1
00:00:00,000 --> 00:00:06,200
Once there's a fast quantum computer out there, sub 10 minutes,

2
00:00:06,980 --> 00:00:11,600
Bitcoin as we know it today is basically not usable because the coins will just be

3
00:00:11,600 --> 00:00:18,200
stolen by someone else once they try to move. As soon as there is evidence of clear logical

4
00:00:18,200 --> 00:00:25,100
qubit scaling, of progressive factorization, there's a very clear evidentiary standard here.

5
00:00:25,720 --> 00:00:29,380
And as soon as that evidentiary standard is met, we have to start doing stuff.

6
00:00:29,380 --> 00:00:37,140
I can make a very defensible, logically tight claim that there is no way a cryptographically

7
00:00:37,140 --> 00:00:39,520
relevant quantum computer will happen in the next 10 years.

8
00:00:39,940 --> 00:00:41,580
What I think, never.

9
00:00:41,720 --> 00:00:43,780
I think it is never going to happen.

10
00:00:46,960 --> 00:00:48,480
Brandon, how are you doing, man?

11
00:00:49,280 --> 00:00:50,200
I'm doing well.

12
00:00:50,240 --> 00:00:50,600
How are you, Danny?

13
00:00:51,680 --> 00:00:53,200
I am very good, thank you.

14
00:00:53,240 --> 00:00:54,560
I'm really looking forward to this conversation.

15
00:00:54,560 --> 00:00:59,380
So you've been kind of the quantum naysayer for a little bit now.

16
00:00:59,980 --> 00:01:03,900
And I've been going down this rabbit hole more and more over the last few days.

17
00:01:04,520 --> 00:01:08,780
Because maybe I'll frame this by like Nick Carter is someone who I know is controversial

18
00:01:08,780 --> 00:01:09,460
amongst Bitcoiners.

19
00:01:09,560 --> 00:01:10,440
I think he's really smart.

20
00:01:10,540 --> 00:01:12,060
I think he writes some amazing pieces on Bitcoin.

21
00:01:12,940 --> 00:01:16,780
And when I first read his sort of quantum series that he started doing a couple of weeks,

22
00:01:16,880 --> 00:01:21,980
well, maybe a month ago or so, it made me think that quantum might be quite close.

23
00:01:21,980 --> 00:01:33,000
But the more and more I've kind of listened to stuff that you've been talking about and done my own research on this, I'm starting to think that I don't want to call it a scam, but maybe it's much further away than we actually think.

24
00:01:33,380 --> 00:01:39,800
So I want to get into all of that with you today. But you want to start just by introducing yourself. This is the first time you've been on What Bitcoin Did.

25
00:01:40,520 --> 00:01:51,160
Yeah, sure. So yeah, I'm Brandon Black, root in code on the internet. I'm a software engineer. I guess I was a software engineer for 20 years, some of that time in management.

26
00:01:51,980 --> 00:01:54,920
About eight years of that in Bitcoin full time.

27
00:01:56,080 --> 00:02:07,360
In Bitcoin, probably my greatest accomplishments in Bitcoin were when I worked at BitGo, launching the first ever Taproot multi-sig wallet and the first ever Musig2 into production.

28
00:02:07,360 --> 00:02:10,520
So those are things I kind of did in the Bitcoin world.

29
00:02:10,680 --> 00:02:17,660
That was the way I got introduced to talking with people at Bitcoin conferences was doing research to do the Musig2 implementation.

30
00:02:17,660 --> 00:02:22,300
I met Jonas Nick and I was asking him questions about it and got into the implementation and

31
00:02:22,300 --> 00:02:28,080
everything there with him. So I guess that's like the broad strokes of me these days. This year,

32
00:02:28,120 --> 00:02:33,500
I'm launching a Bitcoin consultancy. I'm going to try to help people and businesses with their

33
00:02:33,500 --> 00:02:40,240
Bitcoin scripting needs and just general learning about Bitcoin. So if anyone needs me to come talk

34
00:02:40,240 --> 00:02:45,820
to them about Bitcoin, I do that. And yeah, that's what I'm up to. So when did you get into the

35
00:02:45,820 --> 00:02:51,040
quantum stuff? Because as like a Bitcoin developer, someone who, well, maybe not on like core, but

36
00:02:51,040 --> 00:02:54,760
you've been working on like Bitcoin projects, like when did quantum come into this? Because

37
00:02:54,760 --> 00:02:59,160
this FUD has been around for quite a long time now, but it's really ramped up over the last,

38
00:02:59,260 --> 00:03:07,160
I'd say, 12 months. Yeah, yeah. I would say I am naturally interested in all such things,

39
00:03:07,160 --> 00:03:15,080
whether it's cold fusion or quantum physics or, you know, all impossible things are naturally in

40
00:03:15,080 --> 00:03:19,080
my interest. I think that's how I got into Bitcoin in the first place is that same natural interest

41
00:03:19,080 --> 00:03:23,660
in these things. I was like, oh, magic internet money, that doesn't work until it turned out that

42
00:03:23,660 --> 00:03:30,020
it did. So I've naturally drawn to these kinds of things. And so I've been tracking the progress

43
00:03:30,020 --> 00:03:36,820
of quantum computing, gosh, for most of my professional career, just been watching it,

44
00:03:36,860 --> 00:03:43,720
seeing how things are going. So when the FUD comes up, I think I have a decent background

45
00:03:43,720 --> 00:03:51,240
to know, like, have we made a real step change that makes it now instead of 10 years to never

46
00:03:51,240 --> 00:03:55,700
to suddenly five years? And many people, of course, this year are claiming that we've made

47
00:03:55,700 --> 00:04:01,340
that transition from 10 years to never to within five years. And I just don't see any evidence of

48
00:04:01,340 --> 00:04:06,620
that. Okay. So I want to try and like set the table here because quantum computing to me,

49
00:04:06,620 --> 00:04:10,600
when I try and understand it, it still sounds like this Fugazi thing. Like, I think it's really

50
00:04:10,600 --> 00:04:14,840
hard to actually, like if someone asked me to explain what a quantum computer does,

51
00:04:15,400 --> 00:04:19,560
I'd really struggle. Can we start as basic as that? And can you explain what they're trying

52
00:04:19,560 --> 00:04:27,060
to do with quantum computers? Yeah. So the short idea of what quantum computing can do and what

53
00:04:27,060 --> 00:04:33,420
digital computers struggle to do is that a quantum computer can have its qubits, as they're called,

54
00:04:33,420 --> 00:04:37,720
represent a superposition of possible states

55
00:04:37,720 --> 00:04:39,720
as opposed to a single state.

56
00:04:40,140 --> 00:04:41,120
So in a digital computer,

57
00:04:41,460 --> 00:04:44,360
the binary digital computer is what we all use,

58
00:04:44,400 --> 00:04:45,540
is what we're talking over right now.

59
00:04:45,980 --> 00:04:48,880
Every bit of information is either true or false.

60
00:04:48,920 --> 00:04:51,160
It's one or zero, and that's every single thing.

61
00:04:51,600 --> 00:04:52,780
In a quantum computer,

62
00:04:53,320 --> 00:04:58,000
you can have bits that don't have a defined state yet.

63
00:04:58,460 --> 00:05:00,460
And then what you essentially do

64
00:05:00,460 --> 00:05:04,640
is you try to encode the properties of your real world problem,

65
00:05:04,800 --> 00:05:07,700
whether it's cryptographic problems or chemistry simulations,

66
00:05:07,860 --> 00:05:10,760
whatever the physical thing you want to work on

67
00:05:10,760 --> 00:05:13,140
into the possible states of the quantum computer.

68
00:05:14,020 --> 00:05:19,080
And then the properties of the quantum gate interactions

69
00:05:19,080 --> 00:05:23,660
within the qubits can make the correct answers

70
00:05:23,660 --> 00:05:25,300
have a lower energy state

71
00:05:25,300 --> 00:05:28,900
and therefore become more likely when you read out values.

72
00:05:28,900 --> 00:05:33,600
So then you collapse these possible states into real states and read out the values.

73
00:05:34,660 --> 00:05:38,200
And what you get is the low energy state.

74
00:05:38,320 --> 00:05:39,900
So they call this quantum annealing.

75
00:05:40,080 --> 00:05:46,400
And that's kind of the current state of quantum computing is that it can be used in physics simulations to do this quantum annealing,

76
00:05:46,560 --> 00:05:49,580
where you encode the state of a physical system into the quantum bits.

77
00:05:49,880 --> 00:05:55,860
You kind of shake them up, so to speak, and then you read out the low energy state they come to.

78
00:05:55,860 --> 00:06:02,960
And if you do that a few times, with high likelihood, you get the lowest energy state of that encoded system.

79
00:06:03,780 --> 00:06:09,760
So this is the magic of quantum computing and why it can do things that digital computers can't do.

80
00:06:10,060 --> 00:06:17,760
Because with digital computers, to do that same thing, you basically have to look through all the possible states that your system could get into to find the low energy one.

81
00:06:17,760 --> 00:06:35,820
But with quantum computers, because all of the possible states are encoded across the qubits, it can essentially simulate the combination of inputs and states that gets you to that low energy all in one go, instead of having to iterate through every possible value of every input to get there.

82
00:06:35,820 --> 00:06:50,760
So this very clever guy, Shore, forget his first name, came up with an algorithm that can apply those properties of finding a low energy option out of some encoding to cryptographic systems.

83
00:06:50,760 --> 00:07:05,880
And whether that's factoring or elliptic curves, the low energy state that we're looking for there in both cases relates to, well, in factoring, it's kind of a simpler problem to map.

84
00:07:06,080 --> 00:07:07,900
With factoring, you're like, oh, we got 15.

85
00:07:08,000 --> 00:07:09,220
Let me find three and five.

86
00:07:09,540 --> 00:07:18,320
And you can encode that so that the qubits resolve to lower energy state when you have inputs that multiply to the desired value, right?

87
00:07:18,940 --> 00:07:20,180
And so that's factoring.

88
00:07:20,180 --> 00:07:26,200
And then with elliptic curves, you have to map it into this thing called a period finding problem.

89
00:07:26,560 --> 00:07:31,300
But again, it's the same basic problem where the amount of energy encoded in the bit is lower

90
00:07:31,300 --> 00:07:36,840
when you have found the correct sequence of movements through the elliptic curve

91
00:07:36,840 --> 00:07:42,260
to encode the underlying value of the secret key.

92
00:07:43,540 --> 00:07:44,620
And that's roughly what we're doing.

93
00:07:44,620 --> 00:07:52,800
We're using this quantum annealing process to find the period over which the secret key translates in the elliptic curve realm.

94
00:07:53,460 --> 00:07:56,440
And if you can do that, then you can get the secret key out of a public key.

95
00:07:57,020 --> 00:08:03,540
All right. So I do definitely want to get into the issues with factoring and elliptic curves.

96
00:08:03,760 --> 00:08:07,280
But before we do, these computers are never going to be like a general computer, right?

97
00:08:07,280 --> 00:08:11,660
they're going to be specifically targeted to either like breaking encryption or other like

98
00:08:11,660 --> 00:08:18,300
like chemistry applications and things like that. That seems by far the most likely. Some people

99
00:08:18,300 --> 00:08:23,180
have proposed they could also be very useful for AI where essentially it comes out of the same

100
00:08:23,180 --> 00:08:28,600
thing where you would encode the properties of an LLM, let's say, into the quantum states. And

101
00:08:28,600 --> 00:08:33,940
instead of computing through a neural network, the output, you would quantum anneal to the output of

102
00:08:33,940 --> 00:08:39,300
the LLM. So maybe you could also apply it to something like that. But yes, they're never

103
00:08:39,300 --> 00:08:45,500
going to be able to do the hardcore, just sending bits across the internet type stuff, because

104
00:08:45,500 --> 00:08:52,560
there's a setup and a readout process, and it's all about probabilities. There's not a clear answer.

105
00:08:52,720 --> 00:08:55,100
And of course, for sending bits across the internet, you want to have a clear

106
00:08:55,100 --> 00:08:59,540
one or zero on each bit you send, and quantum computers aren't super well suited for that.

107
00:09:00,220 --> 00:09:03,060
Do you wish you could access cash without selling your Bitcoin?

108
00:09:03,060 --> 00:09:07,560
Well, Ledin makes that possible. They're the global leader in Bitcoin-backed lending and

109
00:09:07,560 --> 00:09:12,460
since 2018 they've issued over $9 billion in loans with a perfect record of protecting client

110
00:09:12,460 --> 00:09:17,580
assets. With Ledin, you get full custody loans with no credit checks or monthly repayments,

111
00:09:17,740 --> 00:09:24,020
just easy access to dollars without selling a single sat. As of July 1st, Ledin is Bitcoin only,

112
00:09:24,220 --> 00:09:28,880
meaning they exclusively offer Bitcoin-backed loans with all collateral held by Ledin directly

113
00:09:28,880 --> 00:09:33,120
or their funding partners. Your Bitcoin is never lent out to generate interest.

114
00:09:33,900 --> 00:09:38,240
I recently took out a loan with Ledin. The whole process was super easy. The application took me

115
00:09:38,240 --> 00:09:42,820
less than 15 minutes and in a few hours I had the dollars in my account. It was really smooth.

116
00:09:43,440 --> 00:09:48,860
So if you need cash but you don't want to sell Bitcoin, head over to ledin.io forward slash WBD

117
00:09:48,860 --> 00:09:54,900
and you'll get 0.25% off your first loan. That's ledin.io forward slash WBD.

118
00:09:54,900 --> 00:09:57,480
Privacy was never a priority for mobile networks.

119
00:09:57,860 --> 00:10:00,600
For companies like AT&T, T-Mobile, and Verizon,

120
00:10:00,920 --> 00:10:03,120
data collection and monetization is the default.

121
00:10:03,560 --> 00:10:04,640
But Cape is changing that.

122
00:10:05,180 --> 00:10:06,900
Cape is a premium US mobile carrier

123
00:10:06,900 --> 00:10:09,340
with nationwide coverage designed from the ground up

124
00:10:09,340 --> 00:10:11,300
with privacy and security at the core.

125
00:10:12,080 --> 00:10:12,840
When you sign up,

126
00:10:12,940 --> 00:10:14,880
Cape collects the absolute minimum data required,

127
00:10:15,120 --> 00:10:16,500
stores it for the shortest time possible,

128
00:10:16,680 --> 00:10:17,620
and never sells it.

129
00:10:18,080 --> 00:10:19,560
They also make you significantly harder

130
00:10:19,560 --> 00:10:20,860
to track at the network level

131
00:10:20,860 --> 00:10:22,520
and protect against SIM swap attacks,

132
00:10:22,520 --> 00:10:24,440
which are becoming one of the biggest security risks

133
00:10:24,440 --> 00:10:29,400
out there, especially for Bitcoiners. Cape's SIM swap protection is fundamentally different.

134
00:10:29,760 --> 00:10:34,180
Instead of usernames and passwords, your account is secured by a 24-word passphrase similar to how

135
00:10:34,180 --> 00:10:39,040
Bitcoin wallet works. No one can initiate a SIM swap or take control of your phone number except

136
00:10:39,040 --> 00:10:44,160
you. This isn't a burner phone or a workaround, it's a normal mobile service built properly.

137
00:10:44,760 --> 00:10:49,280
If you care about privacy and security, there is no better mobile carrier. To learn more and get

138
00:10:49,280 --> 00:10:56,200
33% off your first six months, head to cape.co slash WBD and use code WBD at checkout. That's

139
00:10:56,200 --> 00:11:02,160
cape.co slash WBD. With fiat money constantly debasing, wealth preservation isn't optional.

140
00:11:02,680 --> 00:11:07,180
That's why I recommend Swan Bitcoin, a team of dedicated Bitcoiners who work with families and

141
00:11:07,180 --> 00:11:12,180
businesses to build and secure generational wealth with Bitcoin. Strong relationships with clients

142
00:11:12,180 --> 00:11:16,480
are at the center of everything Swan does. A dedicated Swan private wealth representative,

143
00:11:16,480 --> 00:11:18,760
which is a real person that you can text and call,

144
00:11:19,100 --> 00:11:21,020
will help you build a Bitcoin wealth strategy

145
00:11:21,020 --> 00:11:23,980
using Swan's comprehensive platform of Bitcoin services,

146
00:11:24,520 --> 00:11:26,400
including tax-advantaged retirement accounts,

147
00:11:26,760 --> 00:11:29,540
advanced Bitcoin cold storage using collaborative self-custody,

148
00:11:30,000 --> 00:11:32,580
inheritance planning with both trust and entity accounts,

149
00:11:33,020 --> 00:11:35,500
tax loss harvesting, asset-backed loans, and more.

150
00:11:36,320 --> 00:11:38,700
Swan have helped over 100,000 clients since 2020,

151
00:11:39,180 --> 00:11:41,380
and if you're serious about acquiring and securing Bitcoin,

152
00:11:41,540 --> 00:11:42,260
I recommend Swan.

153
00:11:42,820 --> 00:11:45,840
Meet the team at swan.com forward slash WBD,

154
00:11:45,840 --> 00:11:54,320
which is swan.com forward slash wbd. Okay, so let's talk about factoring and elliptic curves,

155
00:11:54,420 --> 00:11:59,640
because factoring isn't really applicable to Bitcoin, but it is to other cryptographic systems.

156
00:11:59,680 --> 00:12:04,420
Is that right? Yeah, so RSA is the famous one that's based on large prime numbers that are

157
00:12:04,420 --> 00:12:09,500
difficult to factor. Okay, and so this is where, like, I don't want to call it a scam, but this is

158
00:12:09,500 --> 00:12:14,040
something that when I was looking at really shocked me, is that the highest number that a

159
00:12:14,040 --> 00:12:19,400
quantum computer can factor right now is 21, I believe. And the bit that really surprised me is

160
00:12:19,400 --> 00:12:24,940
they actually did that in 2012. So all the talk over the last few years of this really accelerating,

161
00:12:25,060 --> 00:12:28,120
they still can't factor a number bigger than that. Do you know why?

162
00:12:29,560 --> 00:12:35,840
So what I observe, and I think many others have made the same observations, I'm not alone on this,

163
00:12:35,840 --> 00:12:42,280
is that the basic properties of quantum of qubits

164
00:12:42,280 --> 00:12:45,420
are pretty well understood at this point.

165
00:12:45,600 --> 00:12:47,240
We can produce some qubits,

166
00:12:47,360 --> 00:12:48,920
we can map some state onto them,

167
00:12:48,940 --> 00:12:52,180
and we can resolve them into a desired state

168
00:12:52,180 --> 00:12:53,400
that we want to measure.

169
00:12:53,760 --> 00:12:55,940
And that enables factoring very small numbers,

170
00:12:56,040 --> 00:12:57,720
because if you can get a couple of qubits together,

171
00:12:58,080 --> 00:13:00,340
you can factor a small number using a quantum computer.

172
00:13:01,620 --> 00:13:03,640
And so what the quantum researchers have been doing

173
00:13:03,640 --> 00:13:04,640
over the last decade

174
00:13:04,640 --> 00:13:09,860
has been trying to make that scale to larger numbers of qubits.

175
00:13:10,760 --> 00:13:13,940
And that turns out to be very, very difficult.

176
00:13:14,200 --> 00:13:17,000
And we've seen hundreds of millions or billions of dollars

177
00:13:17,000 --> 00:13:19,340
poured into trying to get that to scale.

178
00:13:20,280 --> 00:13:23,160
And what seems to happen is that

179
00:13:23,160 --> 00:13:26,720
as you increase the number of qubits you're working with,

180
00:13:27,180 --> 00:13:31,680
the difficulty of correcting the errors in the qubits

181
00:13:31,680 --> 00:13:32,920
goes up exponentially.

182
00:13:32,920 --> 00:13:49,500
And so every time they come up with new technologies that seem to have lower error rates, and they've gotten much, much lower error rates on their individual qubits, but when they start trying to put more qubits together, they keep on seeing that the error rates rise exponentially as they go to more and more qubits.

183
00:13:50,280 --> 00:14:01,500
And so this is something we saw very clearly with the recent Microsoft Majorana announcement where they have built a very cool, very low error rate, single qubit device.

184
00:14:01,500 --> 00:14:07,560
and and so this is this is like oh we can manufacture lots of these single qubit devices

185
00:14:07,560 --> 00:14:12,380
and that's very cool and maybe there are applications for those but what they haven't

186
00:14:12,380 --> 00:14:16,260
demonstrated with all of these new announcements that have been made is no one has been able to

187
00:14:16,260 --> 00:14:20,820
demonstrate the ability to tie a whole bunch of qubits together on a single device and not have

188
00:14:20,820 --> 00:14:24,200
the error rates go up exponentially and so that's that's how we got to this state where

189
00:14:24,200 --> 00:14:28,920
they did this thing many years ago and it hasn't really advanced the the closest to an advancement

190
00:14:28,920 --> 00:14:35,080
we've seen was someone used a quantum computing device to solve a six-bit elliptic curve key. Now,

191
00:14:35,120 --> 00:14:42,420
six bits is ever so slightly larger than 21, but it's still like, I joked that literally a child

192
00:14:42,420 --> 00:14:47,600
can solve a six-bit elliptic curve key. It's not something that requires great computing power.

193
00:14:47,600 --> 00:14:50,340
So it's in a similar realm of those 21-bit factorizations.

194
00:14:51,880 --> 00:14:55,100
Let's give them their dues, right? Let's say they are making breakthroughs,

195
00:14:55,120 --> 00:14:58,220
things are going well in quantum computing world, they're putting billions of dollars in,

196
00:14:58,220 --> 00:15:12,018
What are they excited about Like why are they thinking that this is progressing so fast So a few things I seen that are interesting One is like I said better manufacturability

197
00:15:12,118 --> 00:15:14,018
of individual qubit devices.

198
00:15:14,098 --> 00:15:15,298
That was the Majorana announcement.

199
00:15:15,398 --> 00:15:20,138
So they went from, you know, it takes us like super customized,

200
00:15:20,238 --> 00:15:23,538
like hand-built, we get one working qubit out of 100 attempts,

201
00:15:23,638 --> 00:15:25,598
that level of difficulty of building qubits.

202
00:15:25,678 --> 00:15:27,238
With Majorana, they're like, this is a qubit,

203
00:15:27,298 --> 00:15:29,378
we can produce this qubit in a factory.

204
00:15:30,278 --> 00:15:34,598
Like everything that it takes to build this are things that we know how to put into factories.

205
00:15:34,837 --> 00:15:36,697
So that's a cool innovation right now.

206
00:15:36,738 --> 00:15:43,258
Instead of it being like hand building in labs, individual qubits and getting them to work sometimes, now it's manufacturable.

207
00:15:43,358 --> 00:15:44,998
So that's a big step.

208
00:15:45,478 --> 00:15:53,598
We've also seen pretty big breakthroughs in error correction where for a given error rate in the qubits,

209
00:15:53,598 --> 00:16:05,158
the classical computing or digital computing needed to correct the errors has become a couple order of magnitude faster.

210
00:16:05,678 --> 00:16:07,998
So that's another big breakthrough they've made.

211
00:16:09,298 --> 00:16:17,697
And then another one that we've seen is some folks have made innovations in what they've called gate-based quantum computing,

212
00:16:17,697 --> 00:16:20,618
where they build quantum gates

213
00:16:20,618 --> 00:16:22,678
that are a handful of qubits.

214
00:16:22,758 --> 00:16:23,658
I don't know the exact numbers,

215
00:16:23,758 --> 00:16:26,197
but like less than 10 qubits on a device.

216
00:16:26,438 --> 00:16:27,678
And then they interconnect them,

217
00:16:28,058 --> 00:16:30,158
which is kind of similar

218
00:16:30,158 --> 00:16:31,358
to how we build digital computers.

219
00:16:31,457 --> 00:16:32,038
In digital computers,

220
00:16:32,158 --> 00:16:33,598
we have essentially gates,

221
00:16:33,738 --> 00:16:35,058
you know, NAND, NOR, XOR,

222
00:16:35,158 --> 00:16:36,837
the binary operations.

223
00:16:38,138 --> 00:16:39,738
And we interconnect them on a chip.

224
00:16:40,158 --> 00:16:42,238
So maybe we can do something interesting

225
00:16:42,238 --> 00:16:43,798
with quantum computing in a similar way,

226
00:16:44,298 --> 00:16:45,258
where instead of trying to get

227
00:16:45,258 --> 00:16:47,158
a whole bunch of qubits all tied together,

228
00:16:47,158 --> 00:16:48,598
in a single device,

229
00:16:49,317 --> 00:16:50,398
you build gates

230
00:16:50,398 --> 00:16:52,778
where each gate is a kind of a quantum unit

231
00:16:52,778 --> 00:16:54,138
interconnected to other gates.

232
00:16:54,678 --> 00:16:56,717
And so we've seen developments in that area.

233
00:16:57,478 --> 00:16:59,717
The problem, as far as I understand it,

234
00:16:59,918 --> 00:17:02,158
and I'm not an expert on this,

235
00:17:02,217 --> 00:17:03,577
so maybe someone can correct me,

236
00:17:04,057 --> 00:17:05,797
call in and tell me I'm wrong, essentially,

237
00:17:06,057 --> 00:17:09,378
but I don't think that gate-based quantum computing

238
00:17:09,378 --> 00:17:12,018
can ever break cryptographic systems.

239
00:17:12,598 --> 00:17:14,098
Because if I understand correctly

240
00:17:14,098 --> 00:17:15,618
from reading Shor's algorithm,

241
00:17:15,618 --> 00:17:20,237
you need to have a certain number of qubits

242
00:17:20,237 --> 00:17:22,697
all in a single device working together

243
00:17:22,697 --> 00:17:26,438
to do the one step,

244
00:17:26,638 --> 00:17:28,477
kind of this critical step in the Shor's algorithm

245
00:17:28,477 --> 00:17:30,697
has to have all those qubits together.

246
00:17:30,818 --> 00:17:32,618
It can't be separated across gate boundaries.

247
00:17:33,297 --> 00:17:34,557
Again, someone could correct me if I'm wrong,

248
00:17:34,618 --> 00:17:37,498
but I don't think that's something you can map

249
00:17:37,498 --> 00:17:38,118
between A and B.

250
00:17:38,197 --> 00:17:40,918
So there are other problems in physics and chemistry and stuff

251
00:17:40,918 --> 00:17:43,338
where the gate-based quantum computers may be very useful,

252
00:17:43,557 --> 00:17:45,498
but I don't think they apply to cryptosystems.

253
00:17:45,618 --> 00:17:49,037
So anyway, that's the things that I've heard where they've made major steps in the last

254
00:17:49,037 --> 00:17:50,037
decade or so.

255
00:17:50,037 --> 00:17:54,737
Okay, so one last piece of sort of table setting before we get onto Bitcoin and what this potentially

256
00:17:54,737 --> 00:17:56,217
means for it.

257
00:17:56,217 --> 00:18:01,178
When you say qubits, are you talking about logical qubits or just normal qubits?

258
00:18:01,178 --> 00:18:03,498
And what's the difference between those two?

259
00:18:03,498 --> 00:18:05,178
Yeah, great clarification.

260
00:18:05,178 --> 00:18:07,678
I was wondering if we're going to get into that.

261
00:18:07,678 --> 00:18:10,418
It's important, so you're right.

262
00:18:10,418 --> 00:18:19,918
So because qubits are these kind of flaky, difficult to deal with things, if you take

263
00:18:19,918 --> 00:18:24,598
a single qubit, which essentially a qubit is often a single electron or a single, I think,

264
00:18:24,598 --> 00:18:30,717
muon or kind of other subatomic particles or a single photon, it's very hard to get

265
00:18:30,717 --> 00:18:38,658
any kind of reliable setting and reading with a single physical qubit in one of those subatomic

266
00:18:38,658 --> 00:18:39,658
particles.

267
00:18:39,658 --> 00:18:42,678
is to do actual computation.

268
00:18:43,078 --> 00:18:45,178
They take a whole bunch of physical qubits

269
00:18:45,178 --> 00:18:49,358
and they actually within the physical,

270
00:18:50,578 --> 00:18:52,457
interconnecting the physical qubits

271
00:18:52,457 --> 00:18:55,237
in a physical kind of device, let's say,

272
00:18:55,797 --> 00:18:58,557
they make it so that a bunch of physical qubits

273
00:18:58,557 --> 00:18:59,858
represent one state

274
00:18:59,858 --> 00:19:04,057
and they essentially self-correct each other within that

275
00:19:04,057 --> 00:19:06,018
so that there's a reliable state

276
00:19:06,018 --> 00:19:07,678
instead of a flaky,

277
00:19:07,818 --> 00:19:09,197
we can't really read it or write it state.

278
00:19:09,658 --> 00:19:25,758
So by putting a whole bunch of physical qubits together, they're able to create a somewhat stable, we're talking in many cases stability over microseconds, but somewhat stable qubit state where the physical qubits reinforce each other to create a stable logical state.

279
00:19:25,758 --> 00:19:28,178
And that's what gives rise to the idea of a logical qubit.

280
00:19:28,178 --> 00:19:44,398
Now, there's a lot of nuance and debate about this, and I was having a discussion on X with someone who works in this area where some people have said that doing this whole thing of mapping physical qubits into logical qubits is not necessary, and it's actually wrong-headed.

281
00:19:44,898 --> 00:19:47,518
Basically, if you're doing a whole bunch of work

282
00:19:47,518 --> 00:19:50,118
to take, let's say, tens or hundreds of physical qubits

283
00:19:50,118 --> 00:19:51,778
to make one logical qubit,

284
00:19:52,438 --> 00:19:55,098
and then you're going to try and tie a bunch of logical qubits together

285
00:19:55,098 --> 00:19:57,297
to get a low error rate output,

286
00:19:57,697 --> 00:19:58,998
that's the wrong way to approach it.

287
00:19:59,038 --> 00:20:01,457
And the right way to approach it is to take all those physical qubits,

288
00:20:02,118 --> 00:20:04,457
map your problem directly onto those physical qubits,

289
00:20:04,737 --> 00:20:08,278
and then deal with the fact that they're flaky

290
00:20:08,278 --> 00:20:09,878
kind of internally to your algorithm.

291
00:20:09,878 --> 00:20:19,338
So basically make almost like one giant logical quantum device rather than making a bunch of logical qubits and then tying those together.

292
00:20:19,838 --> 00:20:26,217
So it's not totally clear to me kind of which one of these approaches is going to bear fruit if either of them ever will.

293
00:20:26,658 --> 00:20:32,217
But that's the difference is that a logical qubit is a single stable quantum state represented by a bunch of physical qubits.

294
00:20:32,217 --> 00:20:34,638
and that may be necessary

295
00:20:34,638 --> 00:20:37,118
in order to build computing devices

296
00:20:37,118 --> 00:20:39,457
out of quantum bits

297
00:20:39,457 --> 00:20:41,418
because the physical qubits

298
00:20:41,418 --> 00:20:42,998
are just too flaky to deal with on their own.

299
00:20:44,358 --> 00:20:46,178
Okay, I said last bit of table setting

300
00:20:46,178 --> 00:20:47,338
but I actually have one more question.

301
00:20:47,898 --> 00:20:49,697
Like in classical computing

302
00:20:49,697 --> 00:20:50,758
we have Moore's Law

303
00:20:50,758 --> 00:20:52,418
which I actually don't know the exact definition.

304
00:20:52,578 --> 00:20:54,217
Does it get everything gets twice as fast

305
00:20:54,217 --> 00:20:55,697
and half the price every few years?

306
00:20:56,318 --> 00:20:58,197
It's actually the number of transistors

307
00:20:58,197 --> 00:20:59,977
in the same physical space

308
00:20:59,977 --> 00:21:01,498
doubles every 18 months.

309
00:21:02,217 --> 00:21:17,398
Okay. And that's, you know, over the history of computing, that's probably been accurate. In quantum computing, they have, is it Nevin's law? And can you explain what that is and what they, you know, if you believe in this?

310
00:21:17,398 --> 00:21:46,018
Yeah, yeah. So first, I have to make one important distinction on Moore's Law. It really should have been called Moore's Observation, right? This wasn't Moore saying, oh, I predict that we're going to see a doubling of transistors on a device every 18 months. It was an observation. He looked back at computing, starting with the very early integrated circuits, and noticed that we're seeing a doubling of transistors every 18 months, and said, this may continue for some time into the future, and it got called Moore's Law.

311
00:21:47,398 --> 00:21:59,797
So in quantum computing, they're saying there's Nevin's law, which would, and they talk about this interesting thing where they call it a double exponential in computing power is what the, I believe the claim of Nevin's law.

312
00:21:59,797 --> 00:22:03,838
and the claim is that because

313
00:22:03,838 --> 00:22:07,557
and this is if quantum computing can scale

314
00:22:07,557 --> 00:22:09,078
there is truth to this first part

315
00:22:09,078 --> 00:22:13,038
because quantum computing represents a complete

316
00:22:13,038 --> 00:22:16,957
possible state space in its entangled bits

317
00:22:16,957 --> 00:22:20,998
your equivalent digital computing

318
00:22:20,998 --> 00:22:23,118
goes up exponentially

319
00:22:23,118 --> 00:22:25,697
with the linear increase in qubits

320
00:22:25,697 --> 00:22:28,098
and that's an important thing about quantum computing

321
00:22:28,098 --> 00:22:32,178
and I think why people get excited or scared about it, right?

322
00:22:32,438 --> 00:22:34,138
If we can get quantum computing to work,

323
00:22:34,197 --> 00:22:37,237
it's absolutely true that the equivalent digital computing

324
00:22:37,237 --> 00:22:39,118
that you get out of a quantum device

325
00:22:39,118 --> 00:22:41,538
goes up exponentially with a linear increase

326
00:22:41,538 --> 00:22:43,598
in the number of functional qubits.

327
00:22:44,197 --> 00:22:44,938
That's very important.

328
00:22:46,217 --> 00:22:48,338
Then the claim comes that we will see

329
00:22:48,338 --> 00:22:54,217
a similar exponential rise in the number of useful qubits

330
00:22:54,217 --> 00:22:57,318
as we saw with useful transistors in a device.

331
00:22:58,098 --> 00:23:06,678
So that's a prediction right now, which has no historical justification.

332
00:23:06,678 --> 00:23:12,217
And so unlike Moore's law, Nevin's law is making a prediction, not an observation.

333
00:23:13,018 --> 00:23:15,538
And that's why I am highly skeptical of it.

334
00:23:15,838 --> 00:23:22,378
And I keep saying on X and elsewhere, I'm waiting to see a couple of cycles of, you know,

335
00:23:22,378 --> 00:23:26,338
true scaling of quantum device complexity,

336
00:23:26,338 --> 00:23:29,038
where we see, oh, they factored 21,

337
00:23:29,178 --> 00:23:30,498
they factored 125,

338
00:23:30,977 --> 00:23:32,697
they factored bigger and bigger numbers

339
00:23:32,697 --> 00:23:34,138
over, let's say, five years,

340
00:23:34,477 --> 00:23:36,098
and then we'll be able to know

341
00:23:36,098 --> 00:23:38,298
what the correct rate of growth is for Nevin's Law.

342
00:23:38,898 --> 00:23:40,438
Maybe it's doubling every 10 years.

343
00:23:41,217 --> 00:23:42,638
Maybe it's actually linear.

344
00:23:42,778 --> 00:23:44,758
Maybe it's not exponential the way transistors were.

345
00:23:44,858 --> 00:23:45,938
We just don't know yet

346
00:23:45,938 --> 00:23:47,138
because we don't have the observations.

347
00:23:47,538 --> 00:23:49,258
Moore made a prediction based on observations.

348
00:23:49,818 --> 00:23:51,898
So I don't believe in Nevin's Law as of right now.

349
00:23:52,377 --> 00:23:57,138
I mean, anything that claims a double exponential, I'm immediately skeptical of.

350
00:23:57,938 --> 00:24:03,377
But so in terms of like quantum computing for such a long time has always been this thing that's like a 30 year away problem.

351
00:24:03,678 --> 00:24:12,678
And then in the last couple of years, people think this is becoming more and more imminent to the point where we have people working on quantum proof addresses on Bitcoin, like Hunter Beast doing Bit360.

352
00:24:13,358 --> 00:24:16,118
Nick Carter's saying that this is an imminent threat.

353
00:24:16,618 --> 00:24:17,778
Where are you on that scale?

354
00:24:17,778 --> 00:24:21,358
Do you think it will ever be a threat or how far out do you think it is?

355
00:24:22,377 --> 00:24:25,578
I have two answers to that question.

356
00:24:25,578 --> 00:24:29,197
I think I can make a very defensible,

357
00:24:29,697 --> 00:24:31,858
like logically tight claim

358
00:24:31,858 --> 00:24:33,498
that there is no way

359
00:24:33,498 --> 00:24:35,758
a cryptographically relevant quantum computer

360
00:24:35,758 --> 00:24:37,098
will happen in the next 10 years.

361
00:24:38,118 --> 00:24:40,957
It just requires too much increase

362
00:24:40,957 --> 00:24:43,138
in the number of logical qubits on a device.

363
00:24:43,778 --> 00:24:46,278
There's no way from a kind of human,

364
00:24:46,438 --> 00:24:49,538
even with AI enhancement development process,

365
00:24:49,758 --> 00:24:51,818
the industrial production ability

366
00:24:51,818 --> 00:24:57,358
it can't be developed in less than 10 years. So that's a defensible, I think there's a strong

367
00:24:57,358 --> 00:25:03,858
logical claim of that. My personal emotional, like what I think, never. I think it is never

368
00:25:03,858 --> 00:25:08,717
going to happen. I suspect that there are physical laws that prevent it from happening,

369
00:25:08,957 --> 00:25:13,758
but that's an emotional thing, not something I can like do a scientifically rigorous study of

370
00:25:13,758 --> 00:25:19,438
or anything. Okay, so let's go with the 10-year prediction then. It's at least 10 years off.

371
00:25:19,438 --> 00:25:22,618
is still worth talking about as of maybe now.

372
00:25:23,158 --> 00:25:24,818
10 years isn't that long a time.

373
00:25:24,957 --> 00:25:26,838
And we know that changes take a long time

374
00:25:26,838 --> 00:25:28,298
to be implemented in Bitcoin.

375
00:25:28,858 --> 00:25:31,318
Let's get into what the actual threat to Bitcoin is

376
00:25:31,318 --> 00:25:34,138
if we get a quantum computer that can break cryptography.

377
00:25:34,278 --> 00:25:36,438
So what will happen first?

378
00:25:37,918 --> 00:25:39,178
Yeah, so the relevant thing,

379
00:25:39,237 --> 00:25:40,758
as I mentioned earlier, is Shor's algorithm.

380
00:25:41,278 --> 00:25:44,118
And Shor's algorithm would enable the quantum attacker

381
00:25:44,118 --> 00:25:46,278
with a cryptographically relevant quantum computer

382
00:25:46,278 --> 00:25:48,457
to take our public keys

383
00:25:48,457 --> 00:25:49,877
and turn them into secret keys.

384
00:25:51,298 --> 00:25:52,578
And if they could do that, of course,

385
00:25:52,638 --> 00:25:54,038
then they could spend anyone's Bitcoin.

386
00:25:54,457 --> 00:25:55,717
And so that's the threat.

387
00:25:56,098 --> 00:25:59,217
And so we don't know exactly how that would play out

388
00:25:59,217 --> 00:26:00,717
because it's very hard to game out.

389
00:26:01,558 --> 00:26:03,217
Let's say you are the first person

390
00:26:03,217 --> 00:26:04,877
that develops a cryptographically relevant

391
00:26:04,877 --> 00:26:05,457
quantum computer.

392
00:26:06,318 --> 00:26:07,538
What are you going to do first?

393
00:26:07,778 --> 00:26:09,078
Are you going to go for Satoshi's coins?

394
00:26:10,078 --> 00:26:11,798
Are you going to try and hack into North Korea?

395
00:26:11,938 --> 00:26:12,957
Are you going to try to hack into Russia?

396
00:26:13,058 --> 00:26:14,457
Are you going to try to hack into the US government?

397
00:26:15,197 --> 00:26:17,237
What's the first target if you get that device?

398
00:26:17,237 --> 00:26:41,178
And it's very hard to know because there's different game theoretical reason to do each of those things. You know, a certain entity might find just kind of slowly siphoning Bitcoin away to be the best use of that device. But another entity with the exact same device might think that infiltrating the president of, I don't know, the Ukraine even might be their first target. Like, we just don't know.

399
00:26:41,178 --> 00:26:44,278
and so I don't know what happened first

400
00:26:44,278 --> 00:26:46,098
what I do know is that Bitcoin's

401
00:26:46,098 --> 00:26:48,298
public keys would become

402
00:26:48,298 --> 00:26:49,737
vulnerable to such a device

403
00:26:49,737 --> 00:26:52,178
and this is where we get into then this like

404
00:26:52,178 --> 00:26:54,158
long exposure versus short exposure that

405
00:26:54,158 --> 00:26:56,098
everyone likes to talk about and so

406
00:26:56,098 --> 00:26:58,338
if you are someone whose public

407
00:26:58,338 --> 00:27:00,178
keys are just as secret as your secret

408
00:27:00,178 --> 00:27:02,178
keys meaning you've never used

409
00:27:02,178 --> 00:27:04,118
an address you've never exposed your x

410
00:27:04,118 --> 00:27:05,818
pub you've never exposed your descriptor

411
00:27:05,818 --> 00:27:08,158
you've never you've not used a taproot address

412
00:27:08,158 --> 00:27:10,217
or a pay to pub key address you've only used

413
00:27:10,217 --> 00:27:15,618
address is ending in H, as Hunter likes to call them, and you've been absolutely perfect about

414
00:27:15,618 --> 00:27:19,898
your key discipline, your public keys are as secret as your secret keys, the quantum attacker

415
00:27:19,898 --> 00:27:25,158
cannot steal your Bitcoin at rest because your public keys are not known. And the only thing that

416
00:27:25,158 --> 00:27:30,178
a quantum computer is likely to be able to do when it's first developed, a relevant quantum computer,

417
00:27:30,518 --> 00:27:38,078
is to reverse public keys into secret keys. At some point, and if this happens, it's hard to know

418
00:27:38,078 --> 00:27:39,078
exactly the timeline here,

419
00:27:39,158 --> 00:27:39,898
but at some point

420
00:27:39,898 --> 00:27:41,278
after the first cryptographically

421
00:27:41,278 --> 00:27:42,178
relevant quantum computer,

422
00:27:42,638 --> 00:27:43,358
there will come a time

423
00:27:43,358 --> 00:27:44,578
where such a device

424
00:27:44,578 --> 00:27:46,818
can take a public key

425
00:27:46,818 --> 00:27:47,798
and develop a secret key

426
00:27:47,798 --> 00:27:49,138
in the 10 minutes it takes

427
00:27:49,138 --> 00:27:50,018
to find a Bitcoin block.

428
00:27:50,778 --> 00:27:51,818
And when that happens,

429
00:27:51,877 --> 00:27:52,778
then even those of us

430
00:27:52,778 --> 00:27:54,078
that maybe have perfect

431
00:27:54,078 --> 00:27:55,298
public key security,

432
00:27:55,798 --> 00:27:56,477
our public keys are

433
00:27:56,477 --> 00:27:58,158
absolutely secret from everybody,

434
00:27:58,518 --> 00:27:59,538
are then vulnerable

435
00:27:59,538 --> 00:28:00,658
to the quantum computer as well,

436
00:28:00,697 --> 00:28:02,118
because they can see our transaction,

437
00:28:02,558 --> 00:28:03,258
maybe even see it

438
00:28:03,258 --> 00:28:04,237
after it gets confirmed

439
00:28:04,237 --> 00:28:05,638
and have a deal with a miner

440
00:28:05,638 --> 00:28:07,158
to mine an alternate block

441
00:28:07,158 --> 00:28:09,858
that takes those coins to themselves instead, right?

442
00:28:10,178 --> 00:28:12,737
Once there's a fast quantum computer out there,

443
00:28:12,818 --> 00:28:13,598
if sub 10 minutes,

444
00:28:14,438 --> 00:28:17,398
Bitcoin as we know it today is basically not usable

445
00:28:17,398 --> 00:28:19,877
because the coins will just be stolen by someone else

446
00:28:19,877 --> 00:28:21,638
once they try to move.

447
00:28:22,398 --> 00:28:24,018
So that's the threat to Bitcoin is, yes,

448
00:28:24,078 --> 00:28:24,798
with Shor's algorithm,

449
00:28:24,977 --> 00:28:27,118
a sufficiently powerful quantum computer

450
00:28:27,118 --> 00:28:28,858
can start taking people's coins.

451
00:28:29,558 --> 00:28:31,697
Okay, I think it's worth diving into that a bit deeper

452
00:28:31,697 --> 00:28:35,078
because there's, I think, in Payt's PubKey,

453
00:28:35,158 --> 00:28:36,457
which is a very old address format,

454
00:28:36,457 --> 00:28:40,158
I think there's about 2 million coins that we know of in that address format.

455
00:28:40,338 --> 00:28:45,398
So those ones are essentially gone immediately for the reason being that you can see those

456
00:28:45,398 --> 00:28:47,538
public keys on chain at any one time.

457
00:28:47,598 --> 00:28:48,018
Is that right?

458
00:28:48,538 --> 00:28:51,678
Yeah, they're at risk immediately because the public keys are readily available.

459
00:28:51,877 --> 00:28:51,977
Yep.

460
00:28:52,658 --> 00:28:52,898
Okay.

461
00:28:52,977 --> 00:28:57,457
And then for anyone else who's using like SegWit addresses, the public key isn't exposed

462
00:28:57,457 --> 00:28:58,737
until you actually make a transaction.

463
00:28:58,918 --> 00:29:02,778
So as long as the quantum computer can't break the cryptography in less than 10 minutes,

464
00:29:02,858 --> 00:29:05,038
less than it takes to get in a block, you're going to be okay.

465
00:29:05,038 --> 00:29:08,578
if you've had perfect public key security

466
00:29:08,578 --> 00:29:10,078
and you've never leaked your XPUB

467
00:29:10,078 --> 00:29:12,298
and never leaked your public key

468
00:29:12,298 --> 00:29:13,858
and never leaked your descriptor,

469
00:29:13,998 --> 00:29:16,078
which I've said this in a few other places,

470
00:29:16,078 --> 00:29:20,438
this is a really strange security assumption.

471
00:29:20,818 --> 00:29:23,957
So I like to say there's really no difference

472
00:29:23,957 --> 00:29:25,737
that Bitcoin is broken

473
00:29:25,737 --> 00:29:28,377
once there's even a slow quantum computer,

474
00:29:28,658 --> 00:29:30,258
because to me, the idea

475
00:29:30,258 --> 00:29:31,758
that we're going to hang our hats

476
00:29:31,758 --> 00:29:39,598
on people having perfect public key security is very, very strange. I don't accept that as a

477
00:29:39,598 --> 00:29:45,758
security claim for Bitcoin. No, I totally agree with that. I think I would probably say the

478
00:29:45,758 --> 00:29:51,658
majority of my Bitcoin addresses will be fine, but I bet there's one in there that's not. I wouldn't

479
00:29:51,658 --> 00:29:58,158
be sure. But so why do people focus so much on Bitcoin with quantum computing? Because if quantum

480
00:29:58,158 --> 00:30:03,598
computers do break this cryptography, then there's a wide array of things they can go after.

481
00:30:03,598 --> 00:30:17,435
Why is Bitcoin the focus here So I think there is a legitimacy there in that Bitcoin is hard to change And so you know we saw with the Y2K you know well at least I did

482
00:30:17,475 --> 00:30:19,135
I don't know if you're as old as I am.

483
00:30:19,655 --> 00:30:23,515
But with Y2K, everyone said the world's going to end, but then it didn't.

484
00:30:23,675 --> 00:30:26,495
Why? Because a bunch of software engineers busted ass and made it not end.

485
00:30:26,875 --> 00:30:28,835
And that's true for all the other systems.

486
00:30:28,835 --> 00:30:35,535
You know, if quantum computers look like they're around the corner, you know, governments, most governments and most computers,

487
00:30:35,555 --> 00:30:40,535
operating systems and all that stuff can change to other cryptography in the timeframe necessary

488
00:30:40,535 --> 00:30:45,695
to protect their users. It's less clear that Bitcoin can do that. I think that's where Nick

489
00:30:45,695 --> 00:30:51,975
is kind of doing this like, oh, because of this slowness, we have to act now, even though quantum

490
00:30:51,975 --> 00:30:58,155
is only maybe a thing. It's like, yes, Bitcoin's hard to change. And so that's why the real focus

491
00:30:58,155 --> 00:31:02,475
on Bitcoin is that it might take five years to make a change to Bitcoin, and that might be too

492
00:31:02,475 --> 00:31:07,595
long. So there's a legitimacy there. Yeah, I guess the other thing being there's no recourse with

493
00:31:07,595 --> 00:31:11,755
Bitcoin. So if someone, you know, if you own the private key, you own the Bitcoin. It's not like

494
00:31:11,755 --> 00:31:16,255
you're able to claw back money through the banking system or however it might be. There's no insurance

495
00:31:16,255 --> 00:31:22,315
here. Like this, if you own the private key, you own the Bitcoin. So there are obviously, there are

496
00:31:22,315 --> 00:31:26,655
things we can do to mitigate this attack. But when do you think the conversation really needs to ramp

497
00:31:26,655 --> 00:31:30,715
up from something that is like cool for developers to be working on? We should be thinking about this.

498
00:31:30,715 --> 00:31:34,375
But when does it get to the point where you're like, oh shit, something needs to happen now?

499
00:31:35,135 --> 00:31:39,495
Yeah, so I've said this on the internet before, but I think it's really worth saying in many

500
00:31:39,495 --> 00:31:40,815
different forums over and over.

501
00:31:41,235 --> 00:31:48,835
As soon as there is evidence of clear logical qubit scaling, of progressive factorization

502
00:31:48,835 --> 00:31:55,715
or reversing of electric curve keys that gets greater bit counts reversed with sub-exponential

503
00:31:55,715 --> 00:32:01,635
scaling of energy input, right? There's a very clear evidentiary standard here.

504
00:32:02,055 --> 00:32:05,935
And as soon as that evidentiary standard is met, we have to start doing stuff.

505
00:32:06,955 --> 00:32:14,415
But as of right now, no one has gone from, you know, reversing a 6-bit EC key to a 10-bit EC key

506
00:32:14,415 --> 00:32:19,735
using sub-exponential increases in time and energy. And until there's a sub-exponential

507
00:32:19,735 --> 00:32:24,615
increase in the difficulty of doing that, we still have the exact security model we've always

508
00:32:24,615 --> 00:32:29,175
promised with elliptic curves, which is that it is exponentially difficult to break elliptic curve

509
00:32:29,175 --> 00:32:35,595
keys of a certain size. And as long as it's scaling exponentially with energy input, we still have

510
00:32:35,595 --> 00:32:42,415
that. So it's very easy to change my mind. The evidence can be just put out, hey, look, someone

511
00:32:42,415 --> 00:32:48,215
made a quantum device that for a 6-bit key takes X energy and for a 12-bit key takes Y energy and

512
00:32:48,215 --> 00:32:52,935
that's sub-exponential scaling. Oh, okay, shit, we got to do something. If you're already self-custody

513
00:32:52,935 --> 00:32:57,195
your Bitcoin, you know the deal with hardware wallets. Complex setups, clumsy interfaces,

514
00:32:57,555 --> 00:33:02,875
and a seed phrase that can be lost, stolen, or forgotten. Well, BitKey fixes that. BitKey is a

515
00:33:02,875 --> 00:33:07,595
multi-sig hardware wallet built by the team behind Square and Cash App. It packs a cryptographic

516
00:33:07,595 --> 00:33:12,655
recovery system and built-in inheritance feature into an intuitive, easy-to-use wallet with no

517
00:33:12,655 --> 00:33:18,735
seed phrase to sweat over. It's simple, secure self-custody without the stress. And Time named

518
00:33:18,735 --> 00:33:25,715
Bitkey one of the best inventions of 2024. Get 20% off at bitkey.world when you use the code WBD.

519
00:33:26,235 --> 00:33:33,295
That's B-I-T-K-E-Y dot world and use the code WBD. This episode is brought to you by Anchor Watch.

520
00:33:33,595 --> 00:33:37,955
The thing that keeps me up at night is the idea of a critical error with my Bitcoin cold storage

521
00:33:37,955 --> 00:33:42,315
and this is where Anchor Watch comes in. With Anchor Watch your Bitcoin is insured with your

522
00:33:42,315 --> 00:33:47,615
own A-plus rated Lloyds of London insurance policy and all Bitcoin is held in their time-locked

523
00:33:47,615 --> 00:33:51,915
multi-sig vaults. So you have the peace of mind knowing your Bitcoin is insured while not giving

524
00:33:51,915 --> 00:33:56,255
up custody. So whether you're worried about inheritance planning, wrench attacks, natural

525
00:33:56,255 --> 00:34:01,515
disasters or just your own silly mistakes, you're protected by Anchor Watch. Rates for fully insured

526
00:34:01,515 --> 00:34:06,795
custody start as low as 0.55% and are available for individual and commercial customers located

527
00:34:06,795 --> 00:34:11,875
in the US. Speak to Anchor Watch for a quote and for more details about your security options and

528
00:34:11,875 --> 00:34:16,615
coverage. Visit anchorwatch.com today. That is anchorwatch.com.

529
00:34:17,075 --> 00:34:20,395
What if you could lower your tax bill and stack Bitcoin at the same time?

530
00:34:20,835 --> 00:34:23,475
Well, by mining Bitcoin with Blockware, you can.

531
00:34:24,055 --> 00:34:28,255
New tax guidelines from the Big Beautiful Bill allow American miners to write off 100%

532
00:34:28,255 --> 00:34:33,095
of the cost of their mining hardware in a single tax year. That's right, 100% write-off.

533
00:34:33,735 --> 00:34:38,395
So if you have $100,000 in capital gains or income, you can purchase $100,000 of miners

534
00:34:38,395 --> 00:34:43,595
and offset it entirely. Blockware's mining as a service enables you to start mining Bitcoin right

535
00:34:43,595 --> 00:34:48,395
now without lifting a finger. Blockware handles everything from securing the miners to sourcing

536
00:34:48,395 --> 00:34:53,435
low-cost power to configuring the pool, they do it all. You get to stack Bitcoin at a discount

537
00:34:53,435 --> 00:34:58,715
every single day while also saving big come tax season. Get started today by going to

538
00:34:58,715 --> 00:35:04,395
mining.blockwaresolutions.com forward slash wbd. Of course none of this is tax advice,

539
00:35:04,395 --> 00:35:08,155
speak to your accountant or tax advisor to understand how these rules apply to you

540
00:35:08,395 --> 00:35:12,675
And then head over to mining.blockwaresolutions.com forward slash WBD.

541
00:35:12,855 --> 00:35:17,095
And you'll get one week of free hosting and electricity with each hosted miner purchased.

542
00:35:18,095 --> 00:35:23,095
I mean, it's funny because like I probably buy some of the stuff that Nick's saying.

543
00:35:23,175 --> 00:35:27,795
Like I can believe that there's plenty of people who would willingly allocate money to Bitcoin

544
00:35:27,795 --> 00:35:33,255
that see the quantum computer threat, think that this is real and is closer than it may be in

545
00:35:33,255 --> 00:35:37,755
reality, and then be cautious of ever like touching Bitcoin because of that.

546
00:35:38,135 --> 00:35:39,335
I think that probably exists.

547
00:35:39,475 --> 00:35:41,115
I think it's probably a small minority at the moment,

548
00:35:41,115 --> 00:35:42,515
but it's probably going to grow as well

549
00:35:42,515 --> 00:35:44,235
as the hype around quantum computing grows.

550
00:35:44,595 --> 00:35:47,355
So we do have people working on solutions.

551
00:35:47,935 --> 00:35:49,155
What are those solutions?

552
00:35:49,235 --> 00:35:51,335
So quantum proof cryptography,

553
00:35:51,515 --> 00:35:53,435
like maybe explain how that works

554
00:35:53,435 --> 00:35:55,055
and why it's not an ideal solution,

555
00:35:55,155 --> 00:35:55,895
at least right now.

556
00:35:57,195 --> 00:35:59,795
Yeah, so there's two major camps

557
00:35:59,795 --> 00:36:02,235
of quantum resistant cryptography out there

558
00:36:02,235 --> 00:36:03,955
being proposed for Bitcoin at least right now.

559
00:36:04,135 --> 00:36:06,475
And those are hash-based and lattice-based.

560
00:36:06,475 --> 00:36:08,435
I understand hash-based mostly.

561
00:36:08,555 --> 00:36:10,275
I don't understand lattice-based mostly.

562
00:36:11,115 --> 00:36:18,055
But they, for the moment at least, have somewhat similar problems, which is that they require

563
00:36:18,055 --> 00:36:23,955
much more data to prove ownership and validate signatures than elliptic curves.

564
00:36:24,355 --> 00:36:30,715
Elliptic curves were chosen for a lot of cryptosystems because they have a really, even compared

565
00:36:30,715 --> 00:36:35,055
to RSA, which was a previous cryptosystem that has similar properties in many ways to

566
00:36:35,055 --> 00:36:41,755
elliptic curves. They have much smaller keys and signatures. So like a Bitcoin key is 32 bytes,

567
00:36:41,955 --> 00:36:48,135
33 bytes, depending on things. And a Bitcoin signature is 64 to 70 bytes. These are very

568
00:36:48,135 --> 00:36:52,295
small amounts of data to prove the ownership of something. Really, if you think about what you're

569
00:36:52,295 --> 00:36:56,855
doing and the ability to move Bitcoin and prove ownership with a total of 100 bytes is just

570
00:36:56,855 --> 00:37:03,035
shockingly efficient. And quantum resistant algorithms, whether hash-based or lattice-based,

571
00:37:03,035 --> 00:37:05,155
are many kilobytes.

572
00:37:05,455 --> 00:37:06,935
So not many, several kilobytes at least,

573
00:37:07,015 --> 00:37:08,895
many kilobytes for kind of simple systems

574
00:37:08,895 --> 00:37:10,975
to do that same job of proving ownership

575
00:37:10,975 --> 00:37:11,915
and transfer of Bitcoin.

576
00:37:12,595 --> 00:37:14,015
So that's the big problem right now.

577
00:37:14,655 --> 00:37:17,095
Recent research has made great strides

578
00:37:17,095 --> 00:37:18,755
in reducing the cost of these things,

579
00:37:18,795 --> 00:37:19,795
especially the compute cost.

580
00:37:19,875 --> 00:37:21,535
So there's been some posts going back and forth

581
00:37:21,535 --> 00:37:22,675
on the Bitcoin dev mailing list

582
00:37:22,675 --> 00:37:25,295
about these post-quantum schemes

583
00:37:25,295 --> 00:37:27,295
and how they've really gotten

584
00:37:27,295 --> 00:37:29,015
the signature verification,

585
00:37:29,375 --> 00:37:30,535
which is probably the most important

586
00:37:30,535 --> 00:37:32,635
to optimize portion of the Bitcoin process,

587
00:37:32,635 --> 00:37:36,395
to about the same compute cost as elliptic curve verification.

588
00:37:36,595 --> 00:37:37,655
So that's like a huge innovation

589
00:37:37,655 --> 00:37:39,675
and great progress has been made there.

590
00:37:40,835 --> 00:37:43,435
And so we're just really, I would say in the Bitcoin world,

591
00:37:43,655 --> 00:37:46,195
we're looking for a sufficiently developed

592
00:37:46,195 --> 00:37:47,855
post-quantum crypto system

593
00:37:47,855 --> 00:37:50,635
that doesn't have major downsides for the system

594
00:37:50,635 --> 00:37:52,375
like multi-kilobyte keys.

595
00:37:53,495 --> 00:37:55,775
And the issue there, like in layman's terms,

596
00:37:55,875 --> 00:37:58,375
is it takes Bitcoin from being, I don't know,

597
00:37:58,715 --> 00:38:00,035
do you know how many transactions a second?

598
00:38:00,195 --> 00:38:01,995
I even hate that metric, but how many transactions?

599
00:38:01,995 --> 00:38:03,775
Seven transactions per second is often cited.

600
00:38:03,975 --> 00:38:07,835
Okay, but then it's going to go down to probably less than one if we go to these larger signatures.

601
00:38:08,335 --> 00:38:08,975
Yep, exactly.

602
00:38:08,975 --> 00:38:16,455
And so if we have 10 years, how small do you think we can make those signature sizes?

603
00:38:16,735 --> 00:38:19,375
And how close to what we're working with now?

604
00:38:20,955 --> 00:38:24,695
I am not a cryptographer, so I'm hesitant to even put a prediction on it.

605
00:38:24,695 --> 00:38:42,695
What I would guess is that over 10 years, we can make sufficient developments in kind of layered Bitcoin technologies, whether that's ARCs or Lightning or BitVMs or whatever else people are working on.

606
00:38:43,815 --> 00:38:51,035
So that we can work with the larger signatures, however big, however small we can get them in the same time.

607
00:38:51,095 --> 00:38:53,055
So let's kind of attack it from two sides, right?

608
00:38:53,055 --> 00:38:57,295
make Bitcoin more efficient with layering and reduce the size of the signatures sufficiently

609
00:38:57,295 --> 00:39:01,515
over the course of the time so that we can build a system that really works for people

610
00:39:01,515 --> 00:39:06,635
by the time we need to. It's one of those funny things where I guess the ideal solution here,

611
00:39:06,715 --> 00:39:10,475
if you assume quantum computing is real and it's going to be able to break cryptography at some

612
00:39:10,475 --> 00:39:15,415
point in the future, you want to make the change as late as possible so you have the best solution

613
00:39:15,415 --> 00:39:22,015
possible. But to get that, you do need people working on today, which we do have. But if we

614
00:39:22,015 --> 00:39:26,675
do get there and we need to make a change to Bitcoin, is this a soft fork or a hard fork?

615
00:39:27,895 --> 00:39:31,295
Soft, yeah. Everything here can be done in a soft fork. And I think there'd be a really

616
00:39:31,295 --> 00:39:36,435
interesting discussion on the same way that segregated witness was a soft fork that technically

617
00:39:36,435 --> 00:39:42,895
added block space. If we've made significant strides in the validation cost and we've seen

618
00:39:42,895 --> 00:39:50,055
increases in storage capacity on typical devices and stuff, it might make sense to do another thing

619
00:39:50,055 --> 00:39:56,695
like SegWit where, okay, we get post-quantum and we make a special quantum signature block

620
00:39:56,695 --> 00:40:00,315
that lets us keep about the same transaction throughput. So there's a lot of conversations

621
00:40:00,315 --> 00:40:05,175
to be had here when the time comes. In the meantime, as you said, the right thing to do

622
00:40:05,175 --> 00:40:08,335
right now is to continue developing these post-quantum signature schemes, making them

623
00:40:08,335 --> 00:40:10,655
better and better and better over time so that we're ready.

624
00:40:12,235 --> 00:40:17,955
One of the most interesting things I've heard you say on this is that you think regular computers

625
00:40:17,955 --> 00:40:20,295
might break cryptography before quantum computers.

626
00:40:21,055 --> 00:40:21,875
Explain that.

627
00:40:21,935 --> 00:40:22,375
What do you mean?

628
00:40:24,575 --> 00:40:25,835
As many as people have said,

629
00:40:26,715 --> 00:40:28,115
crypto systems have a shelf life.

630
00:40:28,515 --> 00:40:31,115
You know, RSA 1024-bit was secure

631
00:40:31,115 --> 00:40:32,795
when I first started using cryptography,

632
00:40:33,035 --> 00:40:35,735
but soon after I had to upgrade my RSA keys to 2048,

633
00:40:35,875 --> 00:40:38,295
and then I upgraded further to 4096-bit keys.

634
00:40:39,955 --> 00:40:41,295
SHA-1 is no longer secure.

635
00:40:41,395 --> 00:40:42,415
MD5 is no longer secure.

636
00:40:43,035 --> 00:40:45,255
You know, crypto systems have a shelf life.

637
00:40:45,255 --> 00:40:47,455
and elliptic curves.

638
00:40:48,195 --> 00:40:50,735
The nice thing about modern cryptography,

639
00:40:50,855 --> 00:40:52,395
like we're kind of several generations in,

640
00:40:52,475 --> 00:40:53,535
obviously, of cryptography,

641
00:40:53,915 --> 00:40:59,275
is that we have pretty strong proofs of security.

642
00:40:59,495 --> 00:41:00,615
Like we have, here's the assumption.

643
00:41:01,075 --> 00:41:02,395
If you don't break this assumption,

644
00:41:02,535 --> 00:41:03,415
this system is secure.

645
00:41:05,155 --> 00:41:09,135
But there's still new math being discovered every year.

646
00:41:09,635 --> 00:41:12,535
And there's no way to know for sure

647
00:41:12,535 --> 00:41:17,155
that some genius working in a university somewhere

648
00:41:17,155 --> 00:41:21,735
doesn't come up with a way to basically attack

649
00:41:21,735 --> 00:41:25,375
the underpinnings of the very elliptic curve system.

650
00:41:25,855 --> 00:41:30,615
You know, the SCCP-256K1 curve that we use in Bitcoin

651
00:41:30,615 --> 00:41:32,915
has a specific formula that describes

652
00:41:32,915 --> 00:41:34,555
the shape of the elliptic curve.

653
00:41:35,635 --> 00:41:38,475
And then on that curve, our whole cryptosystem

654
00:41:38,475 --> 00:41:40,955
revolves around moving a point around on that curve

655
00:41:40,955 --> 00:41:42,595
based on secret keys and public keys, right?

656
00:41:44,435 --> 00:41:46,335
What if there's a vulnerability in that curve

657
00:41:46,335 --> 00:41:48,155
and the points on the curve can be predicted

658
00:41:48,155 --> 00:41:49,535
based on the public key?

659
00:41:50,935 --> 00:41:51,835
That's a possibility.

660
00:41:52,175 --> 00:41:53,015
I don't think so.

661
00:41:53,075 --> 00:41:54,075
It's been around for long enough,

662
00:41:54,255 --> 00:41:57,335
but to me, that's a more likely attack vector

663
00:41:57,335 --> 00:41:59,095
for Bitcoin than quantum computing

664
00:41:59,095 --> 00:42:01,295
because we've seen that happen many times.

665
00:42:01,775 --> 00:42:03,435
You'll hear, I kind of go with evidence.

666
00:42:03,935 --> 00:42:05,415
And the evidence is that crypto systems

667
00:42:05,415 --> 00:42:08,815
are broken by innovations in mathematics and cryptography.

668
00:42:08,815 --> 00:42:12,295
they're not broken by like brand new types of computers.

669
00:42:12,295 --> 00:42:13,375
They're broken by math.

670
00:42:13,575 --> 00:42:14,375
And that's what I would think

671
00:42:14,375 --> 00:42:15,555
is the most likely vulnerability.

672
00:42:16,235 --> 00:42:17,195
Now, the great thing is that

673
00:42:17,195 --> 00:42:19,095
the solution to both of these is the same thing.

674
00:42:19,235 --> 00:42:21,295
We should keep developing other crypto systems

675
00:42:21,295 --> 00:42:22,475
that might be suitable for Bitcoin.

676
00:42:22,755 --> 00:42:24,455
And we should implement one in Bitcoin

677
00:42:24,455 --> 00:42:25,375
when the time is right.

678
00:42:25,795 --> 00:42:26,775
Oh, great, we'll do that.

679
00:42:28,155 --> 00:42:31,375
I mean, I imagine AI is gonna also,

680
00:42:31,575 --> 00:42:33,015
like people talk about the breakthroughs

681
00:42:33,015 --> 00:42:34,715
that may come in math and physics through AI.

682
00:42:34,915 --> 00:42:36,955
Like that is probably one of the most realistic ways

683
00:42:36,955 --> 00:42:39,315
that current cryptography is broken.

684
00:42:39,675 --> 00:42:40,055
It's funny.

685
00:42:40,135 --> 00:42:42,495
I was actually talking to AI the other day

686
00:42:42,495 --> 00:42:42,855
and I was like,

687
00:42:42,895 --> 00:42:43,935
how would you break ECDSA?

688
00:42:44,155 --> 00:42:44,915
And it won't tell you.

689
00:42:45,035 --> 00:42:46,655
Like it just refuses to answer that question

690
00:42:46,655 --> 00:42:47,695
or it did for me at least.

691
00:42:48,655 --> 00:42:50,615
I mean, not that I was going to understand

692
00:42:50,615 --> 00:42:51,635
anything that it said anyway.

693
00:42:52,195 --> 00:42:54,095
But so, okay, if we need to soft fork,

694
00:42:55,135 --> 00:42:57,715
then that's one of the really interesting parts

695
00:42:57,715 --> 00:42:58,595
of this conversation is like,

696
00:42:58,595 --> 00:43:00,295
what do we do with the old coins

697
00:43:00,295 --> 00:43:02,315
that will otherwise be stolen?

698
00:43:02,755 --> 00:43:03,795
And there's going to be like

699
00:43:03,795 --> 00:43:04,795
part of the Bitcoin community

700
00:43:04,795 --> 00:43:06,935
that think that those should be confiscated

701
00:43:07,435 --> 00:43:09,055
I, at least as of right now,

702
00:43:09,115 --> 00:43:10,635
I'm very strongly against that.

703
00:43:10,715 --> 00:43:12,575
Like one of the key principles in Bitcoin to me

704
00:43:12,575 --> 00:43:13,275
is property rights.

705
00:43:13,335 --> 00:43:15,335
And I don't think we should steal someone's coins,

706
00:43:15,435 --> 00:43:17,955
just the threat that a bad actor steals the coins later.

707
00:43:18,455 --> 00:43:19,895
Where do you fall on that whole thing?

708
00:43:21,755 --> 00:43:24,515
Yeah, I found people arguing with me on both sides of this

709
00:43:24,515 --> 00:43:25,935
because I'm in the middle.

710
00:43:27,755 --> 00:43:29,955
If there is a sudden quantum break,

711
00:43:29,955 --> 00:43:32,015
let's say tomorrow we find out

712
00:43:32,015 --> 00:43:35,195
that there's actually already a quantum computer

713
00:43:35,195 --> 00:43:36,875
active stealing coins, right?

714
00:43:36,935 --> 00:43:52,295
So I think it's never happening, but tomorrow someone proves, hey, look, I'm doing it right now. Here's your secret key. Then we should confiscate the coins because there was no opportunity for anybody to retain their ownership. And so the entire system's ownership has just been invalidated suddenly.

715
00:43:52,295 --> 00:44:07,975
So in that case, we confiscate the coins and we create some kind of a claim system where some people can reclaim their ownership, but we don't just leave it vulnerable to the quantum attacker because everyone's ownership was simultaneously ruined, essentially.

716
00:44:07,975 --> 00:44:22,735
On the other hand, if what is, I think, a much more reasonable expectation, we see a gradual progression and let's say in five years, quantum computers start factoring 20-bit numbers and we say, okay, it's time to do something.

717
00:44:22,735 --> 00:44:38,395
So a year later, we activate a soft fork that has quantum-resistant cryptography in it. And then that year, they factor 32-bit numbers and they keep progressing. Then people have time from the time we deploy that quantum-resistant soft fork to move their coins to quantum-resistant addresses.

718
00:44:38,395 --> 00:44:40,395
and at some point down the road,

719
00:44:40,855 --> 00:44:42,815
that quantum attacker is going to first break

720
00:44:42,815 --> 00:44:44,115
a meaningful Bitcoin key.

721
00:44:44,215 --> 00:44:45,915
They're going to factor a 256-bit number

722
00:44:45,915 --> 00:44:47,775
and break a Bitcoin key.

723
00:44:48,335 --> 00:44:50,215
But by then, everybody who was active in Bitcoin

724
00:44:50,215 --> 00:44:52,095
has already started using the new soft fork.

725
00:44:52,195 --> 00:44:53,015
So then we just say, fine,

726
00:44:53,055 --> 00:44:53,855
they can have such these coins,

727
00:44:53,915 --> 00:44:54,995
they can have all the dead coins.

728
00:44:55,495 --> 00:44:57,495
It's going to be a race between different quantum actors

729
00:44:57,495 --> 00:44:58,515
who gets which coins.

730
00:44:58,935 --> 00:45:01,055
And as you said, we enforce the ownership and say,

731
00:45:01,115 --> 00:45:03,095
no, whoever gets those keys is the owner

732
00:45:03,095 --> 00:45:04,955
because that's the only thing we know in Bitcoin.

733
00:45:06,035 --> 00:45:08,215
So mostly I'm with you.

734
00:45:08,395 --> 00:45:23,653
unless it a sudden thing where tomorrow someone starts stealing quantum coins or quantum vulnerable coins See I don know if I even agree with that part Like it just it feels very like Ethereum DAO type situation where you not happy with an

735
00:45:23,653 --> 00:45:23,953
outcome.

736
00:45:24,093 --> 00:45:27,232
So therefore you kind of roll back the chain and give people their Bitcoin back.

737
00:45:27,512 --> 00:45:32,673
Like, I don't know if I can ever get on board with that because like one of the issues that

738
00:45:32,673 --> 00:45:36,313
I think people will have here is they'll see that, you know, I don't know, 2 million coins

739
00:45:36,313 --> 00:45:39,633
or whatever it is in paid pub key addresses are going to hit the market.

740
00:45:39,633 --> 00:45:41,813
that's going to have a massive impact on Bitcoin price.

741
00:45:41,992 --> 00:45:43,732
And therefore they're like, that's a bad thing.

742
00:45:44,232 --> 00:45:47,453
But if you completely remove or invalidate

743
00:45:47,453 --> 00:45:48,893
one of the value propositions of Bitcoin,

744
00:45:49,012 --> 00:45:50,052
which is property rights,

745
00:45:50,413 --> 00:45:53,073
like what does the long-term value actually become?

746
00:45:53,373 --> 00:45:55,573
Like, I think the number is far lower in the long-term,

747
00:45:55,673 --> 00:45:58,573
even if you have a short-term huge impact on price.

748
00:45:59,052 --> 00:45:59,673
I don't know.

749
00:45:59,732 --> 00:46:03,453
I feel like confiscating coins at any point is wrong,

750
00:46:03,453 --> 00:46:05,332
whether it's tomorrow or it's in 10 years.

751
00:46:07,373 --> 00:46:08,873
No, I love that perspective.

752
00:46:08,873 --> 00:46:14,273
I mean, I think it's a very kind of baller perspective to take, essentially.

753
00:46:15,012 --> 00:46:22,332
I think realistically, and this is the plain reality of Bitcoin, and I was talking about

754
00:46:22,332 --> 00:46:25,613
it actually on Tone Vase Show the other day, it's a market question.

755
00:46:26,413 --> 00:46:28,532
And neither of us knows what the market will do.

756
00:46:29,453 --> 00:46:34,212
I would bet that if we do see there's a sudden quantum adversary where no one knew it was

757
00:46:34,212 --> 00:46:38,972
coming and suddenly all the coins were starting to move, we would see a chain split rather.

758
00:46:39,472 --> 00:46:43,332
So a chain split would happen. There'd be the one that retains ownership, exactly as you said,

759
00:46:43,393 --> 00:46:47,712
it's just, it's Bitcoin. We add quantum resistance and people start kind of fighting each other to

760
00:46:47,712 --> 00:46:52,153
get their coins quantum resistant. And like, we just kind of YOLO, it's, we're baller,

761
00:46:52,212 --> 00:46:57,352
we're going to go with it. And on the other hand, we have people where we cut off the quantum

762
00:46:57,352 --> 00:47:02,653
vulnerable cryptography and we make some kind of a claim process. And we try to preserve ownership

763
00:47:02,653 --> 00:47:05,252
the best we can within that scenario.

764
00:47:05,732 --> 00:47:08,093
And I have no idea really which one would have more value.

765
00:47:08,193 --> 00:47:09,832
I know that I prefer one of them,

766
00:47:10,212 --> 00:47:12,653
but I'm only one market actor with a small number of coins.

767
00:47:12,953 --> 00:47:14,512
I don't know what the future would hold

768
00:47:14,512 --> 00:47:15,732
on either one of those two coins,

769
00:47:15,813 --> 00:47:17,093
but I think we'd have a clear chain split.

770
00:47:18,133 --> 00:47:19,712
Yeah, and the funny thing is like,

771
00:47:19,773 --> 00:47:21,773
even though I feel quite strongly

772
00:47:21,773 --> 00:47:25,153
that you can't ever confiscate coins under any circumstances,

773
00:47:25,573 --> 00:47:27,073
I also could see a world,

774
00:47:27,153 --> 00:47:29,173
like if you do take the Ethereum DAO analogy,

775
00:47:29,552 --> 00:47:31,832
where I would be on the side of like the Ethereum classic,

776
00:47:31,832 --> 00:47:33,332
which fades into insignificance.

777
00:47:33,552 --> 00:47:34,773
That would definitely be a chain split

778
00:47:34,773 --> 00:47:36,453
where I'd never sell the coins on either side.

779
00:47:37,153 --> 00:47:37,613
Right, exactly.

780
00:47:37,813 --> 00:47:38,752
And I think a lot of people do that.

781
00:47:39,693 --> 00:47:41,972
Yeah, but I do think it's probably

782
00:47:41,972 --> 00:47:43,832
one of the most interesting sort of dilemmas

783
00:47:43,832 --> 00:47:46,052
that may occur if we do actually get this.

784
00:47:47,433 --> 00:47:49,972
If that happens, so if we have a soft fork,

785
00:47:50,193 --> 00:47:53,492
so let's say we know quantum computing is coming,

786
00:47:53,693 --> 00:47:54,832
we've seen the progress,

787
00:47:55,032 --> 00:47:56,573
like it's an imminent threat,

788
00:47:56,653 --> 00:47:58,393
we make a change, we have a soft fork,

789
00:47:58,752 --> 00:48:01,732
would everyone then have to migrate keys

790
00:48:01,732 --> 00:48:03,893
to a quantum proof address.

791
00:48:05,093 --> 00:48:06,512
Yeah, everyone have to move their coins,

792
00:48:06,633 --> 00:48:08,693
which would be a big privacy impact for many people.

793
00:48:09,913 --> 00:48:10,813
Huge privacy impact.

794
00:48:10,873 --> 00:48:13,373
And also it would take a very, very long time, right?

795
00:48:13,413 --> 00:48:14,532
Like block space is limited.

796
00:48:14,712 --> 00:48:16,532
Like how long would that process take?

797
00:48:17,873 --> 00:48:19,972
I think it's like for all the UTXOs,

798
00:48:20,052 --> 00:48:21,732
it's like three or four years right now.

799
00:48:22,913 --> 00:48:23,633
No, not that much.

800
00:48:23,633 --> 00:48:23,873
Okay.

801
00:48:24,212 --> 00:48:24,893
I'm not exaggerating.

802
00:48:25,133 --> 00:48:26,032
It's a good while,

803
00:48:26,032 --> 00:48:28,492
but it's measured in years, not decades for sure.

804
00:48:29,893 --> 00:48:30,093
Okay.

805
00:48:30,093 --> 00:48:35,532
So you basically then are in that situation where you need to move your coins as quickly

806
00:48:35,532 --> 00:48:42,752
as possible because the idea is if the quantum computing progress is so fast that not only

807
00:48:42,752 --> 00:48:46,273
can it break cryptography, but it can now do it in less than 10 minutes, then those

808
00:48:46,273 --> 00:48:48,593
coins are also going to be lost forever, essentially.

809
00:48:50,052 --> 00:48:50,492
Yep.

810
00:48:51,212 --> 00:48:52,852
So miners are going to be happy.

811
00:48:52,992 --> 00:48:54,712
Fees are going to go through the absolute roof.

812
00:48:56,173 --> 00:48:58,212
Yeah, I think that's right.

813
00:48:58,212 --> 00:49:04,773
Does it have any other impacts apart from the, like, does it have any impacts in mining quantum

814
00:49:04,773 --> 00:49:11,732
computing? Very unlikely. There's an algorithm out there called Grover's algorithm that can reduce

815
00:49:11,732 --> 00:49:19,653
the difficulty of finding a SHA-256 collision by a square root factor, basically. But the reality

816
00:49:19,653 --> 00:49:24,212
is that Bitcoin's difficulty adjustment can handle that. So if quantum miner is square root faster,

817
00:49:24,212 --> 00:49:31,133
it's okay it doesn't matter um as far as anyone researching has has figured out so far there's not

818
00:49:31,133 --> 00:49:35,832
really a problem for mining because we have a difficulty adjustment and so as quantum miners

819
00:49:35,832 --> 00:49:41,252
start to roll out like we've seen bigger improvements in mining already in the cpu gpu

820
00:49:41,252 --> 00:49:47,893
asic migration than you would see from a quantum device okay so if anyone listening to this has

821
00:49:47,893 --> 00:49:52,593
been panicking about quantum you think they can relax for a good a good deal of time right now

822
00:49:53,352 --> 00:49:58,492
Yeah, I mean, like I said, I can only say with confidence, like serious confidence, it's going to be more than 10 years.

823
00:49:59,012 --> 00:50:01,373
And so we should watch and we should look for the evidence.

824
00:50:01,773 --> 00:50:11,492
And I think even to go further than that, it makes sense to start taking whatever steps toward improving Bitcoin's resistance to an EC break.

825
00:50:12,593 --> 00:50:17,732
We can take the day that are clearly good steps for Bitcoin, that they don't have downsides, whatever.

826
00:50:17,832 --> 00:50:18,573
We should just do them.

827
00:50:19,273 --> 00:50:21,313
I mentioned, I think, Hunter Beast once in this already.

828
00:50:21,313 --> 00:50:29,352
He developed a BIP with some other folks called BIP360, which adds a taproot type address, but without an exposed public key.

829
00:50:30,273 --> 00:50:32,793
And that would give options for quantum resistance in the future.

830
00:50:33,012 --> 00:50:36,832
It also would reduce the cost of certain types of on-chain contracts that people want.

831
00:50:37,293 --> 00:50:38,613
And so it's like, oh, that's a good thing regardless.

832
00:50:38,793 --> 00:50:39,413
Let's just do it.

833
00:50:39,793 --> 00:50:49,953
I think it makes great sense to take concrete steps today that we know are good steps and keep watching the quantum computing and the post-quantum cryptography in the meantime.

834
00:50:49,953 --> 00:50:57,153
And last question on Bitcoin, well, on this side of Bitcoin, I do want to get into how contentious

835
00:50:57,153 --> 00:51:02,192
a soft fork may be in this scenario, but does it have any impact on mining? Obviously, to open a

836
00:51:02,192 --> 00:51:05,313
channel, you have to do a transaction, so it has an impact there. Is there anything else?

837
00:51:07,153 --> 00:51:13,473
Definitely. Things are so complicated here, right? So we've been developing Bitcoin with

838
00:51:13,473 --> 00:51:20,173
with elliptic curve keys in mind for a decade and a half already.

839
00:51:20,653 --> 00:51:23,712
And so we have hierarchical deterministic wallets

840
00:51:23,712 --> 00:51:27,813
use elliptic curve key transformations to derive the keys.

841
00:51:29,873 --> 00:51:33,473
Musig2 and Frost are elliptic curve key aggregation protocols.

842
00:51:34,453 --> 00:51:38,393
We've got lightning point time lock contracts being worked on.

843
00:51:38,453 --> 00:51:39,173
Those are key based.

844
00:51:39,173 --> 00:51:40,532
We've got silent payments,

845
00:51:40,532 --> 00:51:44,492
which is a very elliptic curve specific method of doing more private addresses.

846
00:51:45,273 --> 00:51:47,732
We've got all this stuff that's all based on elliptic curves.

847
00:51:49,032 --> 00:51:51,552
And so absolutely, it impacts everything about Bitcoin.

848
00:51:51,933 --> 00:51:56,732
You know, DLCs, like all these technologies that we rely on for Bitcoin

849
00:51:56,732 --> 00:51:58,393
are based on the elliptic curve math.

850
00:51:58,512 --> 00:52:00,712
And we would have to kind of redevelop them

851
00:52:00,712 --> 00:52:03,192
for whatever type of post-quantum signature we build.

852
00:52:04,492 --> 00:52:05,613
So there's a lot of work coming up.

853
00:52:06,093 --> 00:52:10,413
Okay, so let's get into the juicy topic of an actual softball.

854
00:52:10,413 --> 00:52:15,973
happening like the idea of a soft fork happening to just improve bitcoin or change bitcoin in a way

855
00:52:15,973 --> 00:52:20,913
that you might get like i don't know ctv or whatever it might be these have become already

856
00:52:20,913 --> 00:52:26,113
like really hot button topics like it seems like i don't think bitcoin's ossified but it's it's

857
00:52:26,113 --> 00:52:29,813
always trending more and more towards ossification like these are getting more and more contentious

858
00:52:29,813 --> 00:52:33,453
do you think this would be an easy soft fork to actually implement

859
00:52:33,453 --> 00:52:55,752
I think that the ossification kind of narrative is largely an artifact, I would say, of the past, of the block-size war and of other things.

860
00:52:55,752 --> 00:52:56,813
and

861
00:52:56,813 --> 00:53:00,332
I don't know

862
00:53:00,332 --> 00:53:02,052
I can't predict the future of course

863
00:53:02,052 --> 00:53:04,433
but I think there is

864
00:53:04,433 --> 00:53:06,232
a memetic shift that will happen

865
00:53:06,232 --> 00:53:08,113
when the right combination

866
00:53:08,113 --> 00:53:10,032
of proposal and person

867
00:53:10,032 --> 00:53:11,752
and timing happens

868
00:53:11,752 --> 00:53:13,933
maybe it's this quantum

869
00:53:13,933 --> 00:53:16,332
narrative right now

870
00:53:16,332 --> 00:53:17,893
that creates that shift

871
00:53:17,893 --> 00:53:19,893
so that we see

872
00:53:19,893 --> 00:53:22,512
BIP360 as the kind of best example

873
00:53:22,512 --> 00:53:24,332
right now being implemented as a soft fork

874
00:53:24,332 --> 00:53:29,953
and people seeing kind of a different viewpoint on Bitcoin softworks,

875
00:53:30,052 --> 00:53:33,633
where instead of Bitcoin softworks being the big SegWit change

876
00:53:33,633 --> 00:53:36,593
that added more block space or the big taproot change

877
00:53:36,593 --> 00:53:42,093
that took away certain kind of weird restrictions

878
00:53:42,093 --> 00:53:43,752
in how Bitcoin could be used,

879
00:53:44,153 --> 00:53:48,953
we see instead, oh, Bitcoin is improving in a slow and steady,

880
00:53:49,093 --> 00:53:51,352
responsible way to mitigate future risks

881
00:53:51,352 --> 00:53:53,252
that maintain the security of the system.

882
00:53:53,992 --> 00:53:58,873
And so I think the whole narrative will change at some point because everything is driven on memes.

883
00:53:59,052 --> 00:54:02,352
That's the fundamental learning of our lives on the internet and of Bitcoin.

884
00:54:03,133 --> 00:54:05,373
And the memes will shift at some point.

885
00:54:05,492 --> 00:54:10,832
And instead of there being this ossification narrative, there'll be just the natural thing of, oh, yeah, we're going to solve quantum gradually.

886
00:54:10,832 --> 00:54:12,293
We're going to do BIP 360 today.

887
00:54:12,552 --> 00:54:14,673
We're going to do some other thing tomorrow.

888
00:54:14,673 --> 00:54:18,232
And we're going to add quantum-resistant signatures at some point later when they're ready.

889
00:54:18,593 --> 00:54:20,413
And that's just the obvious natural thing.

890
00:54:20,413 --> 00:54:22,012
So I do think it'll happen.

891
00:54:22,012 --> 00:54:26,633
I don't know if it's going to be easy, but I think we'll just see a shift in memetics and it'll just happen.

892
00:54:27,992 --> 00:54:31,133
So I'm going to put you on the spot and make you do a prediction.

893
00:54:31,332 --> 00:54:34,832
Do you think a quantum soft fork will be the next soft fork?

894
00:54:41,752 --> 00:54:47,873
If we count BIP 360 as a quantum soft fork, which is how it was originally built, I'm going to go with yes.

895
00:54:48,032 --> 00:54:50,712
I would say yes, that is the most likely next soft fork.

896
00:54:50,712 --> 00:54:56,093
I think the other option that's really out there for our next soft work is probably the consensus cleanup.

897
00:54:56,873 --> 00:54:58,552
And I think that's also a good change.

898
00:54:58,593 --> 00:54:59,893
I think it will go at some point.

899
00:55:00,973 --> 00:55:03,992
But it has much more to talk about.

900
00:55:04,252 --> 00:55:06,293
Not even that any of it is wrong or weird.

901
00:55:06,473 --> 00:55:07,913
It's just got more stuff.

902
00:55:08,232 --> 00:55:14,692
Whereas Bit360 is a very focused, very single change that I think makes sense for a whole bunch of reasons.

903
00:55:15,212 --> 00:55:18,433
And so it's going to be easier to have the conversation about Bit360.

904
00:55:18,433 --> 00:55:21,593
So I think it's more likely to be the one that shifts the memetics.

905
00:55:23,273 --> 00:55:26,593
The funny thing with that one would, if there was a huge breakthrough in quantum computing,

906
00:55:27,012 --> 00:55:32,893
and we thought we didn't have three years or two years, however long it takes to migrate

907
00:55:32,893 --> 00:55:37,473
all of these coins, and we needed to put a block size increase in it as well.

908
00:55:37,712 --> 00:55:39,893
Like, that's how this could get really contentious.

909
00:55:41,793 --> 00:55:42,953
Man, that'd be a mess.

910
00:55:43,492 --> 00:55:43,593
Yeah.

911
00:55:43,673 --> 00:55:45,673
And there's so many messy things out there.

912
00:55:45,673 --> 00:55:48,673
because James and Lott proposed confiscation

913
00:55:48,673 --> 00:55:50,393
and we should start confiscating soon.

914
00:55:51,173 --> 00:55:52,573
And that, of course, was like, whoa.

915
00:55:53,093 --> 00:55:56,173
So there's a lot of controversy to be had.

916
00:55:56,273 --> 00:55:58,953
And I think that's why Bit360 is very appealing

917
00:55:58,953 --> 00:55:59,673
as a next software

918
00:55:59,673 --> 00:56:02,252
because it basically dodges all the controversy.

919
00:56:02,532 --> 00:56:04,113
It's just a stepwise improvement.

920
00:56:04,552 --> 00:56:06,673
It's better for certain contracts today.

921
00:56:06,752 --> 00:56:08,212
It's better for both Quantum in the future.

922
00:56:08,773 --> 00:56:11,093
It has support from all different circles.

923
00:56:11,293 --> 00:56:13,113
Lightning developers like it for certain things.

924
00:56:13,332 --> 00:56:14,573
Arc developers like it for certain things.

925
00:56:14,573 --> 00:56:15,433
Everyone likes it.

926
00:56:15,673 --> 00:56:18,913
So I think, yeah, I'll go with yes.

927
00:56:18,973 --> 00:56:20,692
A quantum change, BIP 360 in particular,

928
00:56:20,933 --> 00:56:21,773
is the next soft fork.

929
00:56:22,613 --> 00:56:23,893
That might be bearish on CTV.

930
00:56:24,293 --> 00:56:26,752
Why does Jameson think we should start confiscating coins now?

931
00:56:26,793 --> 00:56:27,492
I've not seen that.

932
00:56:28,832 --> 00:56:35,832
So I think the idea was that we should deploy post-quantum crypto,

933
00:56:36,413 --> 00:56:37,573
and as soon as it's deployed,

934
00:56:37,933 --> 00:56:42,393
we should start a clock that includes confiscation

935
00:56:42,393 --> 00:56:44,332
X, Y, or Z years

936
00:56:44,332 --> 00:56:46,593
after, like, it stops the

937
00:56:46,593 --> 00:56:48,732
creation of new quantum vulnerable coins after three years

938
00:56:48,732 --> 00:56:50,552
and then stops the spending of quantum

939
00:56:50,552 --> 00:56:52,273
vulnerable coins after five years, something like that.

940
00:56:52,852 --> 00:56:54,692
So it wasn't, like, immediate conversation,

941
00:56:54,813 --> 00:56:56,573
but it was, like, as soon as we can do a post-quantum

942
00:56:56,573 --> 00:56:58,532
fork, we should, and as soon as that's done, we should start a

943
00:56:58,532 --> 00:57:00,552
clock. And I think you and I would agree that

944
00:57:00,552 --> 00:57:02,893
we should not do that. If we have a post-quantum

945
00:57:02,893 --> 00:57:04,613
fork, people can voluntarily move their coins,

946
00:57:04,673 --> 00:57:06,393
and the other coins are the quantum attackers.

947
00:57:06,552 --> 00:57:07,873
That's part of the definition of the system.

948
00:57:08,352 --> 00:57:10,712
But at least, it's not as crazy as I first made it out

949
00:57:10,712 --> 00:57:11,373
to be. Sorry, Jameson.

950
00:57:12,393 --> 00:57:17,712
I think when you, if you look at Satoshi's coins, like, I don't know, people estimate

951
00:57:17,712 --> 00:57:22,933
it has a million. They're in a ton of different PacePub key addresses. He talked about quantum

952
00:57:22,933 --> 00:57:28,893
computing, I think. He must have known that this was a potential future thing and he still

953
00:57:28,893 --> 00:57:32,933
left his keys vulnerable. Like, I think you just have to accept that decision that he made.

954
00:57:34,012 --> 00:57:34,893
I don't disagree. Yeah.

955
00:57:35,673 --> 00:57:39,953
It's going to be interesting. Is there anything else on the quantum side that you're like

956
00:57:39,953 --> 00:57:41,453
really paying attention to at the moment?

957
00:57:42,393 --> 00:57:53,673
Well, the one thing I mentioned already of the idea of maybe we don't need logical qubits and maybe that we should be building logical devices rather than logical qubits and then tying those together.

958
00:57:55,732 --> 00:57:57,332
That's something I'm definitely following.

959
00:57:57,852 --> 00:58:08,192
The particular person I was talking to was not totally compelling on that topic, but if that were to be the case, it could reduce one more hurdle in the way of quantum computing ever working.

960
00:58:08,192 --> 00:58:14,773
So, yeah, I think people who know me know I'm open to having my mind changed, but it

961
00:58:14,773 --> 00:58:15,313
takes evidence.

962
00:58:15,552 --> 00:58:20,752
As of now, there's no evidence that quantum is a kind of a current threat.

963
00:58:21,913 --> 00:58:25,852
Yeah, I mean, I've really liked both listening to your work and seeing what you've been putting

964
00:58:25,852 --> 00:58:29,773
on Twitter, because like I said, I was just kind of going along with the narrative that

965
00:58:29,773 --> 00:58:31,473
a load of changes were happening in quantum.

966
00:58:31,593 --> 00:58:32,613
It was getting really close.

967
00:58:33,252 --> 00:58:37,712
And I think listening to you has made me far less worried about this.

968
00:58:37,712 --> 00:58:41,393
I don't think it's a never problem, but I think it's a long way away.

969
00:58:42,352 --> 00:58:43,933
So Bitcoin's going to be fine.

970
00:58:44,052 --> 00:58:46,093
We're going to make changes if we need to.

971
00:58:46,712 --> 00:58:48,313
And it's going to be interesting.

972
00:58:48,413 --> 00:58:49,613
We'll have a lot of content out of it.

973
00:58:51,373 --> 00:58:53,532
The podcast industrial complex will survive.

974
00:58:54,953 --> 00:58:55,273
Exactly.

975
00:58:55,973 --> 00:58:56,992
Brandon, this has been awesome.

976
00:58:57,133 --> 00:58:57,673
Thank you, man.

977
00:58:57,873 --> 00:59:00,832
Is there any way you want to send anyone if they want to check out everything that you're doing?

978
00:59:01,532 --> 00:59:02,313
Yeah, follow me on X.

979
00:59:02,473 --> 00:59:05,732
I'm mostly, as Robin Linus likes to call me, I'm an addict of X.

980
00:59:05,852 --> 00:59:06,712
I'm there all the time.

981
00:59:07,212 --> 00:59:10,992
I post about Bitcoin stuff and other topics there.

982
00:59:11,832 --> 00:59:14,232
As I said earlier, I'm starting a consulting business this year.

983
00:59:14,293 --> 00:59:16,633
So if you need help with anything Bitcoin related, hit me up.

984
00:59:16,773 --> 00:59:22,373
I'm happy to talk about how we can help with your Bitcoin questions and do Bitcoin stuff.

985
00:59:23,332 --> 00:59:24,252
So what are you doing?

986
00:59:24,373 --> 00:59:27,012
Are you targeting like startups and businesses that might need help?

987
00:59:29,453 --> 00:59:30,212
That's the thought.

988
00:59:30,212 --> 00:59:35,552
Mostly, yeah, is if you're writing a new Bitcoin wallet or a new Bitcoin script, I have expertise

989
00:59:35,552 --> 00:59:38,232
in scripting and kind of security modeling

990
00:59:38,232 --> 00:59:39,732
for Bitcoin applications.

991
00:59:40,073 --> 00:59:42,413
So if someone needs some help,

992
00:59:42,473 --> 00:59:44,212
they're either building it or reviewing it.

993
00:59:45,212 --> 00:59:46,352
I want to make myself available.

994
00:59:47,752 --> 00:59:48,153
Awesome.

995
00:59:48,273 --> 00:59:49,393
If you're vibed coding an app

996
00:59:49,393 --> 00:59:50,813
and you actually can't read code,

997
00:59:51,032 --> 00:59:51,532
speak to Brandon.

998
00:59:52,793 --> 00:59:53,653
This has been awesome.

999
00:59:53,773 --> 00:59:54,352
Thank you, man.

1000
00:59:54,752 --> 00:59:56,953
I will definitely follow up with this.

1001
00:59:57,133 --> 00:59:58,153
Hopefully it's in 10 years

1002
00:59:58,153 --> 00:59:59,273
and it's still 10 years away.

1003
00:59:59,953 --> 01:00:01,433
But appreciate the time, man.

1004
01:00:01,453 --> 01:00:01,873
It's been cool.

1005
01:00:03,012 --> 01:00:03,413
It's been great.

1006
01:00:03,473 --> 01:00:03,732
Thanks, Danny.

1007
01:00:05,552 --> 01:00:07,613
you