1
00:00:00,000 --> 00:00:03,340
special Thursday edition

2
00:00:03,340 --> 00:00:04,760
of the Presidio Bitcoin Jam.

3
00:00:04,760 --> 00:00:05,080
Thursday night,

4
00:00:05,500 --> 00:00:06,260
not football,

5
00:00:06,500 --> 00:00:07,040
not the night,

6
00:00:07,220 --> 00:00:08,560
but Thursday night PBJ.

7
00:00:08,760 --> 00:00:09,940
An emergency jam.

8
00:00:10,160 --> 00:00:10,900
An emergency jam.

9
00:00:10,920 --> 00:00:11,560
And the emergency is

10
00:00:11,560 --> 00:00:11,980
we're all here

11
00:00:11,980 --> 00:00:13,140
and it's a good,

12
00:00:13,300 --> 00:00:13,840
vulnerable time.

13
00:00:14,040 --> 00:00:15,320
Well, we missed last week.

14
00:00:15,420 --> 00:00:15,900
We missed last week.

15
00:00:16,160 --> 00:00:17,640
So we're missing the show.

16
00:00:17,820 --> 00:00:17,920
Yeah.

17
00:00:18,080 --> 00:00:19,100
So we're here.

18
00:00:19,220 --> 00:00:20,960
And then we are at

19
00:00:20,960 --> 00:00:22,740
the Presidio Bitcoin

20
00:00:22,740 --> 00:00:24,120
Open Source AI Summit.

21
00:00:24,280 --> 00:00:24,500
Yep.

22
00:00:24,860 --> 00:00:25,180
Today.

23
00:00:25,500 --> 00:00:25,660
Yeah.

24
00:00:25,660 --> 00:00:26,060
And tomorrow.

25
00:00:26,520 --> 00:00:26,700
Yep.

26
00:00:27,040 --> 00:00:28,460
So no jam tomorrow.

27
00:00:29,000 --> 00:00:29,880
So no jam tomorrow.

28
00:00:30,000 --> 00:00:31,620
But we're like, hey, let's do it.

29
00:00:31,740 --> 00:00:35,160
Let's do an impromptu evening jam.

30
00:00:35,440 --> 00:00:35,700
That's right.

31
00:00:36,040 --> 00:00:39,100
And DK's in town, so we got to take advantage to be back in the studio.

32
00:00:39,180 --> 00:00:39,660
And next week, too.

33
00:00:40,160 --> 00:00:41,100
Oh, even better.

34
00:00:41,280 --> 00:00:41,520
All right.

35
00:00:41,760 --> 00:00:42,200
So tune around.

36
00:00:42,440 --> 00:00:44,340
So we'll have to make sure we come to the studio.

37
00:00:44,540 --> 00:00:44,940
There we go.

38
00:00:45,160 --> 00:00:45,540
Yeah, yeah.

39
00:00:45,740 --> 00:00:46,080
I'm here.

40
00:00:46,260 --> 00:00:46,760
I'm back, too.

41
00:00:47,380 --> 00:00:53,680
And we're going to try to keep it to, well, maybe 47 minutes until 6 or so Pacific.

42
00:00:53,800 --> 00:00:54,260
But anyway.

43
00:00:54,940 --> 00:00:56,540
We have a lot to cover, though, right?

44
00:00:56,880 --> 00:00:58,580
We definitely want to talk about today's event.

45
00:00:58,580 --> 00:00:59,080
Oh, yeah.

46
00:00:59,080 --> 00:01:02,1000
I want to touch on Blockstream Liquid a bit

47
00:01:02,1000 --> 00:01:04,340
because that was a huge hack.

48
00:01:05,460 --> 00:01:06,320
Interesting dynamic.

49
00:01:06,460 --> 00:01:08,180
There's some lessons learned there

50
00:01:08,180 --> 00:01:10,640
and some interesting conversation about that.

51
00:01:10,720 --> 00:01:11,380
You want to start there?

52
00:01:11,480 --> 00:01:12,940
You want to start with...

53
00:01:12,940 --> 00:01:13,980
Yeah, and then let's time boxes

54
00:01:13,980 --> 00:01:15,060
so we don't spend the whole show

55
00:01:15,060 --> 00:01:16,200
talking about Blockstream Liquid.

56
00:01:16,400 --> 00:01:18,380
But last weekend, Labor Day weekend,

57
00:01:20,500 --> 00:01:24,640
so for those that don't know what Blockstream Liquid is,

58
00:01:24,640 --> 00:01:25,640
it is a...

59
00:01:26,600 --> 00:01:27,740
I don't know.

60
00:01:27,740 --> 00:01:29,320
the controversy about it's a sidechain,

61
00:01:29,400 --> 00:01:32,460
but it's a system in which you can deposit Bitcoin into it.

62
00:01:32,900 --> 00:01:34,160
It's a federated system

63
00:01:34,160 --> 00:01:37,980
where I believe they have like 15 different entities,

64
00:01:38,680 --> 00:01:40,060
each hold a key,

65
00:01:40,720 --> 00:01:42,520
11 of the 15 have to sign

66
00:01:42,520 --> 00:01:47,120
to spend out of that system.

67
00:01:47,120 --> 00:01:49,460
Do you know who are the people who,

68
00:01:49,620 --> 00:01:51,660
like who wants to put money in there?

69
00:01:51,960 --> 00:01:54,080
Are there certain use cases where that kind of thing?

70
00:01:54,140 --> 00:01:56,280
It does support confidential transactions,

71
00:01:56,280 --> 00:01:59,580
So if really good privacy is important to you,

72
00:02:00,360 --> 00:02:01,780
that would be a reason you use it.

73
00:02:02,500 --> 00:02:05,860
You need your counterparty you're paying to also be using that?

74
00:02:06,200 --> 00:02:07,280
Is that what you're doing?

75
00:02:07,900 --> 00:02:11,940
So when you deposit real Bitcoin into liquid,

76
00:02:12,160 --> 00:02:15,940
you get LLBDC or liquid BTC token,

77
00:02:16,760 --> 00:02:19,060
which then you can do things within that system.

78
00:02:19,220 --> 00:02:20,260
And then whenever you want out,

79
00:02:20,260 --> 00:02:29,940
You can exit as long as the 11 of the 15 entities in the federation sign your transaction and you can exit.

80
00:02:31,040 --> 00:02:33,380
I think there's some stable coins on there.

81
00:02:33,620 --> 00:02:42,020
I think like Aqua Wallet and Mana Wallet and a few other wallets use it as well as like a scaling solution.

82
00:02:42,320 --> 00:02:45,380
Reach more users, lower fees, etc.

83
00:02:45,380 --> 00:02:49,820
And interestingly, you can hook up, for example, like Lightning on top of Liquid as well.

84
00:02:49,820 --> 00:02:53,760
So you can have potentially swaps between Bitcoin and Lightning via the Lightning Network too.

85
00:02:53,960 --> 00:02:55,240
I don't know if that's popular.

86
00:02:55,520 --> 00:02:59,620
Yeah, I mean, this has been a product Blockstream's worked on for literally like 10 years.

87
00:03:00,620 --> 00:03:08,240
It's also like kind of a testbed of new Bitcoin concepts that haven't yet gotten consensus or enough traction.

88
00:03:10,260 --> 00:03:15,740
Since Blockstream controls the Liquid system, they're able to add new features.

89
00:03:15,1000 --> 00:03:19,100
So it's kind of an interesting experimental testbed as well.

90
00:03:19,100 --> 00:03:23,920
In any case, I didn't know how much money was in Liquid, but I've now said...

91
00:03:23,920 --> 00:03:24,440
We know now.

92
00:03:24,600 --> 00:03:25,500
How much money was not in Liquid?

93
00:03:25,500 --> 00:03:32,300
There's like 4,000, around 4,200 BTC in Liquid.

94
00:03:33,840 --> 00:03:37,640
Something like $350 to $400 million of value in Liquid.

95
00:03:37,800 --> 00:03:39,040
So substantial amount of money.

96
00:03:39,080 --> 00:03:39,880
Yeah, it's more than I thought to, actually.

97
00:03:40,540 --> 00:03:46,420
And then, so the event last weekend, all of a sudden, there were red alerts on social media

98
00:03:46,420 --> 00:03:53,900
saying that 4,000 Bitcoin were withdrawn from Liquid

99
00:03:53,900 --> 00:03:56,240
to one address, to one actor,

100
00:03:56,920 --> 00:04:02,660
which was certainly not an expected transaction to occur.

101
00:04:02,760 --> 00:04:04,300
So then it's like, oh my God, there's been a theft.

102
00:04:04,880 --> 00:04:06,440
How did it occur, right?

103
00:04:06,440 --> 00:04:08,940
And so, I mean, what we know now is

104
00:04:08,940 --> 00:04:15,600
there was a vulnerability in the consensus algorithm

105
00:04:16,560 --> 00:04:22,320
that allowed for inflation so lbtc was created that shouldn't have been created

106
00:04:22,880 --> 00:04:30,400
uh and this uh attacker like learned of this created 4 000 lbcc out of thin air then formed

107
00:04:30,400 --> 00:04:37,760
a valid transaction so the uh the signers the 11 signers had to to sign it are like looked at it and

108
00:04:37,760 --> 00:04:43,440
this i mean their software looked at it and it's like it's valid and so it exited and then it exited

109
00:04:43,440 --> 00:04:43,940
to,

110
00:04:44,680 --> 00:04:47,520
the money can only exit

111
00:04:47,520 --> 00:04:48,080
to

112
00:04:48,080 --> 00:04:51,080
certain entities. I think the

113
00:04:51,080 --> 00:04:53,380
same or some, either a

114
00:04:53,380 --> 00:04:55,460
superset or the same entities that

115
00:04:55,460 --> 00:04:57,700
are the functionaries

116
00:04:57,700 --> 00:04:58,880
of this federation.

117
00:05:00,100 --> 00:05:01,540
And then what I've

118
00:05:01,540 --> 00:05:03,380
learned is I think many or most

119
00:05:03,380 --> 00:05:05,000
have

120
00:05:05,000 --> 00:05:07,340
certain policies in place when it is

121
00:05:07,340 --> 00:05:09,240
withdrawn. Like you, like

122
00:05:09,240 --> 00:05:11,160
there might be velocity

123
00:05:11,160 --> 00:05:13,420
controls where you can only withdraw

124
00:05:13,420 --> 00:05:18,280
so much out of their, I mean, once it's gone out of liquid into their own Bitcoin hot wallet

125
00:05:18,280 --> 00:05:25,720
or cold wallet, there might be velocity controls on how fast it can be withdrawn out of that

126
00:05:25,720 --> 00:05:27,440
wallet to the ultimate recipient.

127
00:05:27,440 --> 00:05:32,040
And the signers themselves are some sort of like a, they're sort of like a financial services

128
00:05:32,040 --> 00:05:34,860
or a Bitcoin focused financial service, like an exchange?

129
00:05:35,200 --> 00:05:37,840
In theory, one thing though, it's never.

130
00:05:38,080 --> 00:05:38,720
It's not public, right?

131
00:05:38,720 --> 00:05:39,300
It's not public.

132
00:05:39,400 --> 00:05:39,500
Yeah.

133
00:05:39,500 --> 00:05:40,940
They used to be public, but they're not public.

134
00:05:41,020 --> 00:05:41,780
So we actually don't know.

135
00:05:41,780 --> 00:05:44,020
You can guess a lot of the major players.

136
00:05:44,180 --> 00:05:44,740
You can guess.

137
00:05:44,880 --> 00:05:46,240
Probably Tether was one,

138
00:05:46,520 --> 00:05:47,620
and maybe Breeze was one.

139
00:05:47,700 --> 00:05:48,000
I don't know.

140
00:05:48,480 --> 00:05:49,680
We don't know who they are.

141
00:05:50,520 --> 00:05:51,220
That was intentional.

142
00:05:51,440 --> 00:05:52,340
There's a trade-off there.

143
00:05:52,820 --> 00:05:55,180
I think they want to reduce regulatory exposure.

144
00:05:55,460 --> 00:05:57,600
That's why they wanted to be more on the down low.

145
00:05:58,600 --> 00:06:00,020
But then as a user of the system,

146
00:06:00,380 --> 00:06:01,520
it's way less transparent.

147
00:06:02,360 --> 00:06:04,760
In the most adversarial-minded case,

148
00:06:05,000 --> 00:06:06,280
it could just be like Blockstream

149
00:06:06,280 --> 00:06:08,140
holds all 15 keys.

150
00:06:08,140 --> 00:06:09,020
It's just one.

151
00:06:09,020 --> 00:06:13,700
And, you know, so, I mean, I don't think that's actually the case, but, you know, it could be.

152
00:06:13,700 --> 00:06:21,140
So anyway, but it turns out there was a service called Sideswap, which was one of the functionaries.

153
00:06:22,080 --> 00:06:27,620
And they did not implement KYC or velocity controls.

154
00:06:28,200 --> 00:06:38,600
So if you use their service to swap out, then you withdraw LBDC to real Bitcoin through the 11, and that's an 11 and 15 multisig.

155
00:06:38,600 --> 00:06:40,880
then that went into a side swap

156
00:06:40,880 --> 00:06:41,680
wallet

157
00:06:41,680 --> 00:06:44,960
and then from there you can withdraw

158
00:06:44,960 --> 00:06:47,000
to your own wallet and all that can be

159
00:06:47,000 --> 00:06:48,860
done instantly, no velocity

160
00:06:48,860 --> 00:06:50,940
controls, no KYC. So that's the

161
00:06:50,940 --> 00:06:52,100
path the attacker

162
00:06:52,100 --> 00:06:53,200
took

163
00:06:53,200 --> 00:06:56,680
to steal

164
00:06:56,680 --> 00:06:58,760
these Bitcoin. Now I'm calling it

165
00:06:58,760 --> 00:07:00,200
stealing and theft.

166
00:07:00,740 --> 00:07:02,680
This attacker left

167
00:07:02,680 --> 00:07:04,320
a message in the Bitcoin blockchain

168
00:07:04,320 --> 00:07:06,560
claiming that they're a white hat

169
00:07:06,560 --> 00:07:07,920
hacker initially.

170
00:07:07,920 --> 00:07:10,620
That was with the move of the 4,000.

171
00:07:11,340 --> 00:07:12,200
I'm White Hat.

172
00:07:12,340 --> 00:07:13,200
We're just here to help you.

173
00:07:13,700 --> 00:07:14,460
I'm a White Hat hacker.

174
00:07:15,120 --> 00:07:17,020
I think it was White Hat hacker is how it was described,

175
00:07:17,080 --> 00:07:17,780
but definitely White Hat.

176
00:07:18,520 --> 00:07:22,280
And at the time, no one knew, right?

177
00:07:22,360 --> 00:07:24,880
All we knew is that 4,000 Bitcoin were just withdrawn.

178
00:07:25,280 --> 00:07:26,200
Holy shit.

179
00:07:26,860 --> 00:07:27,1000
And it says White Hat hacker.

180
00:07:28,1000 --> 00:07:30,960
And then what transpired was fascinating

181
00:07:30,960 --> 00:07:33,700
because then there was an exchange back and forth

182
00:07:33,700 --> 00:07:36,680
over the Bitcoin blockchain in op returns

183
00:07:36,680 --> 00:07:39,180
between Blockstream and this attacker

184
00:07:39,180 --> 00:07:41,360
to communicate.

185
00:07:41,520 --> 00:07:42,700
And some of that was in plain text

186
00:07:42,700 --> 00:07:44,060
and some of it was encrypted

187
00:07:44,060 --> 00:07:46,620
where only those two parties

188
00:07:46,620 --> 00:07:48,000
could see the messages.

189
00:07:48,480 --> 00:07:50,700
So we don't know everything that was exchanged.

190
00:07:51,020 --> 00:07:54,520
But to cut a long story short,

191
00:07:55,800 --> 00:08:00,540
the attacker did return 3,400 BTC.

192
00:08:01,480 --> 00:08:03,460
So a huge portion,

193
00:08:03,460 --> 00:08:07,220
but still kept 15% or 600 BTC.

194
00:08:09,440 --> 00:08:12,260
And, you know, because a lot of the communication we can't see,

195
00:08:12,360 --> 00:08:15,840
it's unclear if that's just like the next step of the negotiation

196
00:08:15,840 --> 00:08:18,340
or from the attacker's perspective, that's just like,

197
00:08:19,260 --> 00:08:21,400
they're claiming that's their bounty.

198
00:08:21,400 --> 00:08:22,400
$50 million bounty.

199
00:08:24,660 --> 00:08:28,460
And as far as I know, that's sort of where things stand.

200
00:08:28,780 --> 00:08:31,000
And there's been some drama back and forth, no, between them.

201
00:08:31,000 --> 00:08:36,660
Yeah, well, and then it's also interesting to see the chatter and commentary on the situation.

202
00:08:36,860 --> 00:08:47,400
There's some people on Twitter and in the Bitcoin space that cheer this person on and don't think it's criminal and they endorse it.

203
00:08:47,600 --> 00:08:50,740
And I find that difficult to swallow.

204
00:08:50,740 --> 00:08:55,100
if you take someone's property

205
00:08:55,100 --> 00:08:57,600
and then you decide that you're going to keep 15%

206
00:08:57,600 --> 00:08:58,260
and give it back to them

207
00:08:58,260 --> 00:09:02,300
because they didn't have the most superior lock on their door

208
00:09:02,300 --> 00:09:02,740
or whatever

209
00:09:02,740 --> 00:09:07,980
I'm sorry but that's still property rights violation and theft

210
00:09:07,980 --> 00:09:12,860
I think they went far beyond

211
00:09:12,860 --> 00:09:15,680
what any sensible definition of white hat hacker is

212
00:09:15,680 --> 00:09:19,080
just from a purely ethical standpoint

213
00:09:19,080 --> 00:09:21,040
the legal system can deal with it as well.

214
00:09:21,120 --> 00:09:22,500
And I'm quite certain the legal system

215
00:09:22,500 --> 00:09:24,980
would call that a criminal case.

216
00:09:25,120 --> 00:09:26,000
Yeah, I mean, I'm curious.

217
00:09:26,160 --> 00:09:28,320
I think the sort of the other side would be,

218
00:09:28,520 --> 00:09:29,080
and we don't know

219
00:09:29,080 --> 00:09:31,240
because a lot of this was encrypted communications,

220
00:09:31,500 --> 00:09:34,740
but if I guess Liquid or Blockstream

221
00:09:34,740 --> 00:09:35,880
agreed to those numbers,

222
00:09:35,960 --> 00:09:37,120
now one can say, you know,

223
00:09:37,140 --> 00:09:39,560
it's one thing to negotiate when you're sitting there.

224
00:09:39,660 --> 00:09:40,720
Hey, I've got all of your money.

225
00:09:40,900 --> 00:09:43,280
I'll give you back 85% if you agree not to prosecute,

226
00:09:43,580 --> 00:09:45,580
which again kind of changes it from white to gray

227
00:09:45,580 --> 00:09:47,420
because it's, I mean, the cynical look on that, right,

228
00:09:47,420 --> 00:09:51,080
is if the hacker didn't have that

229
00:09:51,080 --> 00:09:51,700
and they sort of said,

230
00:09:51,780 --> 00:09:53,300
hey, I've got a bug bounty ahead of time,

231
00:09:53,420 --> 00:09:56,220
would you give me $50 million to not reveal?

232
00:09:56,620 --> 00:09:58,000
That's a very different negotiation.

233
00:09:59,000 --> 00:10:00,000
Another thing I learned,

234
00:10:00,120 --> 00:10:01,240
I mean, it would have been my intuition,

235
00:10:01,360 --> 00:10:01,900
but I learned also,

236
00:10:02,460 --> 00:10:05,280
the criminal case is not impacted by

237
00:10:05,280 --> 00:10:06,780
if Blockstream agrees.

238
00:10:07,040 --> 00:10:08,300
There could still be a criminal case.

239
00:10:08,480 --> 00:10:09,260
Interest of theft.

240
00:10:09,720 --> 00:10:12,440
Why is that if they gave consent to keep it back?

241
00:10:12,580 --> 00:10:14,380
Or I guess, would that be a civil case or something?

242
00:10:15,000 --> 00:10:16,020
I suspect it's because

243
00:10:16,020 --> 00:10:17,740
when you're under duress like that.

244
00:10:17,860 --> 00:10:20,580
I mean, the criminal law is sort of independent

245
00:10:20,580 --> 00:10:22,960
of the duress of the victim.

246
00:10:24,120 --> 00:10:26,500
But in any case, that's where that lands.

247
00:10:26,840 --> 00:10:29,040
Now, another interesting tidbit is,

248
00:10:29,300 --> 00:10:31,200
we just got done speaking.

249
00:10:31,660 --> 00:10:33,540
We just wrapped up day one of the summit.

250
00:10:34,260 --> 00:10:36,400
And a panel that I was on with three other folks,

251
00:10:36,400 --> 00:10:38,840
we talked about AI security scanning.

252
00:10:39,980 --> 00:10:45,200
And, well, it turns out that the red team,

253
00:10:45,200 --> 00:10:48,860
So we had CEO Rob Hamilton on my panel.

254
00:10:49,060 --> 00:10:49,480
Super CEO.

255
00:10:49,620 --> 00:10:51,240
We love talking about him on the show.

256
00:10:51,400 --> 00:10:52,100
And super CEO.

257
00:10:53,420 --> 00:10:54,640
Doesn't really know how to code.

258
00:10:55,640 --> 00:10:56,320
Just kidding, Rob.

259
00:10:57,220 --> 00:10:58,600
But he jokes himself.

260
00:10:58,740 --> 00:11:01,320
He's a security professional now in the past six weeks, right?

261
00:11:01,500 --> 00:11:03,200
Because ever since the cold card incident,

262
00:11:03,460 --> 00:11:06,860
he jumped into gear and has been burning the midnight oil,

263
00:11:07,300 --> 00:11:10,820
doing security scanning of many open source Bitcoin projects.

264
00:11:11,200 --> 00:11:13,700
Many other Bitcoiners joined him.

265
00:11:13,700 --> 00:11:36,040
So there were folks from the red team that found a vulnerability in Liquid, reported it, and that was the one, it's public information, that was a vulnerability dating back to like 2018 or 2019.

266
00:11:36,040 --> 00:11:39,100
so it's been in the code for a long, long time.

267
00:11:40,920 --> 00:11:45,700
And then the fix for that

268
00:11:45,700 --> 00:11:48,720
introduced a worse vulnerability.

269
00:11:49,220 --> 00:11:50,720
It was like easier to exploit.

270
00:11:51,160 --> 00:11:53,960
It was a similar type of...

271
00:11:53,960 --> 00:11:55,300
So this has to do with the...

272
00:11:55,300 --> 00:11:57,760
It actually has to do with the confidential transactions

273
00:11:57,760 --> 00:11:59,660
and the range proofs.

274
00:12:00,520 --> 00:12:02,440
Anyway, I won't get into the technical details,

275
00:12:02,600 --> 00:12:04,640
but all good reason it's not implemented on Bitcoin.

276
00:12:04,640 --> 00:12:10,280
But the fix didn't, or anyway, the fix introduced a worse bug and that's what was exploited.

277
00:12:10,280 --> 00:12:19,200
And there's drama on Twitter today, you know, where I think the red team saying they also

278
00:12:19,200 --> 00:12:35,925
reported the correct fix And then there a dispute on whether that was looked at or not by Blockstream Reported the correct fix ahead of time Yeah Or meaning like explicitly noted the wrong fix was a concern

279
00:12:36,085 --> 00:12:36,905
and here's the right thing.

280
00:12:36,945 --> 00:12:38,185
There's, like, a second email.

281
00:12:38,365 --> 00:12:38,585
I see.

282
00:12:39,225 --> 00:12:41,625
So we'll see how that transpires.

283
00:12:41,625 --> 00:12:43,585
But that was the actual vulnerability.

284
00:12:43,845 --> 00:12:46,085
So, again, it gets back to this AI security scanning.

285
00:12:46,305 --> 00:12:49,205
Like, AI is able to find these vulnerabilities so quickly and easily.

286
00:12:49,205 --> 00:12:52,865
another thing that arose from this that I found

287
00:12:52,865 --> 00:12:55,545
interesting is like the people were criticizing

288
00:12:55,545 --> 00:12:58,965
federations and the value of federations and it actually got me

289
00:12:58,965 --> 00:13:01,785
to thinking in ways I hadn't really thought about before

290
00:13:01,785 --> 00:13:05,085
because in general I just naively thought like

291
00:13:05,085 --> 00:13:07,665
you know federation gives you more security

292
00:13:07,665 --> 00:13:11,105
I'm not just trusting you but I'm you know it's you

293
00:13:11,105 --> 00:13:13,485
you and a couple others

294
00:13:13,485 --> 00:13:16,525
Well, no, even if...

295
00:13:16,525 --> 00:13:20,105
Let's say that we know that it's distinct entities, right?

296
00:13:20,205 --> 00:13:21,805
I mean, even if we...

297
00:13:21,805 --> 00:13:23,925
Let's just hold that as a given.

298
00:13:24,965 --> 00:13:26,645
But that's not necessarily the case.

299
00:13:26,745 --> 00:13:31,285
So in this situation, there's 15 different entities.

300
00:13:32,405 --> 00:13:34,105
But as far as we can tell,

301
00:13:34,105 --> 00:13:35,865
they would...

302
00:13:35,865 --> 00:13:38,345
Whatever the software releases that Blockstream produced,

303
00:13:38,425 --> 00:13:39,425
they just updated.

304
00:13:39,745 --> 00:13:41,045
They didn't like...

305
00:13:41,045 --> 00:13:42,305
And I don't know if this is fact,

306
00:13:42,305 --> 00:13:44,885
but it appears to be the case.

307
00:13:45,365 --> 00:13:47,105
They just update the software,

308
00:13:47,565 --> 00:13:48,185
trust Blockstream,

309
00:13:48,565 --> 00:13:50,185
and don't audit it themselves,

310
00:13:50,585 --> 00:13:54,325
don't do any kind of checking there.

311
00:13:54,325 --> 00:13:56,745
And so that clearly,

312
00:13:57,205 --> 00:13:59,765
there's many aspects of centralization,

313
00:13:59,925 --> 00:14:02,365
but that aspect would then be centralized

314
00:14:02,365 --> 00:14:03,605
because Blockstream,

315
00:14:04,265 --> 00:14:05,025
I mean, it ended up being,

316
00:14:05,245 --> 00:14:07,325
one developer did the fix,

317
00:14:07,405 --> 00:14:08,525
one reviewed it, it got merged,

318
00:14:08,685 --> 00:14:11,005
and then my guess is no one else looked at it.

319
00:14:11,005 --> 00:14:13,925
And they just sort of accepted the software.

320
00:14:14,865 --> 00:14:16,425
So that seems pretty weak.

321
00:14:16,605 --> 00:14:20,365
And then another weakness that was revealed

322
00:14:20,365 --> 00:14:25,865
was this, again, getting back to when Bitcoin's withdrawn from Liquid,

323
00:14:26,725 --> 00:14:29,585
I learned that it's like best practices

324
00:14:29,585 --> 00:14:32,825
that are given to the entities or the functionaries or whatever

325
00:14:32,825 --> 00:14:36,085
to do spending velocity controls,

326
00:14:36,645 --> 00:14:38,665
maybe do KYC as well,

327
00:14:38,665 --> 00:14:40,385
but put these protections in place.

328
00:14:40,385 --> 00:14:41,945
cold store.

329
00:14:42,285 --> 00:14:44,465
It goes from liquid to their system.

330
00:14:44,565 --> 00:14:46,165
But it's a holding period

331
00:14:46,165 --> 00:14:48,525
and there's controls in place

332
00:14:48,525 --> 00:14:50,485
such that

333
00:14:50,485 --> 00:14:52,365
the entire liquidity

334
00:14:52,365 --> 00:14:54,125
in liquid can't be exited

335
00:14:54,125 --> 00:14:56,405
in one transaction all

336
00:14:56,405 --> 00:14:56,885
in one go.

337
00:14:58,645 --> 00:15:00,285
But one of the functionary

338
00:15:00,285 --> 00:15:01,985
side swap wasn't doing that.

339
00:15:02,345 --> 00:15:03,785
So there was

340
00:15:03,785 --> 00:15:06,385
a weakness there as well.

341
00:15:06,745 --> 00:15:08,445
So I don't know what the right answer is.

342
00:15:08,525 --> 00:15:09,905
Should the liquid protocol

343
00:15:09,905 --> 00:15:12,025
enforce these, like, add spending

344
00:15:12,025 --> 00:15:14,125
velocity controls, and if so, what

345
00:15:14,125 --> 00:15:16,265
should they be? That gets tricky, you know.

346
00:15:16,705 --> 00:15:18,045
And I mean, you know,

347
00:15:18,085 --> 00:15:20,085
adding KYC, I don't think

348
00:15:20,085 --> 00:15:21,705
it's probably the right answer to the protocol.

349
00:15:23,065 --> 00:15:24,025
But it raised that

350
00:15:24,025 --> 00:15:25,705
issue as well. And then, like I said,

351
00:15:25,885 --> 00:15:27,725
some people tweeting, comparing, like,

352
00:15:28,065 --> 00:15:29,705
custodians to federations, and

353
00:15:29,705 --> 00:15:31,565
you know,

354
00:15:31,785 --> 00:15:33,925
coming out of some formula where, like, you know,

355
00:15:33,925 --> 00:15:35,085
if, like,

356
00:15:35,685 --> 00:15:37,685
two-thirds of the

357
00:15:37,685 --> 00:15:39,845
entities in your federation are worse

358
00:15:39,845 --> 00:15:42,525
lower quality than like one custodian,

359
00:15:42,585 --> 00:15:43,805
you're better off with just one custodian.

360
00:15:43,885 --> 00:15:44,125
Yes.

361
00:15:45,005 --> 00:15:46,545
Not exactly what you said in the past,

362
00:15:46,605 --> 00:15:49,405
but it's in the same direction as what you said.

363
00:15:49,525 --> 00:15:50,485
Yeah, and I actually, I mean,

364
00:15:50,505 --> 00:15:51,925
I was going to kind of make that point again.

365
00:15:52,045 --> 00:15:52,845
I mean, to me,

366
00:15:53,525 --> 00:15:55,425
I think there's a lot of interesting points in this,

367
00:15:55,425 --> 00:15:56,805
but, you know,

368
00:15:57,365 --> 00:15:59,045
I actually think Saylor actually put that

369
00:15:59,045 --> 00:16:01,425
and he had that nice like tweet storm,

370
00:16:02,345 --> 00:16:03,125
I guess, I don't know if it was like

371
00:16:03,125 --> 00:16:04,285
the Bitcoin maximum or whatever.

372
00:16:04,765 --> 00:16:06,005
Sorry, where he made the point,

373
00:16:06,065 --> 00:16:09,085
it's like, I think we from the early days of Bitcoin,

374
00:16:09,085 --> 00:16:13,185
you know, maybe took a lesson that was important, but made it the only lesson, which is not your

375
00:16:13,185 --> 00:16:18,625
keys, not your coins, or like, you know, um, the, uh, right to hold keys is an obligation to hold

376
00:16:18,625 --> 00:16:24,205
keys. And I think, you know, the way I'm seeing this move increasingly is like, it's not that like

377
00:16:24,205 --> 00:16:28,905
you should never trust a custodian or, you know, take everything yourself, uh, or never trust a

378
00:16:28,905 --> 00:16:32,785
side chain ever. Like all side chains are bad or all lightning network nodes are bad or whatever.

379
00:16:32,785 --> 00:16:36,985
I think the, the point this all raises to me is like in an open system with free competition

380
00:16:36,985 --> 00:16:44,165
and transparency, you should evaluate and really see which custodians and counterparties you trust.

381
00:16:44,405 --> 00:16:46,865
And obviously, I'm still saying you should have the right to hold keys if you want. But for most

382
00:16:46,865 --> 00:16:50,785
people, which custodians you trust, which side chains, which lightning implementations, which

383
00:16:50,785 --> 00:16:54,825
lightning network nodes. And I think one of the big takeaways that's been interesting for me is I

384
00:16:54,825 --> 00:16:59,485
think as a community, you know, I don't think it's a bad thing. You know, and look, we're, you know,

385
00:16:59,925 --> 00:17:03,365
we're the canary in the coal mine. We're, you know, we're the ones taking the AI stuff first.

386
00:17:03,365 --> 00:17:10,565
but um you know i think that a lot of this has come out of sort of like sort of hacker culture

387
00:17:10,565 --> 00:17:14,925
side projects maybe companies that you know like i'd be curious within blockstream and not to like

388
00:17:14,925 --> 00:17:19,165
call them out or whatever but like um i'm guessing liquid was not their number one focus at this time

389
00:17:19,165 --> 00:17:22,605
um and it's like you know if you're a company where maybe you at one point you were focused

390
00:17:22,605 --> 00:17:26,285
on something and then you know it's not like a fault of your own but like priorities shifted

391
00:17:26,285 --> 00:17:29,745
there's new business opportunities and maybe there's just like one person kind of like left

392
00:17:29,745 --> 00:17:33,985
on the maintenance staff, then, you know, if that's still securing serious funds, then that's

393
00:17:33,985 --> 00:17:38,885
a problem. So I, I think, you know, where I hope all this is going to go in the open system is

394
00:17:38,885 --> 00:17:44,085
the services that people use and trust. Hopefully there's an easier way for people to, to, you know,

395
00:17:44,085 --> 00:17:47,485
to realize like cussing with block or using something that came out of block software,

396
00:17:47,485 --> 00:17:53,665
which has, you know, whatever, a lot of really qualified people, top of the line models is not

397
00:17:53,665 --> 00:17:57,725
the same as like the weekend hacker project, the story in 4,000 of your BTC. And there's a wide

398
00:17:57,725 --> 00:17:58,485
spectrum in between.

399
00:18:00,345 --> 00:18:02,105
I'm not saying that blockchain was that, but

400
00:18:02,105 --> 00:18:04,065
the last

401
00:18:04,065 --> 00:18:06,005
aspect that got brought up

402
00:18:06,005 --> 00:18:07,865
that didn't really

403
00:18:07,865 --> 00:18:09,705
get enough discussion, I don't think, but someone

404
00:18:09,705 --> 00:18:11,105
raised the point about liability

405
00:18:11,105 --> 00:18:13,685
from the entities. So there's 15

406
00:18:13,685 --> 00:18:14,745
different companies

407
00:18:14,745 --> 00:18:16,965
and

408
00:18:16,965 --> 00:18:19,545
it's unclear how much liability

409
00:18:19,545 --> 00:18:21,445
each of those companies has

410
00:18:21,445 --> 00:18:22,905
with this staff.

411
00:18:23,525 --> 00:18:25,365
Could there be lawsuits for people who

412
00:18:25,365 --> 00:18:26,425
lost money?

413
00:18:27,725 --> 00:18:31,305
And if that were ever the case

414
00:18:31,305 --> 00:18:33,845
and it was established that there is liability with those entities,

415
00:18:34,645 --> 00:18:36,385
is there enough value in it for them?

416
00:18:36,825 --> 00:18:40,565
Would that dissuade anyone from being part of the Federation?

417
00:18:40,565 --> 00:18:42,405
To being a signatory or whatever?

418
00:18:43,065 --> 00:18:45,765
Is there any benefit to running one of those?

419
00:18:46,185 --> 00:18:49,725
I mean, mainly that you at least get your sort of vote

420
00:18:49,725 --> 00:18:54,145
in terms of like, you might be using the system,

421
00:18:54,405 --> 00:18:56,445
so you want honest actors.

422
00:18:56,445 --> 00:18:59,185
so you want to at least be one of them.

423
00:19:00,285 --> 00:19:01,645
I mean, Spark has the same,

424
00:19:02,705 --> 00:19:05,365
Spark is federated, FedAdmin is federated.

425
00:19:05,465 --> 00:19:07,105
Those are two other systems in the ecosystem

426
00:19:07,105 --> 00:19:08,545
that have federations.

427
00:19:08,725 --> 00:19:10,625
And so with Spark, there's one entity,

428
00:19:10,785 --> 00:19:11,525
which is LightSpark,

429
00:19:11,925 --> 00:19:13,185
but then there's Spark operators.

430
00:19:14,085 --> 00:19:16,405
And I don't know where they're at now,

431
00:19:16,445 --> 00:19:17,265
but they started with two

432
00:19:17,265 --> 00:19:19,385
and now I think they're at three.

433
00:19:19,505 --> 00:19:20,305
Maybe they have more now,

434
00:19:20,305 --> 00:19:23,665
but there'd be the same question there.

435
00:19:23,665 --> 00:19:26,525
why do you want to do it?

436
00:19:26,925 --> 00:19:29,385
Now, with the way Spark works,

437
00:19:29,945 --> 00:19:31,605
let's say we built a wallet

438
00:19:31,605 --> 00:19:34,925
and it was a Spark-based wallet.

439
00:19:35,245 --> 00:19:37,845
Our incentive to be an operator in Spark

440
00:19:37,845 --> 00:19:40,205
would be that Spark only needs

441
00:19:40,205 --> 00:19:43,105
one honest actor to make sure.

442
00:19:43,265 --> 00:19:46,065
And so then we just have to trust ourselves, right?

443
00:19:47,105 --> 00:19:49,125
And that's easier than trusting

444
00:19:49,125 --> 00:19:50,585
a bunch of other companies that you don't have.

445
00:19:50,585 --> 00:19:54,225
You're not able to observe their internet.

446
00:19:54,385 --> 00:19:55,705
So that would be a reason.

447
00:19:56,025 --> 00:19:57,385
Also, potentially business model,

448
00:19:57,485 --> 00:19:58,825
they can cut fees and stuff like that.

449
00:19:58,905 --> 00:20:00,745
So there definitely can be incentives,

450
00:20:00,905 --> 00:20:02,345
but it'll be interesting to see

451
00:20:02,345 --> 00:20:03,885
if that's tested around liability.

452
00:20:04,265 --> 00:20:05,625
Now, where do things wind up?

453
00:20:07,345 --> 00:20:09,065
Blockstream just...

454
00:20:09,065 --> 00:20:13,925
So they paused the Liquid system last weekend.

455
00:20:14,065 --> 00:20:15,465
They just started it again today.

456
00:20:15,465 --> 00:20:17,205
I think last I checked,

457
00:20:17,385 --> 00:20:18,865
withdrawals are still not enabled,

458
00:20:18,865 --> 00:20:21,365
but the actual blockchain is producing blocks again.

459
00:20:22,465 --> 00:20:25,605
So this liquid is a block-based system.

460
00:20:26,685 --> 00:20:29,605
Not block ink, but blocks of transactions.

461
00:20:30,945 --> 00:20:31,845
A chain of blocks?

462
00:20:32,245 --> 00:20:33,205
A chain of blocks.

463
00:20:34,605 --> 00:20:37,365
And they restarted that

464
00:20:37,365 --> 00:20:41,445
and Atomback said that it will be one for one.

465
00:20:41,825 --> 00:20:42,145
Interesting.

466
00:20:42,325 --> 00:20:43,245
So they're plugging the hole then.

467
00:20:43,885 --> 00:20:44,865
So somehow...

468
00:20:44,865 --> 00:20:46,445
It'll be one for one meaning

469
00:20:46,445 --> 00:20:48,225
anybody's going to be made whole.

470
00:20:48,225 --> 00:20:56,345
Yeah, because with liquid, if it's a 50% haircut, the way liquid works is it's not evenly applied to everyone.

471
00:20:56,705 --> 00:20:57,685
It's a rush for the exits.

472
00:20:58,825 --> 00:21:06,105
Like, let's say we're all in liquid, and the moment they turn that on, it would be a fight for who gets their money out first.

473
00:21:06,165 --> 00:21:08,765
Because the last 15% gets zero.

474
00:21:09,425 --> 00:21:10,305
It's like the banking system.

475
00:21:10,425 --> 00:21:16,105
So it would be unfair, and it would probably be the death of liquid.

476
00:21:16,105 --> 00:21:18,925
I mean, is Liquid going to survive anyway?

477
00:21:19,045 --> 00:21:20,645
I don't know, but that would definitely be the death.

478
00:21:21,465 --> 00:21:24,345
But no, he announced that it will be one for one.

479
00:21:24,645 --> 00:21:29,705
So somehow, some way, 600 BTC are being plugged,

480
00:21:30,145 --> 00:21:34,025
apparently independent of whether they get it back from this attack.

481
00:21:35,005 --> 00:21:36,165
Yeah, very interesting story.

482
00:21:36,265 --> 00:21:37,065
That's the Liquid story.

483
00:21:37,765 --> 00:21:39,205
Is it kind of settled and done now,

484
00:21:39,285 --> 00:21:41,285
or are we going to keep following this and there's more stuff?

485
00:21:41,585 --> 00:21:44,665
Well, mainly, I mean, even through today,

486
00:21:44,665 --> 00:21:46,665
there's still op return messages

487
00:21:46,665 --> 00:21:47,605
going back and forth.

488
00:21:48,065 --> 00:21:48,985
Oh, they are with the hacker.

489
00:21:49,225 --> 00:21:51,665
Yeah, so are those 600 Bitcoin returned?

490
00:21:52,365 --> 00:21:53,665
Is this person apprehended?

491
00:21:54,685 --> 00:21:56,305
I mean, that's a future thing.

492
00:21:56,425 --> 00:21:59,425
And then who made Blockstream whole?

493
00:21:59,525 --> 00:22:01,385
Is it from Blockstream Treasury

494
00:22:01,385 --> 00:22:04,085
or individuals or investors?

495
00:22:04,465 --> 00:22:06,545
Like Tether's been a historical backer,

496
00:22:06,725 --> 00:22:10,125
so they'd have the funds to do it,

497
00:22:10,225 --> 00:22:12,825
but what's their incentive if they did?

498
00:22:12,825 --> 00:22:13,225
Yeah.

499
00:22:13,225 --> 00:22:13,345
Yeah.

500
00:22:14,145 --> 00:22:14,245
Right.

501
00:22:14,345 --> 00:22:14,625
Anyhow,

502
00:22:15,425 --> 00:22:16,285
and like with,

503
00:22:16,425 --> 00:22:17,165
you know,

504
00:22:17,205 --> 00:22:20,725
BinFinex had a big hack long ago,

505
00:22:20,965 --> 00:22:23,105
like 2015 or 2016.

506
00:22:23,385 --> 00:22:23,545
Yeah.

507
00:22:24,105 --> 00:22:27,665
In which way more than 4,000 Bitcoin were stolen.

508
00:22:27,665 --> 00:22:28,565
I think it was like,

509
00:22:29,125 --> 00:22:30,265
it was a lot.

510
00:22:30,505 --> 00:22:33,865
I think it was like 100,000 or 180,000 Bitcoin or something.

511
00:22:33,865 --> 00:22:34,405
Huge amount, yeah.

512
00:22:34,425 --> 00:22:35,905
Because Bitcoin was not,

513
00:22:36,105 --> 00:22:37,605
you know, it was like $200 of Bitcoin then.

514
00:22:37,865 --> 00:22:38,005
Okay.

515
00:22:39,705 --> 00:22:41,905
But they got themselves out of that

516
00:22:41,905 --> 00:22:42,925
by issuing a token

517
00:22:42,925 --> 00:22:44,665
I actually think it was one of the more

518
00:22:44,665 --> 00:22:45,225
I actually

519
00:22:45,225 --> 00:22:47,965
it's one of the few tokens

520
00:22:47,965 --> 00:22:49,285
It's called the USD

521
00:22:49,285 --> 00:22:50,245
and

522
00:22:50,245 --> 00:22:54,625
this token actually

523
00:22:54,625 --> 00:22:56,465
is like a reasonable

524
00:22:56,465 --> 00:22:57,805
use case in my opinion

525
00:22:57,805 --> 00:23:00,485
and that it's almost like

526
00:23:00,485 --> 00:23:02,625
a futures contract. I mean, you know, it's kind of like

527
00:23:02,625 --> 00:23:04,405
you think they're going to

528
00:23:04,405 --> 00:23:06,165
eventually make it whole and I don't

529
00:23:06,165 --> 00:23:08,385
so maybe I sell you my

530
00:23:08,385 --> 00:23:10,285
and then I was a depositor

531
00:23:10,285 --> 00:23:11,865
so they gave me some tokens

532
00:23:11,865 --> 00:23:14,145
for the haircut I took.

533
00:23:14,285 --> 00:23:16,245
And I'm like, I don't believe in this feature,

534
00:23:16,365 --> 00:23:17,305
so I sell it to you.

535
00:23:17,305 --> 00:23:20,565
And it seems like, and it just gets around.

536
00:23:21,365 --> 00:23:23,865
You can probably do that through some bankruptcy court

537
00:23:23,865 --> 00:23:26,145
or regulatory court, or it'd take 10 years.

538
00:23:26,285 --> 00:23:27,385
Right, like an on-chain version.

539
00:23:27,625 --> 00:23:28,725
This they could just do right away.

540
00:23:28,945 --> 00:23:29,425
So I don't know.

541
00:23:29,545 --> 00:23:31,005
This is actually an interesting token.

542
00:23:31,365 --> 00:23:31,945
But anyway.

543
00:23:32,765 --> 00:23:34,085
And did they get any,

544
00:23:34,145 --> 00:23:37,045
is there any kind of fingerprint on the hack

545
00:23:37,045 --> 00:23:40,545
that suggests it happened because of a brand new AI model,

546
00:23:40,545 --> 00:23:43,025
or it was just kind of sitting out there hidden in plain sight

547
00:23:43,025 --> 00:23:44,105
and anybody could have done this?

548
00:23:46,145 --> 00:23:51,045
Well, that fix that ended up having a worse vulnerability

549
00:23:51,045 --> 00:23:56,505
than what it fixed was made open source September 1st.

550
00:23:57,385 --> 00:24:01,625
And then September 7th or 8th is when the hack occurred, I think.

551
00:24:01,785 --> 00:24:03,025
So within a week.

552
00:24:03,685 --> 00:24:05,225
So those are the facts.

553
00:24:05,785 --> 00:24:08,405
So I don't think there was a new model in that week.

554
00:24:09,205 --> 00:24:10,265
I don't think that would explain it.

555
00:24:10,265 --> 00:24:12,245
but I've got to believe

556
00:24:12,245 --> 00:24:13,545
probably just people taking

557
00:24:13,545 --> 00:24:15,065
some of the new models

558
00:24:15,065 --> 00:24:16,085
and looking around

559
00:24:16,085 --> 00:24:16,265
yeah

560
00:24:16,265 --> 00:24:17,105
there's

561
00:24:17,105 --> 00:24:18,365
yeah if there's

562
00:24:18,365 --> 00:24:20,385
4,000 Bitcoin bounties

563
00:24:20,385 --> 00:24:21,245
I'm going to be looking at

564
00:24:21,245 --> 00:24:22,925
and I want to steal it

565
00:24:22,925 --> 00:24:24,425
I'm going to be looking at

566
00:24:24,425 --> 00:24:25,765
every new patch

567
00:24:25,765 --> 00:24:26,825
from Liquid

568
00:24:26,825 --> 00:24:28,765
and toss it into my

569
00:24:28,765 --> 00:24:30,025
security scanner

570
00:24:30,025 --> 00:24:30,565
and then be like

571
00:24:30,565 --> 00:24:32,065
okay I'll take 4,000 Bitcoin

572
00:24:32,065 --> 00:24:32,565
right

573
00:24:32,565 --> 00:24:34,845
I mean do you guys have any opinion

574
00:24:34,845 --> 00:24:36,845
on the white slash gray hat thing here

575
00:24:36,845 --> 00:24:37,545
I mean do you think

576
00:24:37,545 --> 00:24:38,725
I mean that's kind of unprecedented

577
00:24:38,725 --> 00:24:39,765
like do you have any strong

578
00:24:39,765 --> 00:24:40,965
strong feelings on that.

579
00:24:41,585 --> 00:24:43,765
Yeah, I think this is egregious

580
00:24:43,765 --> 00:24:46,605
and criminal and unethical

581
00:24:46,605 --> 00:24:49,225
and should be condemned, not encouraged.

582
00:24:49,465 --> 00:24:59,431
Yeah I think if you saying you I think you can claim white hat and then decide how much reward you get to keep Yeah Like why don I say I a white hat and I keeping 99

583
00:24:59,431 --> 00:24:59,851
Exactly.

584
00:25:00,451 --> 00:25:02,891
It's just, you're going to go rob a jewelry store.

585
00:25:03,531 --> 00:25:04,551
I totally agree.

586
00:25:04,691 --> 00:25:09,971
Just to feel, man, I mean, if someone took the other side, they would be like, well, it just needs to be reasonable, right?

587
00:25:10,011 --> 00:25:13,751
So, like, if they took 1%, then it'd be, oh, that was a reasonable bounty.

588
00:25:14,651 --> 00:25:17,371
So then it gets under a slippery slope.

589
00:25:17,651 --> 00:25:18,531
Reasonable according to who?

590
00:25:18,531 --> 00:25:23,511
So the amount or percentage is up for debate.

591
00:25:24,111 --> 00:25:26,191
But then also, who's dictating the terms?

592
00:25:26,351 --> 00:25:30,051
And I think historical norms for white hat hacking is liquid.

593
00:25:30,951 --> 00:25:31,091
Yeah.

594
00:25:31,811 --> 00:25:37,691
You as the white hat don't dictate the terms because that's extortion.

595
00:25:37,951 --> 00:25:38,191
Yeah.

596
00:25:38,671 --> 00:25:39,711
It's not white hat anymore.

597
00:25:40,291 --> 00:25:42,171
I mean, it's literally criminal extortion.

598
00:25:42,171 --> 00:25:42,831
It's a black hat.

599
00:25:42,831 --> 00:25:43,011
Yeah.

600
00:25:43,011 --> 00:25:49,991
So, yeah, to me, it's black and white, not gray on that.

601
00:25:51,211 --> 00:25:55,571
But it's actually really odd to me that Bitcoiners

602
00:25:55,571 --> 00:25:58,831
who are always about taxation is theft,

603
00:25:58,931 --> 00:26:01,831
it's like, oh, but it's only fine to steal 4,000 Bitcoin

604
00:26:01,831 --> 00:26:03,471
and keep 15% of it.

605
00:26:03,831 --> 00:26:04,671
That's not theft.

606
00:26:04,711 --> 00:26:05,271
It's a tax.

607
00:26:07,271 --> 00:26:09,351
Anyhow, all right, let's talk about...

608
00:26:09,351 --> 00:26:10,711
Yeah, let's talk about some open source AI,

609
00:26:10,711 --> 00:26:11,511
because that was today.

610
00:26:11,511 --> 00:26:13,971
So, where do we begin?

611
00:26:14,431 --> 00:26:15,411
Yeah, well, we're doing,

612
00:26:15,571 --> 00:26:17,891
there's a two-day event in the Presidio

613
00:26:17,891 --> 00:26:19,811
hosted by Presidio Bitcoin,

614
00:26:20,071 --> 00:26:21,571
the Open Source AI Summit.

615
00:26:21,791 --> 00:26:21,931
Yep.

616
00:26:23,811 --> 00:26:25,571
First of all, let me give a shout out

617
00:26:25,571 --> 00:26:27,011
to the people who have made this happen.

618
00:26:27,111 --> 00:26:27,351
Oh, yeah.

619
00:26:27,691 --> 00:26:28,931
Haley on my team.

620
00:26:29,291 --> 00:26:30,931
Austin, who's Spiral Grantee.

621
00:26:31,611 --> 00:26:33,271
And then Alex who leads PB.

622
00:26:33,631 --> 00:26:35,871
And the three of them work here out of PB.

623
00:26:36,091 --> 00:26:36,871
They've done a lot of work.

624
00:26:37,431 --> 00:26:38,251
They've done a lot of work.

625
00:26:38,251 --> 00:26:40,591
We've put on a lot of great summits,

626
00:26:40,591 --> 00:26:42,451
including your type 1 energy summit.

627
00:26:43,151 --> 00:26:45,431
But this one is like next level.

628
00:26:45,571 --> 00:26:45,971
It's bigger.

629
00:26:46,271 --> 00:26:47,531
We're not even hosting it here at PB

630
00:26:47,531 --> 00:26:51,951
because we had over 130 people today.

631
00:26:52,491 --> 00:26:53,211
And I think after tomorrow,

632
00:26:53,331 --> 00:26:57,191
we'll have like around 150 unique people

633
00:26:57,191 --> 00:26:58,571
who are coming to this summit.

634
00:27:00,331 --> 00:27:03,751
And yeah, today was out of the Golden Gate Club.

635
00:27:04,131 --> 00:27:05,731
Beautiful venue, really nice.

636
00:27:05,731 --> 00:27:06,491
Amazing, yeah.

637
00:27:06,751 --> 00:27:10,231
And we were able to get like an all-star

638
00:27:10,231 --> 00:27:12,631
set of folks as well.

639
00:27:12,751 --> 00:27:13,971
I mean, we had talked about on the show

640
00:27:13,971 --> 00:27:16,411
maybe a month ago that David Sachs

641
00:27:16,411 --> 00:27:18,491
agreed to speak, so

642
00:27:18,491 --> 00:27:20,731
I was able to have a great conversation with him

643
00:27:20,731 --> 00:27:22,611
and we can dive into that more if you'd like.

644
00:27:24,191 --> 00:27:24,711
But not,

645
00:27:24,891 --> 00:27:26,631
I mean, obviously it's a big name, but we've

646
00:27:26,631 --> 00:27:27,791
had many others as well.

647
00:27:28,911 --> 00:27:30,691
Including tomorrow,

648
00:27:31,391 --> 00:27:32,711
I'll let you guys talk a little bit

649
00:27:32,711 --> 00:27:34,791
about the speakers today and what you found interesting,

650
00:27:34,911 --> 00:27:36,651
but tomorrow we have a VP

651
00:27:36,651 --> 00:27:38,391
from Hugging Face opening up tomorrow

652
00:27:38,391 --> 00:27:41,271
and then a VP, the guy in charge of NemoTron,

653
00:27:41,431 --> 00:27:43,611
open source model for NVIDIA, closes tomorrow.

654
00:27:43,911 --> 00:27:44,411
That's a pretty big deal.

655
00:27:44,811 --> 00:27:47,291
It's a really good cast of folks.

656
00:27:47,391 --> 00:27:48,611
Yeah, and I was super impressed to see,

657
00:27:48,711 --> 00:27:50,551
I mean, just really in the last couple of weeks,

658
00:27:50,951 --> 00:27:53,711
yeah, I mean, it came together and it started being,

659
00:27:53,911 --> 00:27:55,571
obviously with the Good Summit,

660
00:27:55,691 --> 00:27:57,351
hopefully this is what happens is it starts off,

661
00:27:57,451 --> 00:27:59,091
it's like a little bit slow to get going,

662
00:27:59,151 --> 00:28:00,811
then as you get going, everyone wants in,

663
00:28:00,851 --> 00:28:03,151
then all of a sudden, sorry, the scare seats are taken.

664
00:28:03,151 --> 00:28:07,231
But it was really cool just to see how much demand there was.

665
00:28:07,231 --> 00:28:09,291
it was really cool to see that this summit was a mix.

666
00:28:09,491 --> 00:28:11,131
You know, there's people from the, you know,

667
00:28:11,151 --> 00:28:13,531
sort of Bitcoin world that are mixing with like the actual,

668
00:28:13,751 --> 00:28:15,351
you know, people that are building the open source AI,

669
00:28:15,531 --> 00:28:16,411
you know, the, you know,

670
00:28:16,471 --> 00:28:18,591
sort of Vanguard here in Silicon Valley.

671
00:28:18,731 --> 00:28:20,511
And this is the first event that I've seen

672
00:28:20,511 --> 00:28:22,031
where like that was like really mixed.

673
00:28:22,151 --> 00:28:23,091
And I hope crossover.

674
00:28:23,431 --> 00:28:24,651
I hope really good stuff will come out.

675
00:28:24,651 --> 00:28:26,911
I think the number, like two thirds of the attendees

676
00:28:26,911 --> 00:28:29,311
are like not Bitcoin.

677
00:28:29,611 --> 00:28:30,191
They're AI people.

678
00:28:30,371 --> 00:28:30,571
Even better.

679
00:28:30,751 --> 00:28:32,331
And a third are Bitcoin.

680
00:28:32,671 --> 00:28:33,991
But even the third that are Bitcoin

681
00:28:33,991 --> 00:28:36,011
are people who are very much in AI,

682
00:28:36,011 --> 00:28:38,071
using AI or very AI-filled.

683
00:28:38,331 --> 00:28:38,971
Yeah, yeah.

684
00:28:38,971 --> 00:28:39,751
So that was cool.

685
00:28:40,151 --> 00:28:43,191
Who are some of your favorite speakers

686
00:28:43,191 --> 00:28:44,491
or topics or panels?

687
00:28:44,631 --> 00:28:45,251
Or people you met.

688
00:28:45,751 --> 00:28:48,411
I want to make sure I get his name right.

689
00:28:48,471 --> 00:28:49,611
Is it Ramez Nam?

690
00:28:49,851 --> 00:28:50,851
Yeah, Ramez.

691
00:28:51,071 --> 00:28:51,591
Ramez, okay.

692
00:28:52,031 --> 00:28:52,751
I liked his,

693
00:28:52,851 --> 00:28:54,351
I wasn't familiar with his work,

694
00:28:54,411 --> 00:28:56,191
but I guess he's a sci-fi author.

695
00:28:56,351 --> 00:28:57,911
So he's got three sci-fi books.

696
00:28:58,771 --> 00:29:01,691
And I liked the way he kind of showed

697
00:29:01,691 --> 00:29:04,211
the transition of what's happening

698
00:29:04,211 --> 00:29:09,391
with kind of, you know, I think a lot of time, you know, in 2022, everybody said, oh, open AI,

699
00:29:09,911 --> 00:29:15,191
fully central, they're going to just run the table with this. And he sort of showed how I think in

700
00:29:15,191 --> 00:29:23,431
2022 or so they had like a 20 month lead. And then in I think 2023 or four, they that shrunk to five

701
00:29:23,431 --> 00:29:28,871
months lead. And now it's basically, you know, every few weeks, there can be like a new leader.

702
00:29:28,871 --> 00:29:33,271
I think that's like a really promising idea to have illustrated that way.

703
00:29:34,091 --> 00:29:39,091
And especially to do it in, you know, instead of AI being this huge centralizing force,

704
00:29:39,151 --> 00:29:40,611
it could be a plurality of AIs.

705
00:29:41,491 --> 00:29:45,411
And so I thought it was actually interesting to see him bring that data to the story.

706
00:29:46,071 --> 00:29:48,351
And I mean, I'm optimistic.

707
00:29:48,611 --> 00:29:52,571
I think I've been optimistic and continue to be optimistic that there's going to be intelligence everywhere.

708
00:29:52,691 --> 00:29:55,331
It's not going to be a single monopolistic winner.

709
00:29:55,331 --> 00:30:07,911
I don't think, you know, even if there's recursive self-improvement, I think there will be, I know Max may disagree, but I think there's going to be a lot of AI and it's going to be a benefit of humanity to have so many different alternative options.

710
00:30:07,911 --> 00:30:10,831
And he created a really compelling case that that is going to be true, right?

711
00:30:10,991 --> 00:30:12,871
So, yeah, I thought, I totally agree with that.

712
00:30:13,071 --> 00:30:14,371
I thought he brought great energy.

713
00:30:14,451 --> 00:30:15,311
He opened the event, right?

714
00:30:15,331 --> 00:30:16,231
So he brought great energy.

715
00:30:16,691 --> 00:30:18,071
He also opened Type 1, right?

716
00:30:18,091 --> 00:30:18,651
He also opened Type 1.

717
00:30:18,671 --> 00:30:19,211
Oh, yeah.

718
00:30:19,331 --> 00:30:20,991
And I got to give a shout out to him.

719
00:30:21,231 --> 00:30:24,551
Mez is probably, I mean, certainly one of my favorite living futurists.

720
00:30:24,551 --> 00:30:27,891
he was the guy that actually solar pilled me back in like 2015.

721
00:30:28,251 --> 00:30:30,591
He was the first one to really show the Moore's Law of Solar,

722
00:30:30,931 --> 00:30:33,251
I think in like 2011, an article for Scientific American.

723
00:30:33,451 --> 00:30:35,571
So he's really always ahead of the curve.

724
00:30:35,811 --> 00:30:37,771
And I mean, yeah, it's a very hopeful message that,

725
00:30:38,031 --> 00:30:39,811
I mean, I want to believe I'm getting there,

726
00:30:39,971 --> 00:30:42,531
but regardless, it's a hopeful message.

727
00:30:42,531 --> 00:30:45,431
Yeah, we're lucky to,

728
00:30:45,711 --> 00:30:47,611
we're fortunate to have him come to our events

729
00:30:47,611 --> 00:30:48,771
and help out that way.

730
00:30:49,791 --> 00:30:51,591
And just for those listening,

731
00:30:51,591 --> 00:30:53,911
and you're maybe curious

732
00:30:53,911 --> 00:30:54,971
and want to watch the summit,

733
00:30:55,171 --> 00:30:56,131
it was recorded.

734
00:30:56,291 --> 00:30:57,371
It wasn't live streamed though.

735
00:30:57,631 --> 00:30:59,271
So we're,

736
00:30:59,431 --> 00:31:01,351
I think it'll probably be next week

737
00:31:01,351 --> 00:31:05,291
when all the recordings are made available,

738
00:31:05,431 --> 00:31:06,651
but we will be publishing it,

739
00:31:06,831 --> 00:31:07,751
including remesses.

740
00:31:08,051 --> 00:31:08,231
Sweet.

741
00:31:08,231 --> 00:31:12,331
We're trying to rush the David Sachs one

742
00:31:12,331 --> 00:31:15,331
to get it out by tomorrow, midday.

743
00:31:16,131 --> 00:31:17,831
Because, I mean,

744
00:31:17,871 --> 00:31:19,831
we're excited to have him be part of the summit.

745
00:31:19,831 --> 00:31:24,351
and then what happened yesterday, right?

746
00:31:24,911 --> 00:31:25,751
I was offline.

747
00:31:25,911 --> 00:31:26,331
I missed this.

748
00:31:27,851 --> 00:31:28,891
Give me the setup.

749
00:31:29,011 --> 00:31:29,351
What happened?

750
00:31:29,691 --> 00:31:33,431
Well, so there is this employee at Anthropic

751
00:31:33,431 --> 00:31:34,431
who just quit.

752
00:31:34,611 --> 00:31:36,611
So ex-employee just quit Anthropic,

753
00:31:37,091 --> 00:31:39,451
tweeted out talking about

754
00:31:39,451 --> 00:31:42,151
he's worried about the end of humanity

755
00:31:42,151 --> 00:31:44,051
greater than 10% chance

756
00:31:44,051 --> 00:31:44,871
that we're all going to die

757
00:31:44,871 --> 00:31:45,811
in the next couple of years.

758
00:31:46,411 --> 00:31:47,331
Very dramatic.

759
00:31:47,331 --> 00:31:48,411
Standard Anthropic fan.

760
00:31:48,411 --> 00:31:52,731
very dramatic and apocalyptic message,

761
00:31:52,891 --> 00:31:54,711
but with no substance or it's not like,

762
00:31:55,411 --> 00:31:56,591
I mean, as David...

763
00:31:56,591 --> 00:31:57,791
He didn't reveal any new information.

764
00:31:57,951 --> 00:31:59,031
Yeah, I mean, as David...

765
00:31:59,031 --> 00:32:00,011
Zach sort of joked,

766
00:32:00,011 --> 00:32:02,051
I mean, usually if you're a whistleblower,

767
00:32:02,211 --> 00:32:03,451
I mean, he's being framed as a whistleblower.

768
00:32:03,591 --> 00:32:05,931
Number one, usually the company that you left

769
00:32:05,931 --> 00:32:08,271
is not on your side and supporting you.

770
00:32:08,371 --> 00:32:10,371
And in this case, many Anthropoc employees

771
00:32:10,371 --> 00:32:12,471
were retweeting it and supporting

772
00:32:12,471 --> 00:32:14,051
this apocalyptic message.

773
00:32:15,191 --> 00:32:18,311
Also, this person only worked there six weeks.

774
00:32:18,411 --> 00:32:19,851
only worked at Anthropic

775
00:32:19,851 --> 00:32:20,411
for six weeks

776
00:32:20,411 --> 00:32:20,911
for six weeks

777
00:32:20,911 --> 00:32:21,771
and he had like

778
00:32:21,771 --> 00:32:22,631
no Twitter history

779
00:32:22,631 --> 00:32:23,651
and I don't know

780
00:32:23,651 --> 00:32:24,271
I don't know

781
00:32:24,271 --> 00:32:25,051
I don't know how old he is

782
00:32:25,051 --> 00:32:25,331
but I mean

783
00:32:25,331 --> 00:32:26,331
he looks quite young

784
00:32:26,331 --> 00:32:26,691
I mean

785
00:32:26,691 --> 00:32:27,871
I think he's early

786
00:32:27,871 --> 00:32:29,091
early in his career

787
00:32:29,091 --> 00:32:30,031
Sachs called it a style

788
00:32:30,031 --> 00:32:30,371
and then

789
00:32:30,371 --> 00:32:31,311
and then like

790
00:32:31,311 --> 00:32:32,691
and then in the same

791
00:32:32,691 --> 00:32:33,951
or I think he tweeted

792
00:32:33,951 --> 00:32:34,751
maybe Tuesday night

793
00:32:34,751 --> 00:32:35,011
or something

794
00:32:35,011 --> 00:32:35,551
but then like

795
00:32:35,551 --> 00:32:36,151
all day Wednesday

796
00:32:36,151 --> 00:32:37,011
he's on like

797
00:32:37,011 --> 00:32:37,891
all these

798
00:32:37,891 --> 00:32:39,031
mainstream media

799
00:32:39,031 --> 00:32:40,391
interviews

800
00:32:40,391 --> 00:32:41,111
they say he was like

801
00:32:41,111 --> 00:32:42,371
on national news

802
00:32:42,371 --> 00:32:43,471
and is he on Rogan

803
00:32:43,471 --> 00:32:44,871
well Rogan

804
00:32:44,871 --> 00:32:45,711
Rogan covered

805
00:32:45,711 --> 00:32:47,031
so the Wall Street Journal

806
00:32:47,031 --> 00:32:48,151
or either the New York Times

807
00:32:48,151 --> 00:32:49,971
Wall Street Journal also published an article

808
00:32:49,971 --> 00:32:51,751
like 18 minutes before the tweet

809
00:32:51,751 --> 00:32:54,331
that was like about or the tweet.

810
00:32:55,131 --> 00:32:56,051
And then it was Joe Rogan.

811
00:32:56,111 --> 00:32:57,531
It was like clearly orchestrated,

812
00:32:57,691 --> 00:32:59,971
like dozens of politicians retweeted this

813
00:32:59,971 --> 00:33:01,791
with their message to their constituents, etc.

814
00:33:01,931 --> 00:33:04,171
So clearly, I mean, obviously the stuff

815
00:33:04,171 --> 00:33:08,211
wasn't just spontaneous, I think.

816
00:33:08,291 --> 00:33:10,731
But in any case, it created a firestorm.

817
00:33:10,991 --> 00:33:13,451
And I mean, the concern here,

818
00:33:13,451 --> 00:33:15,911
at least from if you're a supporter

819
00:33:15,911 --> 00:33:17,791
of open-weight models,

820
00:33:18,151 --> 00:33:20,891
and distributed versus centralized,

821
00:33:22,031 --> 00:33:25,471
then the concern here is that this is going to create

822
00:33:25,471 --> 00:33:29,251
rapid momentum where Congress acts

823
00:33:29,251 --> 00:33:32,251
and introduces legislation and possibly regulatory frameworks

824
00:33:32,251 --> 00:33:33,771
that are very draconian,

825
00:33:34,371 --> 00:33:38,131
where companies like Anthropoc and OpenAI have the,

826
00:33:38,711 --> 00:33:41,031
it just creates, it's regulatory capture

827
00:33:41,031 --> 00:33:42,071
and creates a moat for them,

828
00:33:42,571 --> 00:33:44,171
and everyone else in open models

829
00:33:44,171 --> 00:33:45,611
might not even be able to comply.

830
00:33:45,611 --> 00:33:47,631
And specifically, I mean, I think it's worth

831
00:33:47,631 --> 00:33:51,691
stating it kind of, I don't know the exact words he used, but if

832
00:33:51,691 --> 00:33:55,991
regulatory capture, if it basically can create a panic, then

833
00:33:55,991 --> 00:34:00,131
the regulators say, okay, we have to deal with the panic. The way to deal

834
00:34:00,131 --> 00:34:04,231
with the panic is to talk to somebody, the leaders of OpenAI, leaders of

835
00:34:04,231 --> 00:34:08,391
Claude, you know, Anthropic. And then you build

836
00:34:08,391 --> 00:34:11,511
legislation and regulation that

837
00:34:11,511 --> 00:34:16,291
supports their needs, but then any of the open source stuff is sort of out in the

838
00:34:16,291 --> 00:34:20,151
cold. There's nobody to talk to. They're not involved in the conversation. And so I think

839
00:34:20,151 --> 00:34:25,931
part of the story he was saying is, let's figure out how to, as a kind of open source,

840
00:34:26,171 --> 00:34:31,911
you know, plurality of models, how do those things sort of have a voice at the table and

841
00:34:31,911 --> 00:34:35,271
sort of avoid the regulatory capture from being able to be completed, right?

842
00:34:35,411 --> 00:34:37,971
Yeah, and I think that's an important point that it doesn't necessarily have to be as

843
00:34:37,971 --> 00:34:42,011
draconian as open source is banned. But like, you know, for example, if you have like, you

844
00:34:42,011 --> 00:34:45,331
know, a FinTech regulation that says we have to speak to your CEO, it's like, well, Bitcoin

845
00:34:45,331 --> 00:34:46,011
doesn't have a CEO.

846
00:34:46,550 --> 00:34:47,171
So I mean,

847
00:34:47,731 --> 00:34:49,191
I'd like to believe that

848
00:34:49,191 --> 00:34:50,491
people are going to

849
00:34:50,491 --> 00:34:51,731
see through some of that,

850
00:34:51,911 --> 00:34:53,291
but I don't know.

851
00:34:53,491 --> 00:34:53,631
I mean,

852
00:34:53,751 --> 00:34:54,571
people are really,

853
00:34:55,451 --> 00:34:55,911
the PSYOP,

854
00:34:56,071 --> 00:34:57,411
the orchestration of the media,

855
00:34:57,611 --> 00:34:59,411
you can create a lot of confusion.

856
00:34:59,411 --> 00:34:59,631
Also,

857
00:34:59,891 --> 00:35:00,951
and you already have,

858
00:35:02,171 --> 00:35:02,451
you know,

859
00:35:02,531 --> 00:35:03,031
it's common,

860
00:35:03,711 --> 00:35:04,231
it's,

861
00:35:05,151 --> 00:35:06,471
the AI sentiment

862
00:35:06,471 --> 00:35:07,371
is quite negative

863
00:35:07,371 --> 00:35:08,151
already

864
00:35:08,151 --> 00:35:09,731
among a lot of mainstream folks.

865
00:35:09,731 --> 00:35:10,311
Right, right.

866
00:35:10,371 --> 00:35:11,851
So you have that

867
00:35:11,851 --> 00:35:13,431
and then these extreme messages

868
00:35:13,431 --> 00:35:17,711
just confirms their existing priors.

869
00:35:18,391 --> 00:35:20,631
So if something confirms your priors,

870
00:35:20,811 --> 00:35:22,171
you're less inquisitive.

871
00:35:22,731 --> 00:35:24,550
You're less like, where's the data?

872
00:35:24,771 --> 00:35:25,751
Where's the evidence?

873
00:35:26,251 --> 00:35:27,491
That's what I thought was going to happen.

874
00:35:27,491 --> 00:35:29,351
Yeah, it's like, yep, that's what I thought.

875
00:35:29,851 --> 00:35:30,871
AI is bad for us.

876
00:35:31,111 --> 00:35:33,851
So it is kind of set up for a disaster

877
00:35:33,851 --> 00:35:35,011
where then, of course,

878
00:35:35,011 --> 00:35:37,071
I mean, why are the politicians

879
00:35:37,071 --> 00:35:38,251
climbing onto this?

880
00:35:38,591 --> 00:35:40,050
I actually don't think they really care

881
00:35:40,050 --> 00:35:41,030
about AI safety

882
00:35:41,030 --> 00:35:42,891
or think that we're all going to die in two years.

883
00:35:42,891 --> 00:35:45,231
they want votes and they want donations.

884
00:35:45,731 --> 00:35:46,991
Well, I mean, to be fair,

885
00:35:47,050 --> 00:35:48,491
just to play devil's advocate a little bit on this,

886
00:35:48,550 --> 00:35:50,651
I mean, I think it's probably true for a lot of them.

887
00:35:51,011 --> 00:35:54,351
I do actually believe a lot of people at Anthropic

888
00:35:54,351 --> 00:35:55,451
really do believe this.

889
00:35:55,531 --> 00:35:57,331
And I think increasingly, if you keep saying that,

890
00:35:57,391 --> 00:35:58,591
some of these people and they're telling their story,

891
00:35:59,011 --> 00:36:00,550
I do think a lot of people probably are,

892
00:36:00,711 --> 00:36:03,191
even in politics, generally afraid of that message.

893
00:36:04,231 --> 00:36:05,791
So I think there's both sides to that.

894
00:36:05,791 --> 00:36:09,631
I agree with that, but I just think that's lunacy

895
00:36:09,631 --> 00:36:13,331
in the sense that no one's answering,

896
00:36:13,550 --> 00:36:14,771
what about the Chinese open?

897
00:36:15,131 --> 00:36:15,311
Yeah.

898
00:36:15,671 --> 00:36:17,011
And I asked David that.

899
00:36:17,151 --> 00:36:18,131
And that was a great question for me.

900
00:36:18,291 --> 00:36:19,731
In fact, I thought that was one of your best questions

901
00:36:19,731 --> 00:36:21,331
was like, okay, you know,

902
00:36:21,731 --> 00:36:22,911
if we try and ban open source

903
00:36:22,911 --> 00:36:24,371
and just cede it to the Chinese, then what?

904
00:36:24,451 --> 00:36:26,691
And like, I thought his response is also pretty funny.

905
00:36:26,691 --> 00:36:27,611
Well, I asked him to steal that

906
00:36:27,611 --> 00:36:30,111
because I know he doesn't hold that position.

907
00:36:30,291 --> 00:36:30,550
Yeah.

908
00:36:30,971 --> 00:36:32,331
So I'm like, steal man it.

909
00:36:32,451 --> 00:36:34,331
And, you know, he did try to steal man it,

910
00:36:34,431 --> 00:36:36,411
which was, you know, they would say that we need,

911
00:36:36,911 --> 00:36:37,631
he didn't say it this way,

912
00:36:37,631 --> 00:36:38,271
but like sort of a,

913
00:36:38,271 --> 00:36:41,591
nuclear or non-proliferation

914
00:36:41,591 --> 00:36:43,531
treaty with China

915
00:36:43,531 --> 00:36:44,051
or something.

916
00:36:44,511 --> 00:36:45,891
And I tend to agree

917
00:36:45,891 --> 00:36:46,551
with him on this point.

918
00:36:47,091 --> 00:36:47,551
Yeah, exactly.

919
00:36:47,751 --> 00:36:48,651
A global pause.

920
00:36:48,771 --> 00:36:49,451
I mean, he basically said

921
00:36:49,451 --> 00:36:50,011
these people are living

922
00:36:50,011 --> 00:36:51,091
in a fantasy land.

923
00:36:51,171 --> 00:36:52,111
I tend to agree that

924
00:36:52,111 --> 00:36:52,831
there's no chance

925
00:36:52,831 --> 00:36:53,631
that people are going

926
00:36:53,631 --> 00:36:54,151
to agree to pause.

927
00:36:54,171 --> 00:36:55,031
Because one data point

928
00:36:55,031 --> 00:36:55,771
that you point out was

929
00:36:55,771 --> 00:36:56,031
Remember when Elon

930
00:36:56,031 --> 00:36:56,591
was trying to do

931
00:36:56,591 --> 00:36:57,731
the global pause on it?

932
00:36:58,371 --> 00:36:58,631
Elon?

933
00:36:58,891 --> 00:36:59,231
Yeah.

934
00:36:59,371 --> 00:36:59,511
Yeah.

935
00:36:59,891 --> 00:37:00,111
When?

936
00:37:00,331 --> 00:37:01,711
Before XAI had a seat

937
00:37:01,711 --> 00:37:02,191
at the table.

938
00:37:02,371 --> 00:37:02,691
He was like,

939
00:37:02,751 --> 00:37:03,891
hey, we got to slow it all down.

940
00:37:04,031 --> 00:37:04,611
Let me catch up.

941
00:37:04,611 --> 00:37:04,771
Yeah.

942
00:37:04,891 --> 00:37:05,471
Sorry, man.

943
00:37:06,511 --> 00:37:08,231
That hurt your fan

944
00:37:08,231 --> 00:37:09,291
I'm a little bit there.

945
00:37:09,651 --> 00:37:12,151
Hey man, listen, I don't agree on everything.

946
00:37:12,391 --> 00:37:13,091
I don't agree on everything.

947
00:37:13,871 --> 00:37:28,476
But a stat that David mentioned was some survey in China like 83 of the population has positive sentiment on AI And I don know what the stats are in the US but it more of the opposite of that right

948
00:37:29,036 --> 00:37:30,616
So, if you

949
00:37:30,616 --> 00:37:32,636
have a population of

950
00:37:32,636 --> 00:37:34,696
over a billion people, 83%

951
00:37:34,696 --> 00:37:35,856
are pro-AI,

952
00:37:36,436 --> 00:37:38,676
and you're doing well with AI,

953
00:37:38,796 --> 00:37:40,516
you see the future of it, are you

954
00:37:40,516 --> 00:37:42,556
going to pause because the US

955
00:37:42,556 --> 00:37:43,856
has a chat with you?

956
00:37:44,816 --> 00:37:46,136
No. Come on.

957
00:37:46,136 --> 00:37:47,536
I think it really, it comes down.

958
00:37:47,616 --> 00:37:52,336
I think that really reinforced for me this idea that China, you know, at writ large, one,

959
00:37:52,616 --> 00:37:54,416
has seen, you know, a lot of people come out of poverty.

960
00:37:54,636 --> 00:37:55,976
They've seen technology be great for them.

961
00:37:56,036 --> 00:37:59,956
So they're like, okay, this is going to be more of the same, more welfare, like more

962
00:37:59,956 --> 00:38:01,296
economic progress, let's say.

963
00:38:02,116 --> 00:38:06,476
But I also think it's interesting and like, you know, I'd be curious to see more of like

964
00:38:06,476 --> 00:38:11,036
a debate between, you know, Chinese lab leaders and lab leaders in the US, because I really

965
00:38:11,036 --> 00:38:12,676
do think there's a fundamental breakdown.

966
00:38:12,676 --> 00:38:15,656
And you could be totally cynical and say, no, it's all just like lying, right?

967
00:38:15,656 --> 00:38:19,496
to capture. Maybe you could be totally there. They're, you know, really believe in it. Maybe

968
00:38:19,496 --> 00:38:23,316
it's a mix of both, but it really does seem like there's a fundamental divide of, you know, in

969
00:38:23,316 --> 00:38:27,616
China, at least from everything I've heard, they see it as a tool, as a new technology, like all

970
00:38:27,616 --> 00:38:32,376
these other technologies akin to something like the web or a train or, you know, airplanes,

971
00:38:32,476 --> 00:38:37,396
whatever. Whereas, you know, this group of sort of the EA people, um, and to be fair, I think a

972
00:38:37,396 --> 00:38:41,556
lot of them really, truly fundamentally believe it, uh, think that this could run away, be gray

973
00:38:41,556 --> 00:38:43,116
goo, paperclip maximization,

974
00:38:43,616 --> 00:38:45,416
biological weapons, whatever.

975
00:38:45,576 --> 00:38:47,536
Pick your flavor on there. So I'd be

976
00:38:47,536 --> 00:38:49,416
curious. I would love to see some of these leaders

977
00:38:49,416 --> 00:38:51,416
from China or whatever, a lot of people from China

978
00:38:51,416 --> 00:38:53,456
and the U.S. actually talk about it and see if

979
00:38:53,456 --> 00:38:55,296
they can move each other's minds at all or

980
00:38:55,296 --> 00:38:57,436
not. But that does seem to be a stark

981
00:38:57,436 --> 00:38:57,856
difference.

982
00:38:59,076 --> 00:39:01,356
We're going to do our best to get the David Sachs

983
00:39:01,356 --> 00:39:03,396
conversation out by

984
00:39:03,396 --> 00:39:05,416
midday tomorrow. Ashu, are you doing

985
00:39:05,416 --> 00:39:05,896
that work?

986
00:39:07,276 --> 00:39:09,516
I think you just got a sign while he's

987
00:39:09,516 --> 00:39:11,256
here producing. Shout out to

988
00:39:11,256 --> 00:39:13,176
Ashu. Ashu bot.

989
00:39:13,476 --> 00:39:14,296
Begin next day.

990
00:39:15,156 --> 00:39:16,556
No, he's...

991
00:39:16,556 --> 00:39:19,216
Yeah, you need like three clones or agents

992
00:39:19,216 --> 00:39:21,096
of yourself, Ashu, because he's been doing

993
00:39:21,096 --> 00:39:22,936
tons of work for the summit today. Ashu's the man.

994
00:39:23,516 --> 00:39:24,576
Then I'm like, hey, Ashu,

995
00:39:24,916 --> 00:39:26,776
can we get this

996
00:39:26,776 --> 00:39:29,556
David Sachs video edited and post-produced

997
00:39:29,556 --> 00:39:30,896
like ASAP? And then, oh,

998
00:39:31,256 --> 00:39:33,456
and then, Ashu, let's do an impromptu

999
00:39:33,456 --> 00:39:34,436
PBJ. Yeah.

1000
00:39:35,176 --> 00:39:37,016
I was just saying, finding Ashu,

1001
00:39:37,236 --> 00:39:39,016
like, best thing that ever came out of the sauna.

1002
00:39:39,156 --> 00:39:39,676
He's the man.

1003
00:39:39,676 --> 00:39:41,816
so thank you

1004
00:39:41,816 --> 00:39:43,336
let's see

1005
00:39:43,336 --> 00:39:47,276
what else about the summit either today

1006
00:39:47,276 --> 00:39:49,376
or is there anything I guess you're not going to be there

1007
00:39:49,376 --> 00:39:51,216
tomorrow but the two of us are I mean you had

1008
00:39:51,216 --> 00:39:52,076
two panels yourself

1009
00:39:52,076 --> 00:39:54,676
I had two panels

1010
00:39:54,676 --> 00:39:56,816
the last one was very interesting

1011
00:39:56,816 --> 00:39:59,196
I guess top of mind for me was on open

1012
00:39:59,196 --> 00:40:01,356
protocols and so that was super cool because

1013
00:40:01,356 --> 00:40:03,096
you know we talked about open protocols

1014
00:40:03,096 --> 00:40:04,876
for identity commerce communications

1015
00:40:04,876 --> 00:40:07,016
and memory and it was cool because I actually had

1016
00:40:07,016 --> 00:40:09,136
four folks on there and I got a kind of a

1017
00:40:09,136 --> 00:40:12,896
it was both a variety of perspectives, but they also had like approached some of these different

1018
00:40:12,896 --> 00:40:16,576
standards from different angles. Um, so we had Elon Strauss, who's more of a sort of like

1019
00:40:16,576 --> 00:40:19,576
philosopher guy that's written a lot with Tim O'Reilly. And it was good because he brought

1020
00:40:19,576 --> 00:40:24,776
that perspective of kind of like, um, you know, free markets, Hayek move, uh, you know, uh,

1021
00:40:24,776 --> 00:40:28,316
information process to the edge. And then we had three builders and it was cool because, um,

1022
00:40:28,956 --> 00:40:32,216
well, three of the four people had actually played with Nostra on identity. So obviously

1023
00:40:32,216 --> 00:40:36,876
I turned a bit of a Nostra commercial and the fourth was very open to it. Um, and so it was

1024
00:40:36,876 --> 00:40:41,796
really cool to hear Tom from Buzz talk about his experience building with Nostra, why they chose it.

1025
00:40:42,136 --> 00:40:46,856
Obviously, Roland from Albie and TaskUl had a lot of experience. Perhaps the most exciting thing to

1026
00:40:46,856 --> 00:40:51,636
me was that this guy, Sam Nightwing from Noose Research. Noose is, you know, they make Hermes,

1027
00:40:51,756 --> 00:40:55,816
which is, I think, the most popular agent right now, even more so than OpenClaw. They're very

1028
00:40:55,816 --> 00:41:01,076
sort of open source, open mind, you know, ethos. They haven't made a lot of these decisions yet.

1029
00:41:01,136 --> 00:41:05,056
So it's interesting for him to get to sort of, you know, learn and kind of ask some questions

1030
00:41:05,056 --> 00:41:06,176
before and after as well.

1031
00:41:06,616 --> 00:41:07,436
He seemed really interested.

1032
00:41:07,976 --> 00:41:10,836
You mentioned open source memory protocols.

1033
00:41:10,916 --> 00:41:11,196
Oh, yeah.

1034
00:41:11,676 --> 00:41:13,076
And I don't know much about it.

1035
00:41:13,156 --> 00:41:14,676
Yeah, I mean, I, to be frank, don't either.

1036
00:41:14,796 --> 00:41:16,756
That was something that Elon brought up to me

1037
00:41:16,756 --> 00:41:17,776
before the panel,

1038
00:41:17,956 --> 00:41:19,416
and he's been pushing for it.

1039
00:41:20,656 --> 00:41:21,456
Yeah, you know,

1040
00:41:21,496 --> 00:41:23,116
I guess they're trying to get some coalition around this.

1041
00:41:23,196 --> 00:41:24,756
I mean, I don't know the details,

1042
00:41:24,856 --> 00:41:25,416
but he's pushing.

1043
00:41:25,576 --> 00:41:25,876
Is it like for agent memory?

1044
00:41:26,276 --> 00:41:27,756
Yes, and I think in general,

1045
00:41:27,856 --> 00:41:28,896
the idea is like,

1046
00:41:28,956 --> 00:41:30,996
you need to be able to have an easy way

1047
00:41:30,996 --> 00:41:32,896
to export your memory or context wherever you go.

1048
00:41:32,896 --> 00:41:34,316
And the more I thought about it,

1049
00:41:34,316 --> 00:41:36,536
which is still limited, that's extremely important

1050
00:41:36,536 --> 00:41:38,196
because that's the way they lock you in.

1051
00:41:38,516 --> 00:41:39,216
That is the lock-in, right?

1052
00:41:39,276 --> 00:41:41,676
So if there's some kind of easy standard,

1053
00:41:41,676 --> 00:41:44,976
I do not expect labs to embrace that,

1054
00:41:45,136 --> 00:41:46,896
but if you can get more of a community sentiment

1055
00:41:46,896 --> 00:41:49,236
or more of a shelling point around how to do that,

1056
00:41:49,576 --> 00:41:51,796
that to me is probably the most important thing

1057
00:41:51,796 --> 00:41:53,196
to keep the ecosystem open.

1058
00:41:53,496 --> 00:41:55,956
But a tool like Buzz would be able to maybe even instigate,

1059
00:41:56,076 --> 00:41:59,316
but certainly use an open source memory protocol.

1060
00:41:59,336 --> 00:41:59,836
100%.

1061
00:41:59,836 --> 00:42:02,136
Is that something you guys have talked about?

1062
00:42:03,356 --> 00:42:04,076
I don't know.

1063
00:42:04,316 --> 00:42:04,996
Yeah, I don't know either.

1064
00:42:05,016 --> 00:42:05,896
Not allowed to talk about it.

1065
00:42:06,396 --> 00:42:08,156
But I am ready for our next topic.

1066
00:42:08,676 --> 00:42:08,996
Oh, boy.

1067
00:42:09,116 --> 00:42:09,776
He's getting ahead.

1068
00:42:10,316 --> 00:42:11,656
But I do think it's interesting.

1069
00:42:11,836 --> 00:42:13,036
I mean, you guys should chat with Elon.

1070
00:42:13,156 --> 00:42:13,476
I'm glad.

1071
00:42:13,576 --> 00:42:15,036
I'm also glad they all got to meet each other

1072
00:42:15,036 --> 00:42:16,776
because like Tom now gets to talk to him a little bit.

1073
00:42:17,296 --> 00:42:19,576
But I also think they were, on the Buzz point,

1074
00:42:19,696 --> 00:42:20,536
they were all very interested

1075
00:42:20,536 --> 00:42:23,096
because this is one of the first environments

1076
00:42:23,096 --> 00:42:25,096
where, you know, the Hermes agent

1077
00:42:25,096 --> 00:42:26,436
can talk with the OpenClaw agent.

1078
00:42:26,756 --> 00:42:28,596
And I was trying to kind of get that point across

1079
00:42:28,596 --> 00:42:31,496
of like, if you have, you know, a forum like Buzz,

1080
00:42:31,936 --> 00:42:34,116
open, you know, that's all based around open protocols,

1081
00:42:34,116 --> 00:42:37,936
they talked about 80A and how it was easy to put Hermes in there,

1082
00:42:38,036 --> 00:42:40,376
even though Hermes hadn't done any work to integrate with Buzz.

1083
00:42:41,036 --> 00:42:42,356
So anyway, I thought that was super cool.

1084
00:42:42,596 --> 00:42:44,096
It showed the importance of open protocols.

1085
00:42:44,896 --> 00:42:46,576
I listened to a lot of it, but not all of it.

1086
00:42:46,616 --> 00:42:52,956
Did you cover standard formats for traces?

1087
00:42:53,996 --> 00:42:55,236
Not on that panel, no.

1088
00:42:55,736 --> 00:42:56,516
That's a good question.

1089
00:42:57,296 --> 00:42:59,776
We didn't, but on the other panel that I did,

1090
00:43:00,476 --> 00:43:03,416
which was kind of focused on the economics of open weights models,

1091
00:43:03,416 --> 00:43:04,996
and then really we covered a lot of stuff.

1092
00:43:05,096 --> 00:43:06,676
I just basically asked Christian Catalini,

1093
00:43:06,776 --> 00:43:07,416
who I like a lot,

1094
00:43:07,796 --> 00:43:09,296
and this guy John Comcoff,

1095
00:43:09,356 --> 00:43:11,836
who's a really, really fascinating thinker.

1096
00:43:12,516 --> 00:43:14,136
A lot of thoughts around autonomous agents.

1097
00:43:14,496 --> 00:43:16,896
Both of them are very interested in,

1098
00:43:17,016 --> 00:43:19,316
I guess, that question,

1099
00:43:19,676 --> 00:43:23,936
how to sign and validate, I guess, traces as well.

1100
00:43:24,896 --> 00:43:26,076
So yeah, I mean,

1101
00:43:26,376 --> 00:43:27,976
I think it would be worth going even deeper

1102
00:43:27,976 --> 00:43:28,616
with them on that.

1103
00:43:29,116 --> 00:43:29,516
Cool.

1104
00:43:31,756 --> 00:43:33,116
Should we talk about Vora a little bit?

1105
00:43:33,116 --> 00:43:33,596
Let's do it.

1106
00:43:34,136 --> 00:43:37,836
So Vora is a company that operates out of PB.

1107
00:43:39,016 --> 00:43:42,556
Folks we know and I guess we're both investors.

1108
00:43:44,496 --> 00:43:49,136
But Jesse and Eric, they've been at PB from the beginning, I think.

1109
00:43:49,876 --> 00:43:53,256
And they did their product reveal today.

1110
00:43:53,456 --> 00:43:57,016
I mean, they've described what their product is about for quite a while now,

1111
00:43:57,016 --> 00:44:02,496
but they actually had hardware in hand and did an Apple-esque reveal,

1112
00:44:02,496 --> 00:44:03,856
which is kind of fun and cool.

1113
00:44:04,916 --> 00:44:08,216
But the industrial design, I think, really caught our eyes.

1114
00:44:08,976 --> 00:44:09,556
I forget who all.

1115
00:44:09,776 --> 00:44:11,496
I mean, I think maybe two of us.

1116
00:44:12,396 --> 00:44:15,456
Yeah, what popped out to you?

1117
00:44:15,556 --> 00:44:16,596
Well, okay, so first I have to say,

1118
00:44:16,656 --> 00:44:18,096
I actually somehow missed it today.

1119
00:44:18,136 --> 00:44:19,136
I didn't realize when Jesse was gone.

1120
00:44:19,216 --> 00:44:20,416
So I want to see the video later.

1121
00:44:20,796 --> 00:44:21,856
So I want to hear from you guys.

1122
00:44:22,016 --> 00:44:24,316
But I did get sort of a private reveal yesterday,

1123
00:44:24,676 --> 00:44:26,176
and I'm blown away.

1124
00:44:26,736 --> 00:44:29,136
And like, you know, Jesse and Eric are obviously very smart.

1125
00:44:29,556 --> 00:44:32,136
And, you know, I know Jesse has a long history with photography.

1126
00:44:32,496 --> 00:44:34,816
So of course, I was excited to back them.

1127
00:44:35,576 --> 00:44:37,556
But it was one thing to hear them talk about.

1128
00:44:37,616 --> 00:44:39,036
It was another thing to see the finished product.

1129
00:44:39,136 --> 00:44:40,436
And the guy they brought on that did hardware

1130
00:44:40,436 --> 00:44:41,616
clearly is like a stud.

1131
00:44:42,496 --> 00:44:43,436
I mean, I was literally,

1132
00:44:43,676 --> 00:44:45,176
you know, because one question I've always had for them

1133
00:44:45,176 --> 00:44:45,676
is like, you know,

1134
00:44:45,816 --> 00:44:47,596
I believe this is going to be very popular.

1135
00:44:47,816 --> 00:44:49,376
I think obviously Apple was super well positioned.

1136
00:44:49,496 --> 00:44:51,216
I think obviously Apple is still very well positioned for it.

1137
00:44:51,496 --> 00:44:52,796
You know, going after the consumer market

1138
00:44:52,796 --> 00:44:54,236
versus enterprise is very challenging.

1139
00:44:55,096 --> 00:44:56,256
But I do have to say,

1140
00:44:56,256 --> 00:44:58,216
like I see more of Jesse's vision now.

1141
00:44:58,616 --> 00:45:00,396
It was truly like a beautiful,

1142
00:45:00,556 --> 00:45:01,476
it was like a piece of wood,

1143
00:45:01,476 --> 00:45:07,516
like real wood embossed with their cool logo like the box like it was it doesn't look like a tech

1144
00:45:07,516 --> 00:45:12,276
product no it was it looks like something you want to display in your living room like it's

1145
00:45:12,276 --> 00:45:18,296
from the 70s like a speaker or something it feels natural like like these guys the biggest thing i

1146
00:45:18,296 --> 00:45:24,396
came away thinking is like this team now um it's craftsmen like the way to get really remarkable

1147
00:45:24,396 --> 00:45:29,296
products is to really give a fuck like really care and then also have taste and these guys do and

1148
00:45:29,296 --> 00:45:31,496
And yeah, I see the vision for like,

1149
00:45:31,576 --> 00:45:32,476
and for those of you who don't know,

1150
00:45:32,596 --> 00:45:33,416
in case we haven't talked about,

1151
00:45:33,696 --> 00:45:36,076
imagine a box you plug in at your house.

1152
00:45:36,396 --> 00:45:38,076
You can run local agents there.

1153
00:45:38,336 --> 00:45:39,836
There's secure elements on the chip.

1154
00:45:39,916 --> 00:45:41,096
They showed me the first motherboards.

1155
00:45:41,296 --> 00:45:42,196
That's so cool.

1156
00:45:42,236 --> 00:45:43,356
And they showed me the secure elements.

1157
00:45:43,356 --> 00:45:46,776
There's a honey badger on the PCB.

1158
00:45:47,296 --> 00:45:48,696
Yeah, and they showed me there's two different secure.

1159
00:45:48,896 --> 00:45:49,696
It's beautiful.

1160
00:45:50,116 --> 00:45:52,236
And also how quickly they got this stuff done.

1161
00:45:52,296 --> 00:45:53,556
And they were mentioning that obviously AI

1162
00:45:53,556 --> 00:45:54,936
is accelerating a lot of the process.

1163
00:45:55,296 --> 00:45:57,356
But like they got a lot of this stuff like manufactured

1164
00:45:57,356 --> 00:45:59,096
and like the first run of the motherboards

1165
00:45:59,096 --> 00:46:01,236
on the order of months, which would

1166
00:46:01,236 --> 00:46:03,176
normally have taken, I'm guessing, years

1167
00:46:03,176 --> 00:46:04,536
in a past life.

1168
00:46:05,556 --> 00:46:06,836
So yeah, I'm really excited.

1169
00:46:07,376 --> 00:46:08,376
So it's a hard robotic.

1170
00:46:08,816 --> 00:46:11,076
As a consumer, you buy it,

1171
00:46:11,416 --> 00:46:12,916
put it in your home, plug it in,

1172
00:46:13,456 --> 00:46:15,396
and it gives you privacy.

1173
00:46:16,176 --> 00:46:17,236
I know they put a lot

1174
00:46:17,236 --> 00:46:19,296
of attention and detail into security as well,

1175
00:46:19,496 --> 00:46:21,076
like protecting his prompt injection

1176
00:46:21,076 --> 00:46:23,296
attacks, which is a very difficult problem to solve.

1177
00:46:24,776 --> 00:46:25,136
And then

1178
00:46:25,136 --> 00:46:27,116
that's

1179
00:46:27,116 --> 00:46:28,356
the value proposition.

1180
00:46:28,356 --> 00:46:38,756
Do you use it basically like a web server to access almost like a chat GPT or a cloud-like interface directly from that as a web server, and then that has its own kind of GPU inside?

1181
00:46:39,396 --> 00:46:42,176
I don't know if you like SSH into it or it's a web server or not.

1182
00:46:42,636 --> 00:46:47,236
Yeah, I mean, I'm not sure of all the standards, but I think the ultimate thing is you have an app on your phone that connects.

1183
00:46:47,336 --> 00:46:48,976
I don't know if it's SSH or not.

1184
00:46:49,776 --> 00:46:54,596
But the one key thing I would call out is, and actually I'd be curious to get feedback from the audience.

1185
00:46:54,596 --> 00:46:59,156
Right now, it's a very capable machine that you can run a lot of good models on.

1186
00:46:59,576 --> 00:47:01,776
But the idea is not just that you do everything locally.

1187
00:47:01,876 --> 00:47:06,336
The idea is you can sort of put a barrier between what you want to be stored locally

1188
00:47:06,336 --> 00:47:08,596
for your memory and everything and outside calls.

1189
00:47:08,956 --> 00:47:13,496
So you could still use this thing to call an instance of Claude

1190
00:47:13,496 --> 00:47:15,456
or an instance of Kodaks or something else.

1191
00:47:15,656 --> 00:47:19,176
And then they're really trying to be, I guess, the protective membrane between those.

1192
00:47:19,676 --> 00:47:22,816
I do think, though, and I'm kind of like maybe pushing them a little bit.

1193
00:47:22,816 --> 00:47:23,896
I'm curious y'all's take.

1194
00:47:24,716 --> 00:47:26,976
They kind of want to go more mass market.

1195
00:47:27,476 --> 00:47:29,176
I'm telling them it would be nice to even consider

1196
00:47:29,176 --> 00:47:30,796
starting with a super premium product

1197
00:47:30,796 --> 00:47:32,516
with max out the RAM,

1198
00:47:32,756 --> 00:47:35,456
letting me run the juiciest, beefiest models possible,

1199
00:47:35,996 --> 00:47:37,536
charge a higher point.

1200
00:47:37,836 --> 00:47:40,476
The product they revealed today is 32 gigabyte?

1201
00:47:40,636 --> 00:47:41,036
64.

1202
00:47:41,416 --> 00:47:41,876
64?

1203
00:47:42,156 --> 00:47:42,856
Yeah, which is good.

1204
00:47:44,296 --> 00:47:46,556
Did they show you the light too?

1205
00:47:47,056 --> 00:47:47,296
Yeah.

1206
00:47:47,356 --> 00:47:48,136
Yeah, the light on the front.

1207
00:47:48,196 --> 00:47:48,596
It's sick.

1208
00:47:49,416 --> 00:47:50,736
Yeah, it's so well done.

1209
00:47:50,736 --> 00:47:51,056
I know.

1210
00:47:51,056 --> 00:47:53,956
I mean, people listening to this are probably like,

1211
00:47:54,036 --> 00:47:54,936
okay, it has a light.

1212
00:47:55,736 --> 00:47:56,996
But you just have to see it.

1213
00:47:56,996 --> 00:47:57,236
There's some wood.

1214
00:47:57,396 --> 00:48:00,276
It's a very warm, welcoming light.

1215
00:48:00,996 --> 00:48:02,636
I forget exactly how they described it,

1216
00:48:02,656 --> 00:48:03,596
but they described it in really wood.

1217
00:48:03,596 --> 00:48:05,296
It's like a natural kind of aesthetic, right?

1218
00:48:05,356 --> 00:48:05,556
Yeah.

1219
00:48:05,716 --> 00:48:07,676
It's anti-blue light, so it's kind of red.

1220
00:48:08,276 --> 00:48:10,336
Yeah, A-plus on the industrial design.

1221
00:48:10,376 --> 00:48:10,716
Unbelievable.

1222
00:48:10,876 --> 00:48:13,276
So hopefully it is super easy to use.

1223
00:48:13,736 --> 00:48:15,236
Like, plug it in and it works.

1224
00:48:15,676 --> 00:48:17,656
I don't know yet, but hopefully that's the case.

1225
00:48:18,916 --> 00:48:19,936
I can't wait to use it.

1226
00:48:19,936 --> 00:48:20,996
I mean, I'm also pushing them,

1227
00:48:21,216 --> 00:48:22,496
whatever we can do with pre-orders,

1228
00:48:22,596 --> 00:48:23,396
if it needs to be more expensive.

1229
00:48:23,396 --> 00:48:24,916
Did they announce a price point?

1230
00:48:25,776 --> 00:48:26,856
Well, I have some ideas,

1231
00:48:26,936 --> 00:48:27,956
but I don't know if that's supposed to be public.

1232
00:48:28,056 --> 00:48:29,196
But again, I told them,

1233
00:48:29,296 --> 00:48:31,176
if you could do a more expensive version

1234
00:48:31,176 --> 00:48:33,336
to get it out faster for the first run.

1235
00:48:33,696 --> 00:48:35,216
It says you can reserve it for $10.

1236
00:48:35,716 --> 00:48:36,596
Yeah, I'm telling you,

1237
00:48:36,656 --> 00:48:37,496
that should be changed.

1238
00:48:37,716 --> 00:48:39,216
It should be a little bit higher than $10.

1239
00:48:39,756 --> 00:48:42,336
But vora.io is the website.

1240
00:48:42,536 --> 00:48:43,976
V-O-R-A dot I-O.

1241
00:48:44,036 --> 00:48:46,096
So could you contribute it to like a mesh LLM?

1242
00:48:46,256 --> 00:48:48,256
So that's where the ultimate vision is, right?

1243
00:48:48,256 --> 00:48:51,516
So I think sort of phase zero is these beautiful machines.

1244
00:48:52,016 --> 00:48:59,036
But something that we've jammed a lot on over the years is like once you have that sort of infrastructure, now what can you build on top?

1245
00:48:59,176 --> 00:49:01,696
If only the people working on Meshulam would want to add Bitcoin.

1246
00:49:01,876 --> 00:49:04,516
And if only the people working on Vora would want to add Bitcoin.

1247
00:49:04,636 --> 00:49:04,856
Yeah.

1248
00:49:05,916 --> 00:49:06,556
If only.

1249
00:49:06,736 --> 00:49:07,296
I just got a lookout.

1250
00:49:08,976 --> 00:49:13,736
Which is one benefit of Bitcoiners immersing themselves in the AI world.

1251
00:49:13,856 --> 00:49:14,116
Yeah.

1252
00:49:14,316 --> 00:49:14,936
But I don't know.

1253
00:49:14,936 --> 00:49:17,256
I mean, Eric was telling me they're kind of leaning towards Sol right now.

1254
00:49:17,356 --> 00:49:17,756
So I will say.

1255
00:49:18,256 --> 00:49:20,076
Zcash

1256
00:49:20,076 --> 00:49:21,796
Zcash

1257
00:49:21,796 --> 00:49:23,196
Zcash please

1258
00:49:23,196 --> 00:49:26,016
this is the shortest episode we've ever done

1259
00:49:26,016 --> 00:49:27,636
I'm actually super happy we did it

1260
00:49:27,636 --> 00:49:29,736
we covered what we were hoping to cover

1261
00:49:29,736 --> 00:49:33,296
we probably didn't have a huge live audience

1262
00:49:33,296 --> 00:49:34,356
given the odd timing

1263
00:49:34,356 --> 00:49:36,516
but hopefully people watch it later

1264
00:49:36,516 --> 00:49:37,396
and

1265
00:49:37,396 --> 00:49:39,776
we can replay it tomorrow

1266
00:49:39,776 --> 00:49:40,776
yeah

1267
00:49:40,776 --> 00:49:44,556
anyway good chat guys

1268
00:49:44,556 --> 00:49:45,316
yeah this was fun

1269
00:49:45,316 --> 00:49:46,656
until next time

1270
00:49:48,256 --> 00:50:18,236
Thank you.