,905 --> 00:39:51,905
like one little thing

616
00:39:51,905 --> 00:39:53,905
with one little market that's yes or no,

617
00:39:53,905 --> 00:39:55,905
and this is not, you know,

618
00:39:55,905 --> 00:39:57,905
how Finney wrote the blog post

619
00:39:57,905 --> 00:39:59,905
about conditional markets.

620
00:39:59,905 --> 00:40:01,905
Conditional markets are where the magic really happens

621
00:40:01,905 --> 00:40:03,905
because we don't just want to know, like,

622
00:40:03,905 --> 00:40:05,905
will someone win some sporting game

623
00:40:05,905 --> 00:40:07,905
or will someone win an election?

624
00:40:07,905 --> 00:40:09,905
We really want to know, like,

625
00:40:09,905 --> 00:40:11,905
we want relationships between things.

626
00:40:11,905 --> 00:40:13,905
So, you know, we have to think about

627
00:40:13,905 --> 00:40:15,905
how does this end up actually solving any problems

628
00:40:15,905 --> 00:40:17,905
between things.

629
00:40:17,905 --> 00:40:19,905
So this is probably too confusing

630
00:40:19,905 --> 00:40:21,905
because it's an earlier version,

631
00:40:21,905 --> 00:40:23,905
but I have a cool example here

632
00:40:23,905 --> 00:40:25,905
that I will now attempt

633
00:40:25,905 --> 00:40:27,905
because I constantly refine this talk

634
00:40:27,905 --> 00:40:29,905
and I'm very unhappy with it, to be honest with you,

635
00:40:29,905 --> 00:40:31,905
but I think it's getting better.

636
00:40:31,905 --> 00:40:33,905
So now I'm trying to teach.

637
00:40:33,905 --> 00:40:35,905
There's no probability.

638
00:40:35,905 --> 00:40:37,905
So here's the thing is that

639
00:40:37,905 --> 00:40:39,905
before we had a market

640
00:40:39,905 --> 00:40:41,905
that had a state

641
00:40:41,905 --> 00:40:43,905
and it was worth $100

642
00:40:43,905 --> 00:40:45,905
if global warming were worth zero otherwise.

643
00:40:45,905 --> 00:40:47,905
And that chart,

644
00:40:47,905 --> 00:40:49,905
there was only one price that was plotted

645
00:40:49,905 --> 00:40:51,905
because it would have been a total waste

646
00:40:51,905 --> 00:40:53,905
of chart space to plot both of the lines

647
00:40:53,905 --> 00:40:55,905
because they would have been

648
00:40:55,905 --> 00:40:57,905
a perfect mirror of each other.

649
00:40:57,905 --> 00:40:59,905
So that would have been a total waste of time.

650
00:40:59,905 --> 00:41:01,905
You could have said the opposite one

651
00:41:01,905 --> 00:41:03,905
would have been worth $100

652
00:41:03,905 --> 00:41:05,905
if global warming was not the warmest year on record

653
00:41:05,905 --> 00:41:07,905
and that one would have started at $60

654
00:41:07,905 --> 00:41:09,905
and it would have gone up to $100.

655
00:41:09,905 --> 00:41:11,905
But that would have been a total waste of ink.

656
00:41:11,905 --> 00:41:13,905
But now it's not going to be a perfect mirror

657
00:41:13,905 --> 00:41:15,905
because we're going to add lots more states

658
00:41:15,905 --> 00:41:17,905
and it's going to be very complicated.

659
00:41:17,905 --> 00:41:19,905
So it'll make for cooler graphs

660
00:41:19,905 --> 00:41:21,905
that you'll see in a second.

661
00:41:21,905 --> 00:41:23,905
But I first have to explain

662
00:41:23,905 --> 00:41:25,905
what on earth is going on in these graphs.

663
00:41:25,905 --> 00:41:27,905
So what I have first is an example

664
00:41:27,905 --> 00:41:29,905
where you flip two coins

665
00:41:29,905 --> 00:41:31,905
and you are betting on

666
00:41:31,905 --> 00:41:33,905
the total outcome of what happened there.

667
00:41:33,905 --> 00:41:35,905
And you say one coin could be heads

668
00:41:35,905 --> 00:41:37,905
and one coin could be tails

669
00:41:37,905 --> 00:41:39,905
and the second coin could be heads

670
00:41:39,905 --> 00:41:41,905
and the third coin could be tails.

671
00:41:41,905 --> 00:41:43,905
And as a result

672
00:41:43,905 --> 00:41:45,905
they should all be trading

673
00:41:45,905 --> 00:41:47,905
at 25 percentage points

674
00:41:47,905 --> 00:41:49,905
because each state is equally likely.

675
00:41:49,905 --> 00:41:51,905
The first coin could be heads

676
00:41:51,905 --> 00:41:53,905
the second coin could be heads or tails

677
00:41:53,905 --> 00:41:55,905
or the first coin could be tails

678
00:41:55,905 --> 00:41:57,905
and the second could be heads or tails.

679
00:41:57,905 --> 00:41:59,905
Now I want to imagine

680
00:41:59,905 --> 00:42:01,905
something really weird

681
00:42:01,905 --> 00:42:03,905
that would never happen in the real world

682
00:42:03,905 --> 00:42:05,905
but will probably maybe help me explain this concept

683
00:42:05,905 --> 00:42:07,905
which is that

684
00:42:07,905 --> 00:42:09,905
what happens if you accidentally

685
00:42:09,905 --> 00:42:11,905
when you were setting this market up or something

686
00:42:11,905 --> 00:42:13,905
like you register all the coins

687
00:42:13,905 --> 00:42:15,905
as different like coin number 3

688
00:42:15,905 --> 00:42:17,905
coin toss number 4, coin toss number 5

689
00:42:17,905 --> 00:42:19,905
and you accidentally have the market be set up

690
00:42:19,905 --> 00:42:21,905
so that both of these axes

691
00:42:21,905 --> 00:42:23,905
are the same coin.

692
00:42:23,905 --> 00:42:25,905
And I would like

693
00:42:25,905 --> 00:42:27,905
this is ask the audience time

694
00:42:27,905 --> 00:42:29,905
because this part is confusing

695
00:42:29,905 --> 00:42:31,905
and I get questions about it later

696
00:42:31,905 --> 00:42:33,905
because I don't explain it very well.

697
00:42:33,905 --> 00:42:35,905
But the question is

698
00:42:35,905 --> 00:42:37,905
if it's the same coin

699
00:42:37,905 --> 00:42:39,905
what should the prices end up being?

700
00:42:39,905 --> 00:42:41,905
Very quickly, yes.

701
00:42:41,905 --> 00:42:43,905
0.15, 0, 0, 0.15

702
00:42:43,905 --> 00:42:45,905
Yes, exactly.

703
00:42:45,905 --> 00:42:47,905
I didn't set up this thing yet

704
00:42:47,905 --> 00:42:49,905
but I'll advance to the next slide.

705
00:42:49,905 --> 00:42:51,905
Maybe you could say this is two dice

706
00:42:51,905 --> 00:42:53,905
and you roll the two dice

707
00:42:53,905 --> 00:42:55,905
and let's say accidentally someone sets it up

708
00:42:55,905 --> 00:42:57,905
so it's the same dice

709
00:42:57,905 --> 00:42:59,905
and whatever the green dice is, is this

710
00:42:59,905 --> 00:43:01,905
and whatever the green dice is, is this

711
00:43:01,905 --> 00:43:03,905
as a mistake.

712
00:43:03,905 --> 00:43:05,905
This would never happen in the real world

713
00:43:05,905 --> 00:43:07,905
but if that did happen

714
00:43:07,905 --> 00:43:09,905
this is just slightly more advanced version of this

715
00:43:09,905 --> 00:43:11,905
but what do you think

716
00:43:11,905 --> 00:43:13,905
these prices would be

717
00:43:13,905 --> 00:43:15,905
moments after you set up

718
00:43:15,905 --> 00:43:17,905
the market.

719
00:43:17,905 --> 00:43:19,905
Zeros on the diagonal.

720
00:43:19,905 --> 00:43:21,905
Yes, precisely, yes.

721
00:43:21,905 --> 00:43:23,905
So what I'm trying to explain here

722
00:43:23,905 --> 00:43:25,905
is that

723
00:43:25,905 --> 00:43:27,905
as the events become more

724
00:43:27,905 --> 00:43:29,905
objectively related to each other

725
00:43:29,905 --> 00:43:31,905
these prices kind of do this weird thing

726
00:43:31,905 --> 00:43:33,905
where they kind of bunch up

727
00:43:33,905 --> 00:43:35,905
in these observable ways

728
00:43:35,905 --> 00:43:37,905
and this is a cool thing

729
00:43:37,905 --> 00:43:39,905
and if you do this

730
00:43:39,905 --> 00:43:41,905
you get everything that you had before

731
00:43:41,905 --> 00:43:43,905
you get what's called the marginal probability

732
00:43:43,905 --> 00:43:45,905
just by adding

733
00:43:45,905 --> 00:43:47,905
so you get just from addition

734
00:43:47,905 --> 00:43:49,905
let me put this back up

735
00:43:49,905 --> 00:43:51,905
you can see that if you just add across the rows

736
00:43:51,905 --> 00:43:53,905
that it's 50-50 for the first coin

737
00:43:53,905 --> 00:43:55,905
and 50-50 for the second coin

738
00:43:55,905 --> 00:43:57,905
and 50-50 for the first coin

739
00:43:57,905 --> 00:43:59,905
and 50-50 for the first coin

740
00:43:59,905 --> 00:44:01,905
and 50-50 by the other dimension

741
00:44:01,905 --> 00:44:03,905
so you get everything

742
00:44:03,905 --> 00:44:05,905
you don't lose a single thing

743
00:44:05,905 --> 00:44:07,905
you get everything you got before

744
00:44:07,905 --> 00:44:09,905
but you get even more

745
00:44:09,905 --> 00:44:11,905
you get these conditional probabilities

746
00:44:11,905 --> 00:44:13,905
that reveal the relationship between things

747
00:44:13,905 --> 00:44:15,905
so you have professional speculators

748
00:44:15,905 --> 00:44:17,905
who research these issues

749
00:44:17,905 --> 00:44:19,905
and become specialists

750
00:44:19,905 --> 00:44:21,905
on whether or not Brexit

751
00:44:21,905 --> 00:44:23,905
would affect whatever

752
00:44:23,905 --> 00:44:25,905
so this is no surrender of sovereignty here

753
00:44:25,905 --> 00:44:27,905
this is just giving advice to the voter

754
00:44:27,905 --> 00:44:29,905
so the voter can still decide what they want

755
00:44:29,905 --> 00:44:31,905
but we're getting very very very high

756
00:44:31,905 --> 00:44:33,905
manipulation resistant

757
00:44:33,905 --> 00:44:35,905
knowledge emerging here

758
00:44:35,905 --> 00:44:37,905
so this is again an old slide

759
00:44:37,905 --> 00:44:39,905
because I made it before

760
00:44:39,905 --> 00:44:41,905
believe it or not I've been doing this

761
00:44:41,905 --> 00:44:43,905
project for a while and this image is very old

762
00:44:43,905 --> 00:44:45,905
in particular but

763
00:44:45,905 --> 00:44:47,905
this is, if you passed

764
00:44:47,905 --> 00:44:49,905
so I'm imagining, you have to imagine that we're in like

765
00:44:49,905 --> 00:44:51,905
2014 or something for this

766
00:44:51,905 --> 00:44:53,905
and so I'm saying next year if we pass some law

767
00:44:53,905 --> 00:44:55,905
will it actually affect

768
00:44:55,905 --> 00:44:57,905
our odds of getting

769
00:44:57,905 --> 00:44:59,905
will the law actually

770
00:44:59,905 --> 00:45:01,905
change our likelihood of getting

771
00:45:01,905 --> 00:45:03,905
the warmest year on record and you do that by just

772
00:45:03,905 --> 00:45:05,905
comparing the ratio of the prices

773
00:45:05,905 --> 00:45:07,905
when you don't do the law

774
00:45:07,905 --> 00:45:09,905
to the ratio of the prices

775
00:45:09,905 --> 00:45:11,905
if you do pass the law

776
00:45:11,905 --> 00:45:13,905
and I have an even more

777
00:45:13,905 --> 00:45:15,905
crazy example here

778
00:45:15,905 --> 00:45:17,905
so this is the guy who's the CEO

779
00:45:17,905 --> 00:45:19,905
of General Electric

780
00:45:19,905 --> 00:45:21,905
a gigantic corporation

781
00:45:21,905 --> 00:45:23,905
in the United States and

782
00:45:23,905 --> 00:45:25,905
I just want to kind of quickly just try to get

783
00:45:25,905 --> 00:45:27,905
through to what's going on because this problem

784
00:45:27,905 --> 00:45:29,905
is not just

785
00:45:29,905 --> 00:45:31,905
you know, the problems I've

786
00:45:31,905 --> 00:45:33,905
mentioned apply equally to actually

787
00:45:33,905 --> 00:45:35,905
CEOs because they're also elected

788
00:45:35,905 --> 00:45:37,905
the board of directors is elected

789
00:45:37,905 --> 00:45:39,905
or at least the bylaws

790
00:45:39,905 --> 00:45:41,905
can determine how that exactly happens but

791
00:45:41,905 --> 00:45:43,905
mostly the shareholders

792
00:45:43,905 --> 00:45:45,905
vote and elect the board of directors

793
00:45:45,905 --> 00:45:47,905
who hires the CEO so this problem

794
00:45:47,905 --> 00:45:49,905
is present in corporations as well

795
00:45:49,905 --> 00:45:51,905
as in the public

796
00:45:51,905 --> 00:45:53,905
sector it's the same voting problem

797
00:45:53,905 --> 00:45:55,905
so I wanted to kind of make that

798
00:45:55,905 --> 00:45:57,905
with this example and it's easier

799
00:45:57,905 --> 00:45:59,905
in the corporation case because you have a clear

800
00:45:59,905 --> 00:46:01,905
metric determining the metric is a little

801
00:46:01,905 --> 00:46:03,905
difficult like what is good

802
00:46:03,905 --> 00:46:05,905
should we reorganize society based on how

803
00:46:05,905 --> 00:46:07,905
likely it is to get global warming down like

804
00:46:07,905 --> 00:46:09,905
probably not reorganize

805
00:46:09,905 --> 00:46:11,905
everything in society you know there's things that

806
00:46:11,905 --> 00:46:13,905
are more important than that like you know

807
00:46:13,905 --> 00:46:15,905
not dying or something so

808
00:46:15,905 --> 00:46:17,905
but with the corporation it's easier

809
00:46:17,905 --> 00:46:19,905
because the stock price

810
00:46:19,905 --> 00:46:21,905
inclusive of any

811
00:46:21,905 --> 00:46:23,905
lawsuits or any kind of

812
00:46:23,905 --> 00:46:25,905
you know being sued for fraud or something

813
00:46:25,905 --> 00:46:27,905
inclusive of all that

814
00:46:27,905 --> 00:46:29,905
is really just the metric and so

815
00:46:29,905 --> 00:46:31,905
I have here this example of this mirroring

816
00:46:31,905 --> 00:46:33,905
that I brought up before so if you just start

817
00:46:33,905 --> 00:46:35,905
with one dimension

818
00:46:35,905 --> 00:46:37,905
and you say will this guy stay as

819
00:46:37,905 --> 00:46:39,905
the CEO and you can see

820
00:46:39,905 --> 00:46:41,905
something weird is happening but these are

821
00:46:41,905 --> 00:46:43,905
the no and the yes are a mirror image

822
00:46:43,905 --> 00:46:45,905
which is why it's pointless to plot it

823
00:46:45,905 --> 00:46:47,905
because it's a waste

824
00:46:47,905 --> 00:46:49,905
of space because there's only one degree of freedom here

825
00:46:49,905 --> 00:46:51,905
so you should only have one line

826
00:46:51,905 --> 00:46:53,905
but then below I have a

827
00:46:53,905 --> 00:46:55,905
second separate market

828
00:46:55,905 --> 00:46:57,905
so it's completely separate and then I'm going to

829
00:46:57,905 --> 00:46:59,905
combine them later and that's the

830
00:46:59,905 --> 00:47:01,905
stock price of GE in the future

831
00:47:01,905 --> 00:47:03,905
so this market is created today

832
00:47:03,905 --> 00:47:05,905
and I'm saying what's the price going to be on

833
00:47:05,905 --> 00:47:07,905
whatever this is May 2nd next

834
00:47:07,905 --> 00:47:09,905
year so and then we're going to

835
00:47:09,905 --> 00:47:11,905
move through time so you have to actually

836
00:47:11,905 --> 00:47:13,905
I think you know kind of

837
00:47:13,905 --> 00:47:15,905
turn your brain to the attention mode

838
00:47:15,905 --> 00:47:17,905
to kind of figure out this asset is

839
00:47:17,905 --> 00:47:19,905
going to be pay out something

840
00:47:19,905 --> 00:47:21,905
based on something that happens in the future

841
00:47:21,905 --> 00:47:23,905
and it's going to be created here

842
00:47:23,905 --> 00:47:25,905
and then we're moving through time

843
00:47:25,905 --> 00:47:27,905
to the present and this

844
00:47:27,905 --> 00:47:29,905
is the present moment which is

845
00:47:29,905 --> 00:47:31,905
after whatever this is 26

846
00:47:31,905 --> 00:47:33,905
weeks after this was created but still

847
00:47:33,905 --> 00:47:35,905
before this

848
00:47:35,905 --> 00:47:37,905
event has happened

849
00:47:37,905 --> 00:47:39,905
but it's a basic law of

850
00:47:39,905 --> 00:47:41,905
finance that whatever

851
00:47:41,905 --> 00:47:43,905
the future value of something should equal the present

852
00:47:43,905 --> 00:47:45,905
value of something minus modulo

853
00:47:45,905 --> 00:47:47,905
the interest rate the time value of money

854
00:47:47,905 --> 00:47:49,905
and so this thing should

855
00:47:49,905 --> 00:47:51,905
your guess at what the stock price will be

856
00:47:51,905 --> 00:47:53,905
next year should really just be

857
00:47:53,905 --> 00:47:55,905
the stock price today so they're

858
00:47:55,905 --> 00:47:57,905
kind of similar

859
00:47:57,905 --> 00:47:59,905
I hope that wasn't confusing because I actually can't

860
00:47:59,905 --> 00:48:01,905
remember why I brought it up at all but the point

861
00:48:01,905 --> 00:48:03,905
is the corporation

862
00:48:03,905 --> 00:48:05,905
stock price is going down and

863
00:48:05,905 --> 00:48:07,905
the CEO may stay or he may

864
00:48:07,905 --> 00:48:09,905
leave seems like there's tremendous

865
00:48:09,905 --> 00:48:11,905
uncertainty there

866
00:48:11,905 --> 00:48:13,905
and we're wondering like is the CEO

867
00:48:13,905 --> 00:48:15,905
jumping ship or should we fire him

868
00:48:15,905 --> 00:48:17,905
because things aren't going very well

869
00:48:18,254 --> 00:48:45,254
and this is a totally contrived example of course I just invented all these numbers
for the example but what I say now is you take the two separate markets and you
combine them and after you combine them you can then re-separate them by just adding
like I did with the coins and if you add this is 1 plus 3 the price of 1 plus 3 and
then it is the green line that you saw before and then it's mirror is the red which
is 1 minus that and then you can

870
00:48:45,254 --> 00:49:15,254
add 1 and 2 here and you get this blue line that you had before so you get
everything that you had before as I said but you get this cool thing where you get
these you get all the dimensions at once and then without making it too kind of like
you know I can explain the details but I think you know I'm worried about losing
people here but long story short this you just do some basic arithmetic you have
here the stock price if the CEO stays in

871
00:49:15,254 --> 00:49:45,134
red and the stock price if the CEO leaves in blue so you can just see as clear as
day that it's higher if he if he stays so in this contrived example that I just
invented you can see that the CEO is sort of doing a better job than whoever the
expected replacement is so I don't want to go into that but yes okay so you can use
this for basically everything that's a slide that I that I removed but it's a kind
of funny one but so

872
00:49:45,134 --> 00:50:14,574
it's like which CEO bakes the stock price the best and again this is the same is a
collective action problem that the shareholders face if you own shares of anything
you get stuff in the mail where it's like should I should you vote for this and I
just you throw it away in the trash immediately because you have no idea what
they're talking about and you have like no 0% of the company but this would actually
be helpful for you to know oh okay should I get rid of these board of director
people or should we take

873
00:50:14,574 --> 00:50:18,654
XYZ action or something and then you say oh which president would make the

874
00:50:18,654 --> 00:50:23,774
unemployment rate the lowest because we don't like unemployment or you could say

875
00:50:23,774 --> 00:50:28,694
which Fed policy helps GDP the most or will this law you know change crime or

876
00:50:28,694 --> 00:50:32,814
will any of these things happen so that's that's the cool thing and now I

877
00:50:32,814 --> 00:50:38,054
have to explain the multi-factor part which is really the same thing but and

878
00:50:38,054 --> 00:50:41,774
again I do apologize I planned on changing these two more European things

879
00:50:41,774 --> 00:50:45,534
but then I decided that that would be just too absurd for me to try to do and

880
00:50:45,534 --> 00:50:48,694
then I would probably say something like embarrassing or something like that so I

881
00:50:48,694 --> 00:50:53,174
kept the Democrat and the Republican are the two major parties in the United

882
00:50:53,174 --> 00:50:58,654
States and Elon Musk is a guy who makes electric cars and giant spaceships so

883
00:50:58,654 --> 00:51:04,454
hopefully you know who that is but I'm saying now that with this multi-dimensional

884
00:51:04,454 --> 00:51:09,414
trick you can erase this this this pernicious effect of why you always have

885
00:51:09,414 --> 00:51:15,414
the two terrible candidates because what you are going to do is condition it

886
00:51:15,414 --> 00:51:20,894
on the likelihood that they win to cancel out all of the the the badness

887
00:51:20,894 --> 00:51:25,174
associated with the fact that some people are unelectable for some mundane

888
00:51:25,174 --> 00:51:29,774
trivial reason like they misspoke or they they tripped or something it's so

889
00:51:29,774 --> 00:51:32,134
funny right like if someone trips something you imagine someone like

890
00:51:32,134 --> 00:51:35,934
walking out into a presidential debate and if you have like a primary debate

891
00:51:35,934 --> 00:51:38,654
and if someone like just tripped and fell over like even if they were the most

892
00:51:38,654 --> 00:51:42,654
even if they're the most the best person for the job on planet Earth they

893
00:51:42,654 --> 00:51:45,294
tripped and fell over and it was on camera it would probably just be that

894
00:51:45,294 --> 00:51:49,814
would be the end of it you know it kind of blows it boggles the mind that how

895
00:51:49,814 --> 00:51:53,534
do we even run society when the best person can be you know you wouldn't do

896
00:51:53,534 --> 00:51:57,174
that if you were hiring someone you're trying to hire someone to like head up

897
00:51:57,174 --> 00:52:02,614
the new division for your corporation or it is for your nonprofit or something

898
00:52:02,614 --> 00:52:06,134
just because they trip you would still probably offer them the job you just be

899
00:52:06,134 --> 00:52:10,214
like oh no you know you trip but in the in the when it comes to voting it's like

900
00:52:10,214 --> 00:52:13,814
if someone does any gaffe it's just over you know it doesn't make any sense to me

901
00:52:13,814 --> 00:52:21,934
but that's okay so the issue is we're going to say what is the okay will

902
00:52:21,934 --> 00:52:27,554
someone win is this dimension and it's split into four states instead of two

903
00:52:27,554 --> 00:52:32,094
just like I split the dice thing into six instead of two and then I'm gonna

904
00:52:32,094 --> 00:52:35,894
say is there a good economy in the United States where I defined it as

905
00:52:35,894 --> 00:52:39,814
less than 5% unemployment as measured by whatever you three years reported by

906
00:52:39,814 --> 00:52:42,294
the Fed so yeah you can have conversations about whether or not those

907
00:52:42,294 --> 00:52:46,214
I'm gonna get to that in a second which is like are those numbers coming in and

908
00:52:46,214 --> 00:52:50,374
he is a garbage in garbage out like what do we do with this these metrics coming

909
00:52:50,374 --> 00:52:54,214
in because that's it's just more again as I mentioned this this topic is very

910
00:52:54,214 --> 00:52:59,694
big and I can only talk about so many aspects of it at once but you know here

911
00:52:59,694 --> 00:53:05,734
we're gonna say does someone win and does or is the economy good and we see

912
00:53:05,734 --> 00:53:10,294
there's almost certainly going to be a terrible economy 0.99 and Elon Musk is

913
00:53:10,294 --> 00:53:15,574
almost almost certainly not going to be elected 0.009 but he's of the 0.009

914
00:53:15,574 --> 00:53:21,614
and of the 0.0095 we've got of being a good economy it all is kind of on the

915
00:53:21,614 --> 00:53:28,374
Elon Musk tile so to speak and so even though Elon Musk is the lost cause you

916
00:53:28,374 --> 00:53:32,774
can do some division here and display this to the user and again I'm gonna

917
00:53:32,774 --> 00:53:37,694
explain like as the user friendliness is obviously terrible at first but I

918
00:53:37,694 --> 00:53:44,694
think it will get there but you can say okay what's if you elect this person

919
00:53:44,694 --> 00:53:48,734
conditional on electing Elon Musk how likely is it that the economy will be

920
00:53:48,734 --> 00:53:55,334
good and then it will be 100% versus 0 and and then I have here that you can

921
00:53:55,334 --> 00:54:01,294
yes so now even though someone so you know that Elon Musk's it's kind of

922
00:54:01,294 --> 00:54:04,614
paradoxical because you measured the electability of Elon Musk and

923
00:54:04,614 --> 00:54:09,294
determined it to be very low but because you have measured it explicitly you can

924
00:54:09,294 --> 00:54:14,254
actually factor it out of the good economy thing just by dividing by it and

925
00:54:14,254 --> 00:54:19,894
so since the election these are all you know prices of stuff that hasn't happened

926
00:54:19,894 --> 00:54:24,054
yet in the future so you're the vision I have is that you people are lining up to

927
00:54:24,054 --> 00:54:27,454
vote and they've done no research and then they log on with their phone and

928
00:54:27,454 --> 00:54:31,294
they see something like two-week moving averages of these prices or something

929
00:54:31,294 --> 00:54:34,494
and they just think well you know I've done no research at all but I really

930
00:54:34,494 --> 00:54:38,614
should vote and I kind of just look up this blue section and they're just like

931
00:54:38,614 --> 00:54:43,454
well you know so that's the vision is that people would stop caring because

932
00:54:43,454 --> 00:54:47,654
they'd be like well you know it says the Elon Musk probably won't win but these

933
00:54:47,654 --> 00:54:51,894
guys are not helping me out at all and you know what the heck does this does

934
00:54:51,894 --> 00:54:57,374
this system know anyway because you know we've elected unelectable people

935
00:54:57,374 --> 00:55:03,814
before you know Donald Trump etc so so then you I would get the voter you the

936
00:55:03,814 --> 00:55:08,174
voter would be like you know come on like who cares about that it says .009

937
00:55:08,174 --> 00:55:12,294
you know I've got nothing to lose this is my vote and then they just say well

938
00:55:12,294 --> 00:55:16,654
uh-huh let's you know all the voters waiting in line are like okay well you

939
00:55:16,654 --> 00:55:20,574
know why not let's just throw let's put the vote on this guy because we know

940
00:55:20,574 --> 00:55:24,254
that he's got the 100 he's actually got the highest number so he actually kind

941
00:55:24,254 --> 00:55:28,894
of becomes electable again and so this is better for all the people who are

942
00:55:28,894 --> 00:55:33,774
uninformed or angry because the less they know the better because they they

943
00:55:33,774 --> 00:55:39,174
should ignore all the weird media campaigns and all of the other attempts

944
00:55:39,174 --> 00:55:43,494
to kind of influence their vote and just focus on things like this we can have a

945
00:55:43,494 --> 00:55:48,694
conversation about manipulation but you know yeah the more the more everything

946
00:55:48,694 --> 00:55:51,974
is just confusing for them the more they might just fall back on relying to this

947
00:55:51,974 --> 00:55:57,374
and again I want to remind you that 71% of Americans don't even have a clue who

948
00:55:57,374 --> 00:56:01,934
their representative is and most of them don't vote at all especially in the

949
00:56:01,934 --> 00:56:06,534
House of Representatives legislation kind of section so almost no one is

950
00:56:06,534 --> 00:56:10,614
voting and almost no one cares about this so those people are like a raw

951
00:56:10,614 --> 00:56:14,654
material like a resource you just get this uninformed vote to become informed

952
00:56:14,654 --> 00:56:20,534
and so it's just like that mentioned before at the restaurant where you had

953
00:56:20,534 --> 00:56:26,134
the three blue bars and one of them became green as the selection but I'd

954
00:56:26,134 --> 00:56:31,774
like to explain a little bit in more detail how we yes so here's the thing is

955
00:56:31,774 --> 00:56:35,974
what if you don't know if Elon Musk and or Peter Thiel is the best because you

956
00:56:35,974 --> 00:56:40,734
think they're both pretty great and maybe they're both unelectable but maybe

957
00:56:40,734 --> 00:56:43,934
and maybe Elon Musk is even more unelectable than Peter Thiel who was

958
00:56:43,934 --> 00:56:49,334
someone who is vaguely he was like a delegate in California and but also a

959
00:56:49,334 --> 00:56:53,654
kind of engineer type or something and then you know I never like Oprah you

960
00:56:53,654 --> 00:56:57,854
know and this is Kim Kardashian you know it's like whatever so you have all

961
00:56:57,854 --> 00:57:00,694
these up and then you think oh no the vote is going to be split between Elon

962
00:57:00,694 --> 00:57:05,894
Musk and Peter Thiel you know the smart person vote is going to be split and so

963
00:57:05,894 --> 00:57:09,254
then I have to go back to voting for the lesser of two evils because the vote is

964
00:57:09,254 --> 00:57:13,134
going to be split but you actually don't because only one of them has the highest

965
00:57:13,134 --> 00:57:17,294
number so the smart people will just say well you know I was pushing for

966
00:57:17,294 --> 00:57:23,734
Peter Thiel but 100% is bigger than 99% so so there it is and then you have the

967
00:57:23,734 --> 00:57:29,854
so then you've escaped electability and then just two more slides about yeah so

968
00:57:29,854 --> 00:57:32,974
there was my plan for like actually making this happen other than you know

969
00:57:32,974 --> 00:57:36,894
writing the code and everything I think you have to kind of ramp it up using

970
00:57:36,894 --> 00:57:41,214
things that people already care about like sports and whatever if you love

971
00:57:41,214 --> 00:57:45,174
sports betting you know probability of someone winning conditional on them

972
00:57:45,174 --> 00:57:48,694
being whatever two on-site kicks or something I don't know this is probably

973
00:57:48,694 --> 00:57:53,534
even the wrong I don't know anything about what would interest people but and

974
00:57:53,534 --> 00:57:57,974
then you just kind of need to do it for a couple years along a variety of things

975
00:57:57,974 --> 00:58:04,054
and this has already been done and the accuracy has been proven and so you just

976
00:58:04,054 --> 00:58:08,734
have to show it to people it's a little bit more of letting them experience kind

977
00:58:08,734 --> 00:58:14,414
of what's going on and again I think people hate the existing system so much

978
00:58:14,414 --> 00:58:18,374
in the United States and most don't vote at all so I think that you know

979
00:58:18,374 --> 00:58:23,814
there's kind of a there's a little niche here for it just kind of work its way in

980
00:58:23,814 --> 00:58:29,654
in fact on CNBC they now I remember 2004 they mentioned it once or twice and

981
00:58:29,654 --> 00:58:33,374
interviewed John Delaney the CEO of in trade where I got that slide I got that

982
00:58:33,374 --> 00:58:39,454
image from those prices and then he died on an unrelated Mount Everest accident

983
00:58:39,454 --> 00:58:46,854
which is real and threw the company into chaos long story but CNBC has done

984
00:58:46,854 --> 00:58:51,454
more and more America television network has done more and more coverage of these

985
00:58:51,454 --> 00:58:56,814
betting market prices even after in trade closed down after 2012 and they the

986
00:58:56,814 --> 00:59:00,894
new website election betting odds showed up and then they do bet fair bet fair

987
00:59:00,894 --> 00:59:04,214
here in Europe on them so I think actually people are kind of getting in

988
00:59:04,214 --> 00:59:08,014
more yeah this is again the thing is topic is so broad that you can't

989
00:59:08,014 --> 00:59:11,654
possibly talk about all of it at once but it's like what on earth are we

990
00:59:11,654 --> 00:59:14,454
gonna actually measure is it gonna be all unemployment rate what should it be

991
00:59:14,454 --> 00:59:18,294
it could be land value I think that's a good one or it could be anything that

992
00:59:18,294 --> 00:59:24,894
humans can measure kind of with a random poll of their kind of jurisdiction or

993
00:59:24,894 --> 00:59:29,214
something this is a very complicated topic and in itself without how to get

994
00:59:29,214 --> 00:59:35,094
that not captured but there are these ways sortition is like randomly picking

995
00:59:35,094 --> 00:59:40,094
names out of a hat and then you can you could have people you could call people

996
00:59:40,094 --> 00:59:43,494
the way we in the United States we do jury duty you just call people over and

997
00:59:43,494 --> 00:59:48,374
say force them to report on you know what their salary is with that what

998
00:59:48,374 --> 00:59:51,774
they were unemployed you know how much they like their government or whatever

999
00:59:51,774 --> 00:59:55,334
just kind of and then there are all these clever tricks random response and

1000
00:59:55,334 --> 00:59:59,934
things to like force some noise but yeah this is a whole topic in itself and we

1001
00:59:59,934 --> 01:00:03,454
don't have time for it so that that is just a crazy introduction to my project

1002
01:00:03,454 --> 01:00:08,814
and it's super weird but it is actually relatively so you can go to this the

1003
01:00:08,814 --> 01:00:13,414
site Bitcoin Hivemind calm and I had a bunch of people look at it including a

1004
01:00:13,414 --> 01:00:19,894
little bit you know including Adam back Peter Todd Andrew Paul straw Roger Ver so

1005
01:00:19,894 --> 01:00:25,014
a lot of people like this idea and it's sort of old so it's time for it to come

1006
01:00:25,014 --> 01:00:28,734
back you saw those the how Finney posts from like 2000 whatever doesn't for

1007
01:00:28,734 --> 01:00:33,454
doesn't eat or whatever that was so it's a it's kind of a cool idea so

1008
01:00:33,454 --> 01:00:36,894
that's the talk and if anyone has any ideas on how to improve the talk please

1009
01:00:36,894 --> 01:00:41,934
come up to me later because it's a very a lot to get across and I'm not sure

1010
01:00:41,934 --> 01:00:48,134
what the best way is so anyway so that is the talk so thank you very much

1011
01:00:48,134 --> 01:01:00,254
oh yes okay so if you go to the site you'll see we don't exactly have because

1012
01:01:00,254 --> 01:01:03,214
the weird thing is I designed it to be a sidechain of Bitcoin and sidechains of

1013
01:01:03,214 --> 01:01:09,454
Bitcoin are also held up but we have a kind of test net that it not only has it

1014
01:01:09,454 --> 01:01:13,814
has like a little GUI with a little little graphs but yeah there's a lot of

1015
01:01:13,814 --> 01:01:20,654
improvements still to be done but there is actually a there is actually a yeah

1016
01:01:20,654 --> 01:01:23,974
there's a version of this on the site and everyone should check it out and we

1017
01:01:23,974 --> 01:01:29,814
did rebase it to a recent Bitcoin core version like I don't know like eight

1018
01:01:29,814 --> 01:01:34,414
months ago so it's not hopelessly obsolete and I've been trying to keep

1019
01:01:34,414 --> 01:01:38,054
all these things synchronized so yeah there is some some stuff there's lots to

1020
01:01:38,054 --> 01:01:43,154
do still but we do there's like a minimum viable thing there that you can

1021
01:01:43,154 --> 01:01:50,234
run on testnet yes so we should we wait for the should we do the thing or should

1022
01:01:50,234 --> 01:01:58,594
we know oh there we go okay I think you made a very persuasive argument for the

1023
01:01:58,594 --> 01:02:02,554
use of prediction markets and I have a kind of philosophical question so I

1024
01:02:02,554 --> 01:02:05,714
think the general argument is that you know we can do this sort of collective

1025
01:02:05,714 --> 01:02:11,034
market mechanism and that humans are good sort of transducers of information

1026
01:02:11,034 --> 01:02:16,114
from the environment in this market mechanism which you then optimize the

1027
01:02:16,114 --> 01:02:21,594
decision so so I guess my question is like do you really just need humans at

1028
01:02:21,594 --> 01:02:27,514
all for the system what if you just had sensors or other other information like

1029
01:02:27,514 --> 01:02:32,274
there's no good questions just since this this can only give advice you know

1030
01:02:32,274 --> 01:02:36,674
it so it will optimize so what it what this will do is attempt to aggregate

1031
01:02:36,674 --> 01:02:41,274
society's information on whether or not someone elected will have some will

1032
01:02:41,274 --> 01:02:45,954
move some metric up or down more than someone else but there's no it says

1033
01:02:45,954 --> 01:02:50,954
nothing about values so maybe you want the unemployment rate to go up for some

1034
01:02:50,954 --> 01:02:55,874
reason you know it's like it says nothing at all about values and you could

1035
01:02:55,874 --> 01:02:58,754
have several different markets all on different things and people could

1036
01:02:58,754 --> 01:03:02,674
aggregate those values in different ways like someone might say oh I don't know

1037
01:03:02,674 --> 01:03:06,434
you know I really don't want someone could say brexit is gonna cost us a lot

1038
01:03:06,434 --> 01:03:10,594
of money as estimated by this market but someone else could say oh brexit is

1039
01:03:10,594 --> 01:03:15,874
going to result in XYZ immigration or something and everyone could be in

1040
01:03:15,874 --> 01:03:19,034
perfect agreement about all those numbers and then just say well I care

1041
01:03:19,034 --> 01:03:24,154
about immigration more than I care about how much money it will cost because the

1042
01:03:24,154 --> 01:03:28,754
UK is a rich whatever blah blah blah country so we can afford it so it says

1043
01:03:28,754 --> 01:03:34,754
nothing about values whatsoever have you read the book brave new world yes think

1044
01:03:34,754 --> 01:03:41,314
about it yes it's a good book in fact Adolphe Huxley's book partially

1045
01:03:41,314 --> 01:03:46,314
inspired Neil Postman's book I'm using ourselves to death which is the book

1046
01:03:46,314 --> 01:03:54,594
that I mentioned before which is about how media ecology has shaped the way we

1047
01:03:54,594 --> 01:03:59,194
run society including the elections where instead of having the three hour

1048
01:03:59,194 --> 01:04:03,714
four hour super debate I've just uninterrupted talking we have like

1049
01:04:03,714 --> 01:04:08,794
televised 30-second rebuttal or something like that but yeah it's good I

1050
01:04:08,794 --> 01:04:16,594
I think I'm more worried about 1984 the thesis of part of the ending of Neil

1051
01:04:16,594 --> 01:04:23,194
Postman's book was that 1984 threat was overemphasized and that the brave new

1052
01:04:23,194 --> 01:04:26,274
world threat was under emphasized and that we should have worried more about

1053
01:04:26,274 --> 01:04:31,954
brave new world but I don't know there may be a tie or something it's not it's

1054
01:04:31,954 --> 01:04:35,594
not looking very good but the 1984 ones is looking especially bad that stuff in

1055
01:04:35,594 --> 01:04:40,154
that stuff in China is just like that's just exactly as Phil Zimmerman said it's

1056
01:04:40,154 --> 01:04:45,594
just a lot just pure stasis forever and that is horrible I mean it's just that

1057
01:04:45,594 --> 01:04:49,834
will those people will be trapped until something it'll be exactly like the

1058
01:04:49,834 --> 01:04:53,234
Great Wall of China really where they'll be trapped into some other technological

1059
01:04:53,234 --> 01:04:58,794
force it's just just like I don't know what I know exactly how to describe what

1060
01:04:58,794 --> 01:05:03,314
I'm saying but until they are disrupted by the the fact that they are they've

1061
01:05:03,314 --> 01:05:10,034
been held back on new ideas forever so that's a sad thought yes does your

1062
01:05:10,034 --> 01:05:15,194
system deal with the fact that like unemployment like you say that these are

1063
01:05:15,194 --> 01:05:21,114
like factors which have long they have like delayed effects on society is that

1064
01:05:21,114 --> 01:05:27,474
calculated in there somehow yeah as I say yes that's a good question it it goes

1065
01:05:27,474 --> 01:05:31,594
into the what I said before about values so it's true that in the world of

1066
01:05:31,594 --> 01:05:35,434
economics it's often said like the stock market is like a leading indicator and

1067
01:05:35,434 --> 01:05:39,234
then there are these other things that are present indicators and then the

1068
01:05:39,234 --> 01:05:43,114
unemployment rate is a lagging indicator and so the economy goes bad the stock

1069
01:05:43,114 --> 01:05:46,434
market crashes immediately but then people don't start getting laid off for

1070
01:05:46,434 --> 01:05:50,874
like a couple months or something but this is just I mean if you wanted to

1071
01:05:50,874 --> 01:05:54,674
account for that you could just put the unemployment rate on different dates you

1072
01:05:54,674 --> 01:05:58,074
could say well what's the unemployment rate what's the stock market gonna be in

1073
01:05:58,074 --> 01:06:03,754
March and what's the unemployment rate going to be in September and try to do

1074
01:06:03,754 --> 01:06:08,714
harmonize them but there's no need for this project to take that type of thing

1075
01:06:08,714 --> 01:06:16,514
that the topic what the the topic of the market like what is it about that is

1076
01:06:16,514 --> 01:06:19,874
there's nothing to do with what exactly how the project works because you can

1077
01:06:19,874 --> 01:06:25,594
change that to try to do anything that you anything you like that it would be a

1078
01:06:25,594 --> 01:06:28,794
waste of time to do something that wasn't measurable like you know what is

1079
01:06:28,794 --> 01:06:33,274
your favorite color and I have to have things in there to kind of jam that

1080
01:06:33,274 --> 01:06:41,754
halfway through so make it impossible to work but if but yeah it's there is this

1081
01:06:41,754 --> 01:06:46,594
takes this project takes no position on what you would want to know so you may

1082
01:06:46,594 --> 01:06:49,434
want to know the unemployment rate or you may or you may not or you may not

1083
01:06:49,514 --> 01:06:54,754
think that the unemployment rate is interesting thoughts on selling your

1084
01:06:54,754 --> 01:06:59,194
vote so if somebody doesn't have like a like they don't value their vote but

1085
01:06:59,194 --> 01:07:04,514
maybe somebody else values I vote for a dollar yes there is a Robin Hansen the

1086
01:07:04,514 --> 01:07:08,714
blog that I quoted earlier that how Finney posted on he wrote something

1087
01:07:08,714 --> 01:07:12,674
about a quadratic voting which I think is unfortunately too complicated for

1088
01:07:12,674 --> 01:07:17,874
anyone to any to anyone to realistically switch from a current mode of voting to

1089
01:07:17,874 --> 01:07:22,674
that but there was a is a kind of weird thing where you could accumulate your

1090
01:07:22,674 --> 01:07:25,674
vote and it would square and then you could sell it or something but again

1091
01:07:25,674 --> 01:07:30,834
this is not this takes no position also on whatever society is really using

1092
01:07:30,834 --> 01:07:36,874
because this is about getting the equal amount of knowledge to everyone at once

1093
01:07:36,874 --> 01:07:39,874
in the form of the price so it's everyone sees they everyone sees the

1094
01:07:39,874 --> 01:07:44,634
same price so it kind of doesn't really matter as I said if you use proportional

1095
01:07:44,634 --> 01:07:49,754
representation or some other kind of voting this should work in all those

1096
01:07:49,754 --> 01:07:54,834
situations and in all those situations you still have the problem that the Ken

1097
01:07:54,834 --> 01:07:59,034
arrow Nobel Prize problem of that whatever you come up with it will

1098
01:07:59,034 --> 01:08:04,674
probably not really work because the idea there are two conflicting ideas

1099
01:08:04,674 --> 01:08:08,714
embedded in voting one is that everyone should be represented and the other is

1100
01:08:08,714 --> 01:08:12,794
that the outcome should be rational and you actually can only have one in the

1101
01:08:13,114 --> 01:08:17,394
mathematics of aggregating the vote so the one that most people pick is that

1102
01:08:17,394 --> 01:08:21,154
the outcome should be rational and that some people should but that's not what

1103
01:08:21,154 --> 01:08:24,434
you picked it with proportional representation but again it's a

1104
01:08:24,434 --> 01:08:32,074
different conversation but you the the idea is that the vote should be a guess

1105
01:08:32,074 --> 01:08:36,834
at what everyone will ultimately agree on they'll ultimately find to be

1106
01:08:36,834 --> 01:08:42,154
non-controversial so if you vote to get rid of slavery or you vote to establish

1107
01:08:42,154 --> 01:08:46,314
women's rights or something that you should that should what's really going

1108
01:08:46,314 --> 01:08:50,234
on is that you're kind of guessing at what everyone will agree was always the

1109
01:08:50,234 --> 01:08:56,234
right answer in the future so these schemes about changing the voting

1110
01:08:56,234 --> 01:09:00,034
mechanism or not I think they're not also not what this project is about

1111
01:09:00,034 --> 01:09:03,834
because this should work for just about any weird voting any weird aggregating

1112
01:09:03,834 --> 01:09:11,274
thing yes sorry that's okay I have a few questions but I have to pick one

1113
01:09:11,274 --> 01:09:16,514
right yeah okay I'm gonna pick one there's issues with liquidity because

1114
01:09:16,514 --> 01:09:21,194
the Elon Musk market would obviously have low liquidity yes liquidity yes

1115
01:09:21,194 --> 01:09:24,754
excellent question so that one that you want me to take them one at a time you

1116
01:09:24,754 --> 01:09:29,514
can probably like say a shit ton yes well there is okay so the answer is

1117
01:09:29,514 --> 01:09:32,594
probably just gonna be you try and get as much liquidity as you can no no no

1118
01:09:33,154 --> 01:09:38,394
you're right that but no that's not you're wrong about the answer but I can

1119
01:09:38,394 --> 01:09:42,594
talk a lot about how that's done because you are completely correct that why

1120
01:09:42,594 --> 01:09:45,874
would anyone buy this weird state in this giant dimensional thing and the

1121
01:09:45,874 --> 01:09:50,194
answer is that in this the market is done in a different way using something

1122
01:09:50,194 --> 01:09:54,554
called a market scoring rule that Robin Hanson actually invented for this for

1123
01:09:54,554 --> 01:09:58,394
this purpose and he's a physicist and I don't know why I brought that up because

1124
01:09:58,394 --> 01:10:05,994
that's a you know appeal to authority but and so is this but but he Robin

1125
01:10:05,994 --> 01:10:09,074
Hanson is the guy who runs his blog and he's over down here and he invented this

1126
01:10:09,074 --> 01:10:12,994
thing where actually no matter how many states the market has you can always

1127
01:10:12,994 --> 01:10:16,514
trade against the market itself and all you don't actually need a counterparty

1128
01:10:16,514 --> 01:10:23,554
so someone gives me someone when the person who creates the market puts a

1129
01:10:23,554 --> 01:10:27,514
little bit of money up as a kind of sacrifice and then it's kind of like you

1130
01:10:27,514 --> 01:10:32,754
have a dish and you have it on balanced on a needle and people put money on and

1131
01:10:32,754 --> 01:10:36,754
the dish kind of tilts around and the angle of the dish is actually what

1132
01:10:36,754 --> 01:10:39,954
determines the prices and how many different marbles people get me one

1133
01:10:39,954 --> 01:10:44,034
state is fixed we only need one underwriting yeah you only need is a

1134
01:10:44,034 --> 01:10:47,954
little bit of money in the center at first and then everyone can show up and

1135
01:10:47,954 --> 01:10:51,994
make trades even without any counterparty at all so the liquidity can

1136
01:10:51,994 --> 01:10:55,834
never fall to zero so it will always be possible to update the market price

1137
01:10:55,834 --> 01:10:59,514
which means it will always be possible for someone who notices an error in the

1138
01:10:59,514 --> 01:11:04,834
entire array of prices to update it so it's actually 100% solved and more to

1139
01:11:04,834 --> 01:11:07,994
the point it's actually much easier in blockchain technology to just have this

1140
01:11:07,994 --> 01:11:12,434
thing just be a sequence of atomic updates in a line so it's actually

1141
01:11:12,434 --> 01:11:15,234
solves several different problems at once but as I mentioned the projects

1142
01:11:15,234 --> 01:11:17,994
kind of big and you can only talk about so many things at a time but you're

1143
01:11:18,354 --> 01:11:22,194
correct that would be fixes it really well so that's cool

1144
01:11:22,194 --> 01:11:28,034
does I'll look into that that is very cool the other question was about that

1145
01:11:28,034 --> 01:11:33,074
efficient market efficiency arguments depend upon bad mark bad money

1146
01:11:33,074 --> 01:11:37,994
eventually leaving the market fully or it's like significantly yeah that's

1147
01:11:37,994 --> 01:11:44,354
another coming in faster than it's leaving and because even though this

1148
01:11:44,354 --> 01:11:48,834
system is like quite you know cyberspace and quite purist and so it

1149
01:11:48,834 --> 01:11:54,394
can hardly be uneconomical it can persist in being uneconomical but the

1150
01:11:54,394 --> 01:12:00,954
money that it's set in could still be secretly not performing the economic

1151
01:12:00,954 --> 01:12:06,594
function yes this is the manipulated prices question right yes I think so no

1152
01:12:07,034 --> 01:12:14,314
okay what was it then I didn't understand so if the general if the rest

1153
01:12:14,314 --> 01:12:20,434
of the world is not economically minded and efficient enough it there is no

1154
01:12:20,434 --> 01:12:24,274
guarantee that your prediction markets will have more good than bad money in it

1155
01:12:24,274 --> 01:12:28,714
and they'll consistently produce bad results due to the inefficiencies of

1156
01:12:28,714 --> 01:12:36,954
actual market like in practice okay so the the question is does I don't not

1157
01:12:36,954 --> 01:12:40,914
sure like what do you know I'm not sure what really good or bad money means but

1158
01:12:40,914 --> 01:12:46,874
I assume that you mean that yes okay well this again another thing I didn't

1159
01:12:46,874 --> 01:12:51,794
explain is how I solved the Oracle problem which is how this technology

1160
01:12:51,794 --> 01:12:54,954
becomes aware of actually what the unemployment rate is or who was actually

1161
01:12:55,074 --> 01:12:59,994
elected which is itself a document that I wrote and is on the front page of the

1162
01:12:59,994 --> 01:13:05,714
website Bitcoin hide my calm and is very long and filled with lots of you know

1163
01:13:05,714 --> 01:13:09,314
writing and math and things in there but I did not explain that at all because I

1164
01:13:09,314 --> 01:13:13,394
tried to give that talk and it's hopelessly boring and no one can I mean

1165
01:13:13,394 --> 01:13:16,194
I don't do a good job of explaining it I guess but in the software I have a

1166
01:13:16,194 --> 01:13:20,218
little simulator you can open it up and click on this little beaker

1167
01:13:20,218 --> 01:13:24,378
thing icon and it will show up and you can say oh what if the people try to

1168
01:13:24,378 --> 01:13:28,178
resolve the outcomes this way or that way and you can see if you can if you

1169
01:13:28,178 --> 01:13:32,538
can try to break any of my security assumptions but I don't think was that

1170
01:13:32,538 --> 01:13:34,858
what you were asking about is how does the protocol figure out what actually

1171
01:13:34,858 --> 01:13:39,338
happened or you're saying that people will bet and they will be bad people who

1172
01:13:39,338 --> 01:13:46,298
will just bet randomly or something oh yes so this is the low knowledge I think

1173
01:13:46,298 --> 01:13:51,418
you're saying okay I think yes so yeah the market can only aggregate knowledge

1174
01:13:51,418 --> 01:13:57,058
that is present in in the minds of the traders so the traders don't know

1175
01:13:57,058 --> 01:14:02,098
anything it's useless so as I explained with the okay we'll get well let me

1176
01:14:02,098 --> 01:14:04,738
finish this one in a sec so it's like if you have a market prediction market on

1177
01:14:04,738 --> 01:14:10,018
the outcome of a fair coin toss then you cannot like gain any you it would be 50

1178
01:14:10,018 --> 01:14:15,498
50 the no one would pay 60 cents for one of these states because the other person

1179
01:14:15,498 --> 01:14:19,458
would buy the alternative state for 40 cents knowing that it's a fair coin and

1180
01:14:19,458 --> 01:14:24,498
that you know over the long run they will they will tend to do to do well so

1181
01:14:24,498 --> 01:14:29,858
but is that what you were saying or no yes

1182
01:14:38,658 --> 01:14:43,138
sure yes I agree

1183
01:14:46,498 --> 01:14:58,778
okay okay yes okay so the question is a kind of plutocracy question that says

1184
01:14:58,778 --> 01:15:03,858
what if most of the people who have money are have an agenda and then they

1185
01:15:03,858 --> 01:15:07,858
can manipulate this market but but this is not a plutocracy this is not like

1186
01:15:07,858 --> 01:15:13,498
$1 one vote this is like $1 risk to one vote so if you have a plutocracy you

1187
01:15:13,498 --> 01:15:17,378
could just keep it in something else and it would be safe but you know I

1188
01:15:17,378 --> 01:15:22,578
would be safely earn or whatever 4% return or whatever we're going on you

1189
01:15:22,578 --> 01:15:26,378
don't have to risk it in this and so if you did you would be risking that you

1190
01:15:26,378 --> 01:15:37,138
would lose it yeah yeah yes no but it is not though because think about this you

1191
01:15:37,138 --> 01:15:41,418
have to get the actual metric up so even if you manipulate the prices

1192
01:15:41,418 --> 01:15:45,738
tremendously you have to actually get in you you manipulate the prices

1193
01:15:45,738 --> 01:15:50,898
tremendously and no one has enough money to defeat you then you will get whoever

1194
01:15:50,898 --> 01:15:56,258
it is Donald Trump you'll get Donald Trump elected instead of Elon Musk yeah

1195
01:15:56,258 --> 01:15:58,618
but no I was let me explain no I think because I think I understand what you're

1196
01:15:58,618 --> 01:16:04,178
saying so it's like you will get you will get your way a few times but

1197
01:16:04,178 --> 01:16:09,218
inevitably you'll get Donald Trump and the unemployment rate with that the

1198
01:16:09,378 --> 01:16:13,338
price of that other thing will remain low or will be at zero instead of a

1199
01:16:13,338 --> 01:16:18,298
hundred so all the people who can cast counter bets and they say well there's

1200
01:16:18,298 --> 01:16:21,818
you see because it's mathematically identical it's like yes is one minus no

1201
01:16:21,818 --> 01:16:28,498
and so if so yeah there's you could buy if you thought Elon Musk would do well

1202
01:16:28,498 --> 01:16:33,058
you can buy this or you can buy everything other than you can just say

1203
01:16:33,058 --> 01:16:37,018
instead I know that whatever I know that Hillary Clinton will screw it up so I buy

1204
01:16:37,018 --> 01:16:41,818
all these other ones so as I know that there's not a chance that she'll do a

1205
01:16:41,818 --> 01:16:45,738
good job and so then you get something for less than $1 that ends up being a

1206
01:16:45,738 --> 01:16:49,858
dollar and so you still make money so even though you can if you have most of

1207
01:16:49,858 --> 01:16:55,018
the money you can manipulate the outcome the few people who are forecasting

1208
01:16:55,018 --> 01:17:00,178
accurately will be multiplying them their wealth by five times or each cycle

1209
01:17:00,178 --> 01:17:02,658
and so eventually they will be the ones that have all of the money and it will

1210
01:17:02,658 --> 01:17:10,618
create exponentially more money to manipulate the system yes

1211
01:17:14,258 --> 01:17:17,538
rational ignorance

1212
01:17:20,338 --> 01:17:23,338
sure

1213
01:17:32,698 --> 01:17:40,858
necessarily because this is again the traders are free to decline they can say

1214
01:17:40,858 --> 01:17:46,178
I can't improve on this array of prices so they can say even though I've got a

1215
01:17:46,178 --> 01:17:48,818
lot of money and I could move the prices to whatever I want they'd say well they

1216
01:17:48,818 --> 01:17:51,658
kind of look correct to me so instead I'm gonna do something else with my

1217
01:17:51,658 --> 01:17:56,098
money I'm gonna go play golf or something you know but if if there's an

1218
01:17:56,098 --> 01:17:59,978
error in the prices anyone can

1219
01:18:03,658 --> 01:18:10,138
yes no but you understand it's different from voting because if you feel very

1220
01:18:10,138 --> 01:18:13,858
strongly about something you can put more of your wealth into it you've been

1221
01:18:13,858 --> 01:18:17,458
maybe and you could even maybe take out a loan or something so this this is not

1222
01:18:17,458 --> 01:18:22,058
necessarily just measuring one vote per person each person is free to modulate

1223
01:18:22,058 --> 01:18:26,058
their influence in the market by how confident they are so a lot of people

1224
01:18:26,058 --> 01:18:29,418
take positions that then then they take positions on global warming and then

1225
01:18:29,418 --> 01:18:34,578
they won't bet in the global market so what did they really believe yes okay

1226
01:18:34,578 --> 01:18:41,098
let's go more questions yes Emanuel Wallerstein he's a world systems

1227
01:18:41,098 --> 01:18:53,378
theorist and what he describes is that we are in what is called a bifurcation

1228
01:18:53,378 --> 01:19:00,418
point which is in the last 100 years or so it was very easy there was a very

1229
01:19:00,418 --> 01:19:07,178
linear kind of course for the destination of events for example like

1230
01:19:07,178 --> 01:19:12,978
when you see the price on a market is going up it's very easy to predict where

1231
01:19:12,978 --> 01:19:21,658
it's going to go next but when in a in a system there are too many when it when

1232
01:19:21,658 --> 01:19:25,858
there are too many when a system isn't able to accurately when a kind of

1233
01:19:25,858 --> 01:19:31,618
worldview isn't able to completely account for all of the variables or all

1234
01:19:31,618 --> 01:19:39,578
of the phenomena those kind of factors can can build up in a system and what

1235
01:19:39,578 --> 01:19:45,458
happened is a complex system reach a kind of tipping point where things start

1236
01:19:45,458 --> 01:19:50,978
to cascade and when a system so for example like when the price starts to go

1237
01:19:51,018 --> 01:19:55,818
down and become very difficult to predict of what rate is falling how far

1238
01:19:55,818 --> 01:20:04,618
it's going to drop and until what level is going to go down and the entire I

1239
01:20:04,618 --> 01:20:15,218
think the entire premise of what you're talking about is the people voted for

1240
01:20:15,218 --> 01:20:21,778
Donald Trump because they were misinformed and if they were informed

1241
01:20:21,778 --> 01:20:29,818
maybe they could make a better decision what if people what if what if people

1242
01:20:29,818 --> 01:20:36,778
what people were informed and still chose to vote for Donald Trump and what

1243
01:20:36,778 --> 01:20:44,778
if what if history shows us that this is the inevitable course of all human

1244
01:20:44,858 --> 01:20:51,738
civilizations and that there is no mechanism that is going to save us from

1245
01:20:51,738 --> 01:20:59,058
a process of renewal this inherent in civilization itself I think that's far

1246
01:20:59,058 --> 01:21:03,898
too pessimistic to make any sense pessimistic is the idea that there's no

1247
01:21:03,898 --> 01:21:08,378
the idea that there is no possibility of technological improvement I think is

1248
01:21:08,378 --> 01:21:15,418
very pessimistic well look at I don't know then if it's a rebirth and what's

1249
01:21:15,418 --> 01:21:19,018
the what is how is it different from what happened the last 30 years so the

1250
01:21:19,018 --> 01:21:26,778
last 70 years well would you say the Roman Empire breaking up and

1251
01:21:26,778 --> 01:21:33,258
Christianity taking over in Europe was progress or was that regression was the

1252
01:21:34,258 --> 01:21:41,858
slavery in Rome was that something that was was bad oh yeah well I think it's

1253
01:21:41,858 --> 01:21:46,498
clear now with hindsight it's interesting today we all everyone agrees

1254
01:21:46,498 --> 01:21:51,258
that slavery was wrong but in the past it was a hotly debated issue and a lot

1255
01:21:51,258 --> 01:21:56,018
of people took it for granted that slavery was the morally correct position

1256
01:21:56,018 --> 01:22:02,298
and that it was um so obviously that's not yeah well I mean the thing is so

1257
01:22:02,298 --> 01:22:05,898
it's a question about about knowledge though I mean it was originally taken a

1258
01:22:05,898 --> 01:22:08,858
lot of things were originally taken for granted and then we had to we had to

1259
01:22:08,858 --> 01:22:15,698
learn more and improve our you know our legal and governance technology you know

1260
01:22:15,698 --> 01:22:19,738
we had to improve our laws and they were improved many different times I mean

1261
01:22:19,738 --> 01:22:23,178
there was a time when there was no democracy whatsoever and then in ancient

1262
01:22:23,178 --> 01:22:28,498
Athens it was accidentally introduced by someone who was using it for selfish

1263
01:22:28,498 --> 01:22:34,098
reasons they just kind of kept doing it and then Rome had all kinds of weird

1264
01:22:34,098 --> 01:22:41,018
tinkering that they were doing with their setup but yeah I don't I don't know

1265
01:22:41,018 --> 01:22:46,018
if I would call the Enlightenment progress and I would call ancient Greece

1266
01:22:46,018 --> 01:22:53,138
progress but I don't know maybe I would call Florence in the 1300s but I don't

1267
01:22:53,138 --> 01:22:56,778
know if I would call it and everything after the Enlightenment in the West I

1268
01:22:56,778 --> 01:23:05,338
would call progress but I don't know if I recall anything else yeah it's funny

1269
01:23:05,338 --> 01:23:12,978
that people do they do that like the World Bank yeah yeah the the joke is I

1270
01:23:12,978 --> 01:23:15,938
should just measure energy consumption and that is something so you know very

1271
01:23:15,938 --> 01:23:19,538
from the stem department right that it's just like doesn't have any of these but

1272
01:23:19,538 --> 01:23:23,978
as I said it's not it this project takes no stance on values it just tries to

1273
01:23:23,978 --> 01:23:27,338
provide people with information and I think it's plausible that if people had

1274
01:23:27,338 --> 01:23:32,618
more information they might still have voted for Donald Trump yeah yeah no I

1275
01:23:32,618 --> 01:23:41,698
thought we have more questions that we can so last question then make this room

1276
01:23:41,698 --> 01:23:49,378
for something completely different yeah what was your what did you think about

1277
01:23:49,378 --> 01:23:54,778
the conference in Lisbon oh yeah yes I was at the Lisbon conference ah that's

1278
01:23:54,778 --> 01:23:58,298
a good one you really kind of well you know actually that's funny that you

1279
01:23:58,298 --> 01:24:00,898
mentioned that I had a great time at a conference in Lisbon building on

1280
01:24:00,898 --> 01:24:05,538
Bitcoin you know it was yesterday in the day before and it's great to see

1281
01:24:05,538 --> 01:24:09,538
everyone as always and people present a lot of interesting ideas but I kind of

1282
01:24:09,538 --> 01:24:14,098
wonder about something I have a very different view and I have no idea if how

1283
01:24:14,098 --> 01:24:18,138
many people share this view but for me there's a kind of drama in the space

1284
01:24:18,138 --> 01:24:22,458
that makes the conference is much more exciting and so they had scaling one in

1285
01:24:22,458 --> 01:24:28,218
Montreal in 2015 and that was just like so exciting and then stuff kind of gets

1286
01:24:28,218 --> 01:24:32,178
old and you know I got a scaling three wasn't as interesting as scaling one and

1287
01:24:32,178 --> 01:24:38,098
two and then it was kind of like breaking Bitcoin kind of tried to it had

1288
01:24:38,098 --> 01:24:42,098
some reason it had more drama and I think these things where all the

1289
01:24:42,098 --> 01:24:46,818
technical people meet have they go through these waves somehow I don't know

1290
01:24:46,818 --> 01:24:50,018
I can't quite explain it and I don't want to attribute it to anyone or

1291
01:24:50,018 --> 01:24:54,298
anything I don't know I don't understand it myself but I think sometimes there's

1292
01:24:54,298 --> 01:24:58,298
just a need for a conference and it just appears and it's a great success it's

1293
01:24:58,298 --> 01:25:03,698
like lightning in a bottle or something and then then people try to replicate it

1294
01:25:03,698 --> 01:25:11,098
and it's parts of it get you know just a little bit less interesting and sometimes

1295
01:25:11,098 --> 01:25:15,578
I feel that way you know there's a kind of capture you know that something is a

1296
01:25:15,658 --> 01:25:18,818
success and then everyone wants to affiliate with the success and then you

1297
01:25:18,818 --> 01:25:24,658
can have these weird cycles and so I enjoyed it a lot I loved especially that

1298
01:25:24,658 --> 01:25:28,378
I presented at the second half of day two and I loved that half not just

1299
01:25:28,378 --> 01:25:32,658
because I was in it but there was a lot of stuff that referenced my kind of

1300
01:25:32,658 --> 01:25:36,978
areas of interest including Sergio's Drivechain thing and also talks about

1301
01:25:36,978 --> 01:25:40,538
like super soft forks and these other kind of theoretical things that I'm

1302
01:25:40,538 --> 01:25:47,098
interested in and yeah so I I like that I I don't know it's like these these

1303
01:25:47,098 --> 01:25:50,418
conferences we you know we do so many conferences and it's very hard to like

1304
01:25:50,418 --> 01:25:56,338
put any of them into context but yeah I thought it was very neat I think the

1305
01:25:56,338 --> 01:26:05,298
community is very it's a very interesting you know I if you saw if you

1306
01:26:05,298 --> 01:26:09,218
watch my talk then it's a little bit of it is about a kind of my frustration

1307
01:26:09,218 --> 01:26:15,138
with how slow things are in terms of the reactivity to altcoins but a lot of

1308
01:26:15,138 --> 01:26:19,898
people are very proud of the fact that they are resistant to any mental

1309
01:26:19,898 --> 01:26:23,938
manipulation at all they say oh we're not gonna let that affect our decisions

1310
01:26:23,938 --> 01:26:27,698
our technical decisions which is admirable in a sense but that is a kind

1311
01:26:27,698 --> 01:26:30,738
of a weird difference of opinion and you do see a lot of that with these

1312
01:26:30,738 --> 01:26:35,618
conferences in the later waves a lot of very academic topics that are kind of

1313
01:26:35,618 --> 01:26:42,498
very you know a lot of math and debatable applicability it's time for a

1314
01:26:42,498 --> 01:26:48,058
closing statement oh yeah statement yeah that's like in an American television

1315
01:26:48,058 --> 01:26:53,418
presidential debate where the 30 seconds long and just have a lot of boring words

1316
01:26:53,418 --> 01:26:57,058
about freedom but yeah I mean I don't think I have a closing statement I'd

1317
01:26:57,058 --> 01:27:01,258
rather ask I'd rather answer more more questions I quite liked a description of

1318
01:27:01,258 --> 01:27:07,058
Lisbon is lightning in a bottle that was great it's a pun on the phrase and

1319
01:27:07,058 --> 01:27:09,778
lightning network

1320
01:27:14,658 --> 01:27:20,578
yeah and this project has to it's so bizarre and experimental that it has to

1321
01:27:20,578 --> 01:27:24,458
be a sidechain or an altcoin if it can't be a sidechain because it's just so

1322
01:27:24,458 --> 01:27:28,298
weird there's no possible way that it you could get people to buy into it and

1323
01:27:28,298 --> 01:27:34,418
you know it's kind of it's like it's a but yeah as I explained you know you'd

1324
01:27:34,418 --> 01:27:38,898
start this off as you started off as something experimental and people would

1325
01:27:38,898 --> 01:27:42,138
try it out to be a little bit like like lightning network and people could try

1326
01:27:42,138 --> 01:27:45,498
out a little bit of it at a time and then they could figure out what what

1327
01:27:45,498 --> 01:27:48,178
about it they like and what about they don't like and how to make it how to

1328
01:27:48,178 --> 01:27:51,418
make it better and then how to kind of scale it up and do all those other

1329
01:27:51,418 --> 01:27:57,178
things so I mean I have a lot of plans for all of those but I think you just

1330
01:27:57,178 --> 01:28:00,378
get people involved and then you do have a have a stuff about like you know

1331
01:28:00,378 --> 01:28:06,418
you try to get something where it's like how can you get I one thing I would do

1332
01:28:06,418 --> 01:28:11,058
is assemble a giant website with all of the entire track record of prediction

1333
01:28:11,058 --> 01:28:14,418
markets and I would try to compare it to other other people to try and demonstrate

1334
01:28:14,418 --> 01:28:20,298
how superior it is and I need I need something to crack that that statistical

1335
01:28:20,298 --> 01:28:24,178
ignorance that pervades the population where the prediction market says that

1336
01:28:24,178 --> 01:28:28,098
something has a 10% chance and then one out of every times you hear from your

1337
01:28:28,098 --> 01:28:31,378
friends an email that the prediction market said that thing wouldn't happen

1338
01:28:31,378 --> 01:28:35,018
but it happened prediction markets are wrong and I don't know what to do about

1339
01:28:35,018 --> 01:28:41,938
that you get those people to the people who are just we know numeral folks just

1340
01:28:41,938 --> 01:28:45,978
afraid of numbers and afraid of measurement and we still have depressing

1341
01:28:45,978 --> 01:28:50,098
one more talk this so she'll have some time for that yeah all right it won't

1342
01:28:50,098 --> 01:28:53,898
give Paul a grand applause here