Bitcoin Security after COLDCARD, PB Media Archive Launch, Announcing the Type II Summit

1
00:00:00,000 --> 00:00:03,880
All right, Em, we're live.
2
00:00:03,900 --> 00:00:04,300
I think we're live.
3
00:00:04,600 --> 00:00:05,340
We are.
4
00:00:05,840 --> 00:00:06,600
Surprise, surprise.
5
00:00:07,200 --> 00:00:09,360
Can you see Under the Table?
6
00:00:09,680 --> 00:00:10,500
Yep, you can see Under the Table.
7
00:00:10,500 --> 00:00:11,440
Yeah, I can see Under the Table.
8
00:00:11,700 --> 00:00:13,700
So the FTSE game is all going live.
9
00:00:13,840 --> 00:00:14,300
I see the...
10
00:00:14,300 --> 00:00:16,200
Our producer's going to fix it a little bit,
11
00:00:16,280 --> 00:00:16,780
although then...
12
00:00:16,780 --> 00:00:17,600
We believe in transparency.
13
00:00:17,600 --> 00:00:18,200
Yeah, that's good.
14
00:00:18,360 --> 00:00:19,060
All right, there we go.
15
00:00:19,460 --> 00:00:20,400
Old transparency.
16
00:00:20,900 --> 00:00:21,340
Old transparency.
17
00:00:21,980 --> 00:00:22,740
Professional show.
18
00:00:22,900 --> 00:00:23,560
Professional show.
19
00:00:23,800 --> 00:00:24,720
Open source of everything.
20
00:00:25,040 --> 00:00:26,320
Open source of the studio.
21
00:00:26,320 --> 00:00:28,200
I wish I had my open source everything hat.
22
00:00:28,440 --> 00:00:29,840
We need to retweet this stuff.
23
00:00:30,000 --> 00:00:30,600
There we go.
24
00:00:30,640 --> 00:00:31,360
Get it out there.
25
00:00:31,780 --> 00:00:32,420
Jobs to do.
26
00:00:33,400 --> 00:00:34,140
You got jobs.
27
00:00:34,900 --> 00:00:35,460
What's going on?
28
00:00:35,740 --> 00:00:38,400
You were telling us about your snip, Max.
29
00:00:39,080 --> 00:00:39,420
Oh, wow.
30
00:00:39,420 --> 00:00:39,940
Your snip max.
31
00:00:40,580 --> 00:00:42,520
I'm not sure which snip you're talking about.
32
00:00:42,520 --> 00:00:49,580
We, our show, our episode last week was by far the most watched show ever.
33
00:00:49,740 --> 00:00:50,240
Oh, that's right.
34
00:00:50,260 --> 00:00:51,520
By like over 2X.
35
00:00:52,300 --> 00:00:53,140
I did predict it.
36
00:00:53,360 --> 00:00:53,780
Yeah, you did.
37
00:00:54,340 --> 00:00:54,600
I mean.
38
00:00:54,900 --> 00:00:55,960
It was also very timely.
39
00:00:56,320 --> 00:00:57,840
Cold guard incident was huge.
40
00:00:57,840 --> 00:01:00,960
we still have a much larger
41
00:01:00,960 --> 00:01:02,900
Bitcoiner audience I think
42
00:01:02,900 --> 00:01:04,340
than AI and tech
43
00:01:04,340 --> 00:01:07,200
we love our Bitcoiners
44
00:01:07,200 --> 00:01:09,260
we welcome Bitcoiners so thanks for listening
45
00:01:09,260 --> 00:01:10,900
we also want to grow our AI
46
00:01:10,900 --> 00:01:12,660
folks but we'll do that over time
47
00:01:12,660 --> 00:01:15,000
anyway glad that people
48
00:01:15,000 --> 00:01:16,680
and it seemed to be well received
49
00:01:16,680 --> 00:01:19,000
the show last week and guess what this week
50
00:01:19,000 --> 00:01:20,160
we're going to talk more about it
51
00:01:20,160 --> 00:01:21,820
less about I think
52
00:01:21,820 --> 00:01:24,960
cold card incidents specifically
53
00:01:24,960 --> 00:01:27,140
but sort of the what's the response
54
00:01:27,140 --> 00:01:29,560
from everyone and there's so much to talk about there.
55
00:01:29,720 --> 00:01:32,040
But we want to cover a couple of things first
56
00:01:32,040 --> 00:01:33,220
so we don't run out of time.
57
00:01:34,200 --> 00:01:35,820
Maybe you want to start us off, Max?
58
00:01:35,880 --> 00:01:37,240
And there was an event here yesterday
59
00:01:37,240 --> 00:01:39,100
which DK and I did not attend.
60
00:01:39,200 --> 00:01:39,520
You did.
61
00:01:39,660 --> 00:01:40,740
So we'd love to hear more about it.
62
00:01:40,740 --> 00:01:41,460
Holding it down.
63
00:01:42,360 --> 00:01:43,980
You know, Presbyterian Bitcoin, obviously,
64
00:01:44,260 --> 00:01:45,740
as you said, continuing to expand.
65
00:01:45,980 --> 00:01:48,060
I feel like all, you know, I've said this before
66
00:01:48,060 --> 00:01:49,700
and I actually believe it that we actually are,
67
00:01:49,780 --> 00:01:52,280
you know, coming from the foothills of the singularity
68
00:01:52,280 --> 00:01:54,060
into just like straight up the singularity right now.
69
00:01:54,160 --> 00:01:55,540
I mean, we'll talk a lot more about that
70
00:01:55,540 --> 00:01:56,820
with the code stuff,
71
00:01:56,840 --> 00:01:58,500
which I feel like it's going to be an ongoing topic.
72
00:01:58,660 --> 00:01:59,940
I mean, can you believe we're barely,
73
00:02:00,140 --> 00:02:01,560
I guess what, you know,
74
00:02:01,780 --> 00:02:03,320
just a little bit over halfway through the year?
75
00:02:03,560 --> 00:02:05,080
Like imagine where we are now
76
00:02:05,080 --> 00:02:06,300
compared to earlier in the year
77
00:02:06,300 --> 00:02:07,300
and imagine where we're going to be at the end.
78
00:02:07,760 --> 00:02:11,020
So anyways, as we approach the singularity,
79
00:02:11,720 --> 00:02:13,220
you know, one thing I think I'm realizing
80
00:02:13,220 --> 00:02:16,200
is that all sort of technologies are converging.
81
00:02:16,620 --> 00:02:18,080
And I think that's kind of
82
00:02:18,080 --> 00:02:18,960
part of the point of the singularity.
83
00:02:18,960 --> 00:02:22,960
So, you know, Bitcoin as open source standard for value,
84
00:02:23,120 --> 00:02:24,760
as energy backed money, as we've discussed,
85
00:02:24,760 --> 00:02:26,940
obviously is now starting to collide with AI.
86
00:02:27,120 --> 00:02:30,500
So we're doing a lot more AI at the office
87
00:02:30,500 --> 00:02:33,340
with the Open Source Summit, more people working here.
88
00:02:33,520 --> 00:02:34,520
We're also doing a lot more with energy.
89
00:02:34,680 --> 00:02:35,920
Last year, we had the Type 1 Summit,
90
00:02:36,040 --> 00:02:38,160
which was a big success and a lot of fun.
91
00:02:38,620 --> 00:02:43,720
Yesterday, there was a nonprofit energy group
92
00:02:43,720 --> 00:02:46,040
that held a working group during the day
93
00:02:46,040 --> 00:02:48,300
on sort of thinking about standards,
94
00:02:48,300 --> 00:02:51,340
open standards for energy transmission.
95
00:02:51,520 --> 00:02:53,240
So thinking about in the same way that we have
96
00:02:53,240 --> 00:02:59,120
open standards for the internet, TCPIP and related protocols, open standards for value
97
00:02:59,120 --> 00:03:06,980
transmission with Bitcoin, Lightning Network, etc. It's really shocking how antiquated the
98
00:03:06,980 --> 00:03:10,280
electricity grid is. This is something that we've talked about in the past, something I've
99
00:03:10,280 --> 00:03:14,860
been relatively in the weeds on for a while, although compared to these folks, definitely
100
00:03:14,860 --> 00:03:19,900
not as much. And it's kind of crazy because it's like, you know, in the same way, I mean,
101
00:03:19,900 --> 00:03:26,380
you would be shocked to see if you wanted to query sort of like data from a utility almost
102
00:03:26,380 --> 00:03:29,960
anywhere in the country, even in some of the more advanced markets like in Texas, you know,
103
00:03:30,020 --> 00:03:34,360
a lot of that data, it's like, even if they wanted to share it, it's extremely difficult,
104
00:03:34,480 --> 00:03:38,100
if possible, to share. I mean, for example, there's, you know, lots of things around,
105
00:03:38,240 --> 00:03:42,980
you know, sort of transmission line planning and, you know, power planning where, you know,
106
00:03:43,100 --> 00:03:48,220
the data may not even be in a database yet. I mean, there's like some crazy stuff. And so,
107
00:03:48,220 --> 00:03:54,920
in the past, that hasn't really been a problem because load growth and demand growth, you know,
108
00:03:54,960 --> 00:04:00,200
the new sort of energy and power coming online and the new sort of, I guess, offtakers of that
109
00:04:00,200 --> 00:04:08,540
electricity, you know, it's been very tiny single digit growth for decades. However, now we are,
110
00:04:08,540 --> 00:04:14,680
you know, with AI basically just completely changing that narrative. We're basically realizing
111
00:04:14,680 --> 00:04:19,160
the old grid is completely scorotic, completely incapable as it stands of expanding. And that was
112
00:04:19,160 --> 00:04:24,680
kind of the thesis of the essay I wrote last year, headmind.dc slash grid, which talked all about how
113
00:04:24,680 --> 00:04:29,300
the grid was basically too scorotic to grow. That was the thesis of the type one summit. I still
114
00:04:29,300 --> 00:04:34,680
personally am of the opinion that we need to be building alternative off-grid solar and storage
115
00:04:34,680 --> 00:04:40,600
private grids, probably DC first grids that are like tailor-made for these large multi-gigawatt
116
00:04:40,600 --> 00:04:44,980
data center loads first that could then potentially grow into new cities and take all kinds of loads.
117
00:04:45,320 --> 00:04:49,300
However, it was interesting to hear at this event, people that are much more focused on
118
00:04:49,300 --> 00:04:55,980
how do we reform the current grid? And basically, with open standards, can we make basically markets
119
00:04:55,980 --> 00:05:00,160
more? Basically, can we create markets in the first place? Can we make them more transparent?
120
00:05:00,540 --> 00:05:05,040
Can that enable, you know, potentially multiples of current load and generation to join the grid?
121
00:05:05,520 --> 00:05:08,800
So anyways, and then there was a public event last night, which I didn't hear everyone there,
122
00:05:08,800 --> 00:05:10,240
but I'm pretty familiar with the folks.
123
00:05:10,980 --> 00:05:15,300
And so we had Sam Brandon from Maginti Infrastructure,
124
00:05:15,400 --> 00:05:17,200
which is a company I'm actually invested in.
125
00:05:17,500 --> 00:05:19,380
They talked about a lot of the work that they're doing
126
00:05:19,380 --> 00:05:21,480
with battery storage systems
127
00:05:21,480 --> 00:05:23,860
to sort of expand the scope of the grid.
128
00:05:25,580 --> 00:05:27,940
One of the other public speakers was Bill Tai,
129
00:05:28,120 --> 00:05:31,280
who I didn't realize he was the first employed at TSMC
130
00:05:31,280 --> 00:05:33,160
or early at TSMC, which is kind of crazy.
131
00:05:33,540 --> 00:05:34,700
He's obviously got a long history.
132
00:05:34,940 --> 00:05:36,840
I think he was at Google too, maybe with you guys.
133
00:05:37,460 --> 00:05:37,780
I don't know.
134
00:05:37,780 --> 00:05:41,560
He does some Hawaii events too.
135
00:05:41,920 --> 00:05:42,020
Yeah.
136
00:05:42,260 --> 00:05:42,840
He does the kite.
137
00:05:42,840 --> 00:05:45,480
He used to do the Mai Tai kite camping, right?
138
00:05:45,540 --> 00:05:46,080
The surf camp.
139
00:05:46,640 --> 00:05:47,520
Yeah, sounds fun.
140
00:05:48,040 --> 00:05:50,840
But obviously, you know, he was involved with Hut 8 and the Bitcoin mining stuff.
141
00:05:50,980 --> 00:05:52,980
So that, you know, it was interesting to hear those different perspectives.
142
00:05:53,320 --> 00:05:59,340
But anyways, it's in the same way that we push for open standards here around value,
143
00:05:59,720 --> 00:06:03,220
identity with Nostr for agents, open source AI.
144
00:06:03,220 --> 00:06:05,940
I think it's just like shocking how backwards
145
00:06:05,940 --> 00:06:08,780
and how far back the energy grid is.
146
00:06:09,280 --> 00:06:11,000
And so, yeah, it's cool to see.
147
00:06:11,140 --> 00:06:13,180
What does the grid need in terms of like open standards?
148
00:06:13,300 --> 00:06:15,480
Does it need open standards for like communication
149
00:06:15,480 --> 00:06:18,080
of supply and demand?
150
00:06:18,300 --> 00:06:20,860
But what are the earliest problems that need to be solved?
151
00:06:20,860 --> 00:06:22,500
Like I have too much supply,
152
00:06:22,740 --> 00:06:25,100
do you take it for me or what kinds of?
153
00:06:25,560 --> 00:06:26,340
I mean, for one thing,
154
00:06:26,400 --> 00:06:28,680
you need to just be able to like query the data, right?
155
00:06:28,780 --> 00:06:30,680
Like, so right now, if you just wanted to see,
156
00:06:30,680 --> 00:06:34,920
I think Texas and ERCOT has by far the best versions of this.
157
00:06:35,200 --> 00:06:36,900
But in a lot of utilities around the country,
158
00:06:37,240 --> 00:06:40,340
basically these are government-granted monopolies.
159
00:06:40,520 --> 00:06:41,800
And so they have very little incentive.
160
00:06:41,920 --> 00:06:45,360
In fact, their pricing is not market-based.
161
00:06:45,560 --> 00:06:48,000
It's like there's a body.
162
00:06:48,520 --> 00:06:50,380
I mean, it's centralized planning.
163
00:06:50,500 --> 00:06:51,780
It's a body that basically says,
164
00:06:52,100 --> 00:06:53,940
okay, this is what you can charge
165
00:06:53,940 --> 00:06:55,900
as a multiple of the capital that you spend.
166
00:06:56,220 --> 00:06:57,460
And so the incentives are messed up
167
00:06:57,460 --> 00:07:00,020
because they're incentivized to pay a huge amount of money.
168
00:07:00,020 --> 00:07:04,360
the old world. I mean, it's kind of like, you know, the industrial complexes of various forms.
169
00:07:04,880 --> 00:07:10,160
But like, if you even wanted to say like, you know, okay, like, I don't know, like how much,
170
00:07:10,320 --> 00:07:15,700
I don't know, like, in real time, like, I want to be able to query exactly how much like one
171
00:07:15,700 --> 00:07:18,600
section of the grid is using versus another section of good, like, that's not, it's not
172
00:07:18,600 --> 00:07:22,220
possible. And when they're doing stuff like trying to do planning for new transmission lines,
173
00:07:22,500 --> 00:07:26,960
a lot of this data doesn't even live like in a single, I mean, in some cases, not even any
174
00:07:26,960 --> 00:07:33,400
database, which is crazy. Is that sometimes strategic to keep it isolated? Like, do the
175
00:07:33,400 --> 00:07:38,420
utilities want to keep them isolated for like strategic reasons or kind of competitive reasons?
176
00:07:38,820 --> 00:07:43,000
I think the biggest thing is just kind of how it evolved, where it's like, you know,
177
00:07:43,160 --> 00:07:48,160
the 1950s and 1960s, and a lot of this stuff was actually being developed. It just hasn't,
178
00:07:48,280 --> 00:07:51,560
because it's not exposed to competitive market forces, it's just not a lot of,
179
00:07:51,560 --> 00:07:57,600
a lot of reason to expand. To be fair, you know, to them, there are reasons to keep this data super
180
00:07:57,600 --> 00:08:02,380
siloed and opaque as well. Some of it, right? Like, I think there was a big push, you know,
181
00:08:02,380 --> 00:08:06,660
the sort of internet of electricity was becoming more of a thing in the late 90s.
182
00:08:07,120 --> 00:08:10,780
And then, you know, in the early 2000s, obviously, you have September 11. And I think that was like a
183
00:08:10,780 --> 00:08:14,940
big thing around sort of terrorist cyber attacks, people don't want the grids to go down, because
184
00:08:14,940 --> 00:08:21,200
that could obviously be really bad. So I think that was one force. And then another was Enron,
185
00:08:21,200 --> 00:08:26,100
basically blowing up because they were some of the most, let's say, exotic market-based
186
00:08:26,100 --> 00:08:27,540
participants at that time.
187
00:08:27,640 --> 00:08:29,480
And obviously, there was a lot of corruption on top of that.
188
00:08:30,520 --> 00:08:32,020
Anyways, it's interesting.
189
00:08:32,020 --> 00:08:39,760
I think the biggest TLDRs to me are we need standards for how data is stored, what kind
190
00:08:39,760 --> 00:08:42,900
of data can be queried, who can query what sort of data.
191
00:08:43,080 --> 00:08:45,620
I mean, a lot of this stuff involves just things we think about.
192
00:08:45,760 --> 00:08:50,600
I even mentioned Nostra to certain people's just very basic things of you need to be able
193
00:08:50,600 --> 00:08:54,780
to sign things with a private key. You need to be able to store it on, you know, is it a centralized
194
00:08:54,780 --> 00:08:58,540
database? Is it many different instances of a database? How do you credential with cookies?
195
00:08:58,540 --> 00:09:02,100
Who gets access to what pieces? All that kind of stuff needs to get figured out. It's very
196
00:09:02,100 --> 00:09:06,460
interesting. But I think to me, the biggest takeaway on the current grid, I've got, I guess,
197
00:09:06,520 --> 00:09:10,740
two big takeaways. On the current grid is the biggest opportunity is definitely in storage
198
00:09:10,740 --> 00:09:15,100
batteries, not a question. I think you guys probably saw this week, you know, Base Power
199
00:09:15,100 --> 00:09:19,960
announced another raise at a north of $10 billion valuation, which is kind of crazy.
200
00:09:20,600 --> 00:09:40,300
And, you know, if you really kind of understand what they're starting to do and some of the other players in this space as well, it's about trying to basically really figure out a network of storage on the grid and be able, whoever has the, not only the biggest, but the most strategically placed and, you know, ideally, whatever, has the most transparent data, which is in Texas.
201
00:09:40,300 --> 00:09:42,300
that's why you're seeing all this happen in Texas on ERCOC,
202
00:09:42,920 --> 00:09:43,780
because they have their own independent grid.
203
00:09:44,380 --> 00:09:51,060
You basically are in a position to be able to effectively market make for the whole grid.
204
00:09:51,260 --> 00:09:53,940
So we haven't really seen sort of the generation of, you know,
205
00:09:54,220 --> 00:09:55,620
Citadel or whatever coming out to the grid.
206
00:09:55,940 --> 00:09:58,620
That's what people, I think, that are betting on these kinds of companies
207
00:09:58,620 --> 00:09:59,720
like Base Power are betting on.
208
00:09:59,940 --> 00:10:01,520
I think they, in particular, I don't,
209
00:10:01,940 --> 00:10:03,760
I'm sure there's things I don't know about their next story
210
00:10:03,760 --> 00:10:06,260
because they're very focused on residential and installing batteries there.
211
00:10:06,420 --> 00:10:08,700
I'm a little bit more skeptical of that approach,
212
00:10:08,700 --> 00:10:12,460
having done residential solar for a long time
213
00:10:12,460 --> 00:10:16,320
versus doing kind of larger utility scale batteries
214
00:10:16,320 --> 00:10:17,340
and a network of that.
215
00:10:17,880 --> 00:10:21,020
And I am involved in some companies doing that right now.
216
00:10:21,540 --> 00:10:23,540
But anyways, I think both are very interesting
217
00:10:23,540 --> 00:10:26,240
and a lot more storage,
218
00:10:26,520 --> 00:10:28,640
a lot more transparency around data
219
00:10:28,640 --> 00:10:31,360
can probably get us to multiples of the current grid.
220
00:10:31,680 --> 00:10:33,380
But my other big takeaway
221
00:10:33,380 --> 00:10:34,620
that I keep pushing on everyone is,
222
00:10:34,620 --> 00:10:37,160
right now Texas has increased their peak load.
223
00:10:37,160 --> 00:10:40,360
I think we just saw the largest peak load ever
224
00:10:40,360 --> 00:10:41,940
in the last couple of days in Texas,
225
00:10:42,040 --> 00:10:43,180
and it was 90 gigawatts.
226
00:10:43,400 --> 00:10:45,380
So that means when everyone's running all their, you know.
227
00:10:45,460 --> 00:10:46,960
They got a heat wave going there.
228
00:10:47,460 --> 00:10:48,560
AC, all that stuff.
229
00:10:48,760 --> 00:10:49,760
I mean, there's heat wave going everywhere.
230
00:10:49,860 --> 00:10:51,560
Thank God for the fog.
231
00:10:52,560 --> 00:10:54,080
In my current state, I need the fog.
232
00:10:54,240 --> 00:10:57,140
But if you think about, you know,
233
00:10:57,140 --> 00:10:59,500
what could storage and, you know,
234
00:10:59,600 --> 00:11:02,220
potentially sort of like speeding up
235
00:11:02,220 --> 00:11:03,540
some of this transmission stuff do,
236
00:11:03,640 --> 00:11:05,180
I mean, you know, I think you could get to
237
00:11:05,180 --> 00:11:07,120
multiples of 90 gigawatts, right?
238
00:11:07,160 --> 00:11:09,540
by more intelligently managing the grid building.
239
00:11:09,640 --> 00:11:10,380
Better utilization, yeah.
240
00:11:10,740 --> 00:11:13,680
Yeah, but still, even let's be super generous
241
00:11:13,680 --> 00:11:15,680
and say you 5X that and you get to 450,
242
00:11:15,740 --> 00:11:16,780
even 500 gigawatts.
243
00:11:17,860 --> 00:11:19,580
The point that I keep coming back to
244
00:11:19,580 --> 00:11:21,020
and the sort of thesis that I'm pursuing
245
00:11:21,020 --> 00:11:23,140
very heavily at HiveMind is that's awesome.
246
00:11:23,300 --> 00:11:24,180
And there will be companies built
247
00:11:24,180 --> 00:11:24,780
that make a lot of money.
248
00:11:24,880 --> 00:11:26,020
It's going to deliver a lot of value.
249
00:11:26,420 --> 00:11:27,700
But at the same time, it's like,
250
00:11:27,880 --> 00:11:29,720
that's still not anywhere near a terawatt.
251
00:11:30,040 --> 00:11:31,180
And if you really believe,
252
00:11:31,260 --> 00:11:33,700
if you're really ASI, whatever you want to call it,
253
00:11:33,700 --> 00:11:35,600
AGI pills, we're going to be terawatts fast
254
00:11:35,600 --> 00:11:37,140
if we're really in the singularity.
255
00:11:37,160 --> 00:11:41,400
And I am just, you know, understanding how sclerotic these institutions are.
256
00:11:41,460 --> 00:11:42,840
I think it's great that there's pushing on it.
257
00:11:43,320 --> 00:11:48,280
I think that sort of building alternative systems that are completely new from the ground up
258
00:11:48,280 --> 00:11:52,040
is the only way that you can reach to the multi-gigawatts.
259
00:11:52,440 --> 00:11:52,840
Space.
260
00:11:53,380 --> 00:11:53,920
And space.
261
00:11:54,060 --> 00:11:57,420
And that's why, you know, I think, you know, people laughed at Elon going to space.
262
00:11:57,520 --> 00:11:59,960
They laughed at us when we had the type one panel on this last year.
263
00:12:00,000 --> 00:12:02,640
But it's like, this gets the other big thing, right?
264
00:12:02,660 --> 00:12:05,720
Is that it's not even a technological problem at this point.
265
00:12:05,720 --> 00:12:10,660
So you just mentioned, so PB hosted the Type 1 Energy Summit last fall.
266
00:12:11,340 --> 00:12:12,200
You led the charge there.
267
00:12:12,320 --> 00:12:12,820
Great job.
268
00:12:14,080 --> 00:12:18,000
Tell us about, well, so yesterday's event, I think, stemmed from that, right?
269
00:12:18,180 --> 00:12:21,640
And then can we expect a Type 2?
270
00:12:21,860 --> 00:12:22,100
Yeah.
271
00:12:23,020 --> 00:12:25,500
Or I don't know if it's, is it Type 2 or Type 1?
272
00:12:25,500 --> 00:12:26,640
Type 1 Part 2?
273
00:12:26,880 --> 00:12:28,720
No, it's Type 2.
274
00:12:28,840 --> 00:12:30,460
And I don't want to over-announce now.
275
00:12:30,540 --> 00:12:33,400
I realize this is the announcement of the announcement, you know, in classic style,
276
00:12:33,400 --> 00:12:35,460
because we don't have the website up yet.
277
00:12:35,720 --> 00:12:39,420
Um, but, uh, but yes, but you get the announcement, the announcement, this is a little bit of an
278
00:12:39,420 --> 00:12:40,640
alpha drop for our PV listeners.
279
00:12:40,780 --> 00:12:42,380
There will be a type two summit.
280
00:12:42,540 --> 00:12:45,820
Of course, type one, for those that don't remember, reference the Carter chef scale
281
00:12:45,820 --> 00:12:47,720
using all of the sun's energy touching the earth.
282
00:12:48,020 --> 00:12:50,640
Type two would be using all of the sun's energy out there, period.
283
00:12:50,740 --> 00:12:54,540
So the Dyson swarms sort of surrounding the sun sounds like crazy and sci-fi, but you
284
00:12:54,540 --> 00:12:56,700
know, when you're in the singularity, I think type two is the galaxy.
285
00:12:57,000 --> 00:12:57,980
Oh, well, there you go.
286
00:12:58,160 --> 00:12:59,160
Well, yeah.
287
00:12:59,340 --> 00:13:00,900
Type three is there's three.
288
00:13:01,060 --> 00:13:01,240
Yeah.
289
00:13:01,400 --> 00:13:01,900
Three is everything.
290
00:13:02,220 --> 00:13:04,140
There's actually like, there's actually like 10 of these.
291
00:13:04,140 --> 00:13:06,220
I thought one was all touching.
292
00:13:06,220 --> 00:13:06,940
Oh, I didn't know it was 10.
293
00:13:06,960 --> 00:13:07,760
Two was the solar system.
294
00:13:07,840 --> 00:13:08,300
Three was the galaxy.
295
00:13:08,480 --> 00:13:10,220
But we'll follow up on that.
296
00:13:11,300 --> 00:13:12,360
Regardless, you know.
297
00:13:12,500 --> 00:13:14,820
If only we had an LM fact-checking us right now.
298
00:13:14,980 --> 00:13:15,340
Exactly.
299
00:13:15,520 --> 00:13:16,420
If only we had.
300
00:13:16,920 --> 00:13:17,300
Gem chat.
301
00:13:17,320 --> 00:13:17,800
Sometimes do.
302
00:13:18,020 --> 00:13:18,760
If only we had gem chat.
303
00:13:19,860 --> 00:13:24,020
If you are interested in that, you know, stay tuned.
304
00:13:24,240 --> 00:13:27,680
I mean, if you, yeah, I mean, you can reach out to PD and let us know.
305
00:13:27,980 --> 00:13:28,520
I remember.
306
00:13:28,900 --> 00:13:31,960
Type 1 is all the energy reaching the Earth.
307
00:13:31,960 --> 00:13:32,280
Exactly.
308
00:13:32,280 --> 00:13:34,460
Type 2 is all of the sun's energy,
309
00:13:34,660 --> 00:13:36,800
and then type 3 is either galaxy or,
310
00:13:36,860 --> 00:13:37,540
I think it's galaxy.
311
00:13:37,660 --> 00:13:39,020
And then it would go beyond that.
312
00:13:39,100 --> 00:13:39,260
Exactly.
313
00:13:39,520 --> 00:13:40,420
You're right.
314
00:13:41,520 --> 00:13:44,200
But regardless, we'll have more information on that.
315
00:13:44,360 --> 00:13:45,1000
It's going to be super limited,
316
00:13:46,320 --> 00:13:48,220
so we're already over half capacity.
317
00:13:48,480 --> 00:13:50,040
I expect we'll fill up immediately.
318
00:13:50,280 --> 00:13:51,660
So if you are super interested,
319
00:13:52,060 --> 00:13:53,040
let us know sooner or later.
320
00:13:53,200 --> 00:13:54,520
We will put up a website at some point.
321
00:13:55,440 --> 00:13:57,080
Anyways, but all that to say,
322
00:13:57,440 --> 00:13:59,600
I do think it's worth just mentioning,
323
00:13:59,600 --> 00:14:05,100
you know, as we go through this singularity period, the reason that people are pushing for
324
00:14:05,100 --> 00:14:09,520
things like the Dyson swarm, um, is like what I was saying, it's not necessarily a political
325
00:14:09,520 --> 00:14:12,440
problem. I mean, sorry, a technical problem. It's more of a political problem at this point.
326
00:14:12,440 --> 00:14:17,400
And so even in Texas where they, you know, there's this whole batch zero process for those
327
00:14:17,400 --> 00:14:22,080
who are following, this is very in the, in the weeds for energy nerds. Um, but for the batch
328
00:14:22,080 --> 00:14:26,700
zero process, they, you know, uh, the governor basically just halted all these new projects
329
00:14:26,700 --> 00:14:30,080
that were supposed to get basically greenlit at one time,
330
00:14:30,200 --> 00:14:32,520
so effectively bringing on a tremendous amount of new,
331
00:14:32,800 --> 00:14:33,920
both loading capacity.
332
00:14:34,820 --> 00:14:40,040
But, you know, as data centers become, you know,
333
00:14:40,100 --> 00:14:41,260
basically more of a hot button issue
334
00:14:41,260 --> 00:14:42,220
for both the left and the right,
335
00:14:42,540 --> 00:14:44,620
I kind of expect this is going to be more of the thing around,
336
00:14:44,840 --> 00:14:46,400
basically, if this is happening in Texas,
337
00:14:46,540 --> 00:15:00,592
you know what going to be happening everywhere else And so it kind of like if we don get that going and maybe you only need one or two more to kind of like really wake the hyperscalers up either going and building new grids or going to space are the only options I see
338
00:15:01,232 --> 00:15:06,912
I like the other option that you didn't mention, which I just got my attention refocused on this week.
339
00:15:06,952 --> 00:15:08,332
I heard the Panthalassa story.
340
00:15:08,612 --> 00:15:09,412
Oh, absolutely.
341
00:15:09,732 --> 00:15:12,472
And you brought him to the Type 1 Summit.
342
00:15:12,952 --> 00:15:15,412
Yeah, I was an early investor in Panthalassa as well.
343
00:15:15,732 --> 00:15:17,232
So I'm a big fan of these guys.
344
00:15:17,232 --> 00:15:18,132
Take it to the oceans.
345
00:15:18,852 --> 00:15:19,732
Yeah, so Pantolosa.
346
00:15:20,612 --> 00:15:21,592
Fund one or fund two?
347
00:15:21,692 --> 00:15:21,972
Fund one.
348
00:15:22,272 --> 00:15:22,812
Okay, good.
349
00:15:23,032 --> 00:15:23,372
Swag.
350
00:15:25,812 --> 00:15:27,392
Yeah, so, and, you know,
351
00:15:27,472 --> 00:15:28,872
Pantolosa is a very interesting story.
352
00:15:29,012 --> 00:15:30,372
They're basically, yeah,
353
00:15:30,392 --> 00:15:32,392
putting data centers at sea.
354
00:15:33,772 --> 00:15:35,932
Totally unregulated, similar to space,
355
00:15:36,172 --> 00:15:38,032
but maybe less challenging
356
00:15:38,032 --> 00:15:39,332
than getting up to space, right?
357
00:15:39,632 --> 00:15:41,592
Yeah, and I highly recommend we have,
358
00:15:41,592 --> 00:15:44,232
if you go to the Presidio Bitcoin video playlist
359
00:15:44,232 --> 00:15:46,112
from Type One Summit last year,
360
00:15:46,112 --> 00:15:50,472
there's a really interesting conversation that I had with Garth, the founder of Pantolaza,
361
00:15:50,992 --> 00:15:56,612
and Shrep, the former meta CTO who's on his board. We had some really, really like pretty out there
362
00:15:56,612 --> 00:16:00,852
conversation around computing at high seas and sort of pirate compute and all that. I think it's
363
00:16:00,852 --> 00:16:05,292
less about the pirate compute and more just, yeah, it's like there's going to be a lot less,
364
00:16:05,432 --> 00:16:10,832
yeah, a lot less red tape, a lot less, I mean, fewer people living there. Let's put it that way.
365
00:16:11,972 --> 00:16:13,712
Security is an interesting question.
366
00:16:13,712 --> 00:16:15,292
Security is a very interesting question.
367
00:16:15,392 --> 00:16:17,232
So to be clear, I think I'm bullish on them.
368
00:16:17,312 --> 00:16:18,192
I'm really excited for them.
369
00:16:18,312 --> 00:16:19,632
I think there's a lot to figure out.
370
00:16:19,712 --> 00:16:22,272
I don't think that's going to be the whole sort of answer to the question.
371
00:16:22,652 --> 00:16:27,152
But I think we're at a point right now where it's like the old world is just showing it
372
00:16:27,152 --> 00:16:28,232
can't adapt quickly enough.
373
00:16:28,492 --> 00:16:33,072
I'm glad there are people pushing on sort of squeezing out what we can from the old world.
374
00:16:33,452 --> 00:16:34,432
New things too, yeah.
375
00:16:34,852 --> 00:16:35,212
Exactly.
376
00:16:35,352 --> 00:16:39,232
But we're in a case where, again, it's like if you really believe in the singularity,
377
00:16:39,252 --> 00:16:43,532
and I think this gets into a much deeper question as well, is of like, if you really think we're
378
00:16:43,532 --> 00:16:44,892
and get to the trillion dollar clusters.
379
00:16:44,972 --> 00:16:45,692
And we haven't even talked about
380
00:16:45,692 --> 00:16:46,832
what's happening in the AI markets
381
00:16:46,832 --> 00:16:47,612
in the last week or two.
382
00:16:48,092 --> 00:16:49,792
You know, how do we finance that?
383
00:16:49,912 --> 00:16:51,932
Do the hyperscalers have enough cash to finance that?
384
00:16:52,012 --> 00:16:53,272
Is the government going to underwrite that?
385
00:16:54,492 --> 00:16:55,632
And if they are,
386
00:16:55,692 --> 00:16:57,112
then they're going to have different sort of,
387
00:16:57,112 --> 00:16:59,172
you know, ideas about where that gets built.
388
00:16:59,292 --> 00:17:01,252
So there's a lot coming.
389
00:17:01,732 --> 00:17:03,572
The energy story, in my opinion,
390
00:17:03,732 --> 00:17:04,852
is, well, I would have said
391
00:17:04,852 --> 00:17:06,392
the hottest, most interesting story in the world,
392
00:17:06,552 --> 00:17:08,572
but we're also seeing, you know,
393
00:17:08,852 --> 00:17:10,492
agents collaborating autonomously.
394
00:17:10,632 --> 00:17:11,952
So that's just, yeah, in the singularity,
395
00:17:12,032 --> 00:17:12,832
it's too hard to pick.
396
00:17:13,532 --> 00:17:14,712
It's too hard to pick what's most interesting.
397
00:17:15,292 --> 00:17:19,212
So Maltbook was actually the sci-fi version of this, right?
398
00:17:19,332 --> 00:17:23,152
It was kind of the performance art version of this thing,
399
00:17:23,192 --> 00:17:24,072
but it actually happened now.
400
00:17:24,112 --> 00:17:24,772
It's happening now.
401
00:17:25,032 --> 00:17:25,172
Yeah.
402
00:17:25,332 --> 00:17:26,852
Should we talk about that or do we do in general?
403
00:17:26,852 --> 00:17:28,992
Yeah, well, let's get another hot topic.
404
00:17:29,112 --> 00:17:29,552
Another hot topic.
405
00:17:29,552 --> 00:17:32,852
Is the PB Media Archive launch.
406
00:17:33,312 --> 00:17:34,412
Actually, I stay corrected.
407
00:17:34,412 --> 00:17:37,092
The hottest topic in the singularity.
408
00:17:38,472 --> 00:17:40,132
Yeah, it might not top these other things,
409
00:17:40,252 --> 00:17:41,472
but it is exciting, I think.
410
00:17:41,472 --> 00:17:44,352
so let's knock it out so we can let people know about it.
411
00:17:45,212 --> 00:17:49,572
So this is episode 78 of our show,
412
00:17:49,672 --> 00:17:50,552
if you can believe it or not.
413
00:17:50,912 --> 00:17:52,232
So, you know, what?
414
00:17:53,172 --> 00:17:56,092
Well over 100 hours of content from the three of us.
415
00:17:56,612 --> 00:17:58,712
Haley has her 21 and 21 show.
416
00:17:58,852 --> 00:18:01,652
I think she's got something like 40 episodes of that.
417
00:18:03,112 --> 00:18:06,432
And then PB has done a number of other events
418
00:18:06,432 --> 00:18:07,972
with a bunch of other content too.
419
00:18:08,152 --> 00:18:10,952
So we've amassed a lot of content.
420
00:18:10,952 --> 00:18:20,252
great start and um a while back uh i was chatting with uh ashu our our producer uh who works out a
421
00:18:20,252 --> 00:18:27,032
pb here and he um about an idea to create like a topic index so something that the bitcoin optech
422
00:18:27,032 --> 00:18:33,532
project did a number of years ago uh which is a they have a really high quality weekly newsletter
423
00:18:33,532 --> 00:18:40,832
so lots of high quality content um hard to find something in the past though too tedious like
424
00:18:40,832 --> 00:18:43,352
it would take too much time. And also, if you just
425
00:18:43,352 --> 00:18:46,772
were interested in one topic, it was hard to know which newsletters
426
00:18:46,772 --> 00:18:49,652
covered it. So they created a topic index where you could just
427
00:18:49,652 --> 00:18:52,852
pick your topic of interest and drill down and see
428
00:18:52,852 --> 00:18:55,592
oh, these three newsletters spoke about it and you just have all the links
429
00:18:55,592 --> 00:18:58,352
available. I found the same problem with
430
00:18:58,352 --> 00:19:01,872
our show and with PB Media, which is
431
00:19:01,872 --> 00:19:03,792
like, I know we talked about
432
00:19:03,792 --> 00:19:07,912
how the security block
433
00:19:07,912 --> 00:19:09,852
used Buzz to
434
00:19:09,852 --> 00:19:12,192
investigate this vulnerability
435
00:19:12,192 --> 00:19:13,192
or what have you.
436
00:19:14,332 --> 00:19:16,172
But it's like, well, in that case,
437
00:19:16,292 --> 00:19:18,212
that actually was a real use case this week, but of course
438
00:19:18,212 --> 00:19:20,332
I knew it was last Friday's show, but still
439
00:19:20,332 --> 00:19:22,332
it's like, which timestamp? I don't want to have to go
440
00:19:22,332 --> 00:19:23,952
re-listen to 90 minutes.
441
00:19:25,892 --> 00:19:26,312
Ashu does
442
00:19:26,312 --> 00:19:28,192
a great job of creating timestamps
443
00:19:28,192 --> 00:19:29,832
for every YouTube video we have.
444
00:19:30,472 --> 00:19:31,992
Maybe like 10 to 15
445
00:19:31,992 --> 00:19:34,192
titles with timestamps.
446
00:19:34,272 --> 00:19:36,032
That's great, but still not exactly
447
00:19:36,032 --> 00:19:37,432
necessarily...
448
00:19:37,432 --> 00:19:39,272
this use case wasn't available.
449
00:19:39,412 --> 00:19:41,232
So I can just ask AI now
450
00:19:41,232 --> 00:19:44,072
at the PV Media Archive website
451
00:19:44,072 --> 00:19:46,532
and it instantly summarizes
452
00:19:46,532 --> 00:19:48,792
what we've talked about
453
00:19:48,792 --> 00:19:49,872
on whatever I ask
454
00:19:49,872 --> 00:19:51,932
and includes every episode
455
00:19:51,932 --> 00:19:53,272
that we've talked about it
456
00:19:53,272 --> 00:19:55,292
with specific timestamp links
457
00:19:55,292 --> 00:19:56,032
to those shows.
458
00:19:56,172 --> 00:19:57,412
So it's just an amazing tool.
459
00:20:01,072 --> 00:20:03,932
So like in the case of self-custody,
460
00:20:03,992 --> 00:20:05,252
we've talked about self-custody
461
00:20:05,252 --> 00:20:06,132
on a number of episodes.
462
00:20:06,132 --> 00:20:08,372
We didn't just start talking about it last week.
463
00:20:08,992 --> 00:20:13,292
So it can be really interesting to go back and find what did these guys talk about.
464
00:20:13,672 --> 00:20:14,892
So I think it's a really powerful tool.
465
00:20:15,032 --> 00:20:19,532
You can both look things up by topic, but you can also ask AI.
466
00:20:19,792 --> 00:20:21,672
And AI will give a really good summary.
467
00:20:22,032 --> 00:20:25,592
Also, humorously, you can ask what are our favorite sports teams.
468
00:20:25,972 --> 00:20:26,192
Oh, yeah.
469
00:20:26,872 --> 00:20:30,672
And it did nail that I'm an Iowa State guy, you're a Memphis guy,
470
00:20:30,972 --> 00:20:34,832
and it correctly called out DK as having gone to the University of Illinois,
471
00:20:34,832 --> 00:20:37,012
but doesn't really know what he's talking about
472
00:20:37,012 --> 00:20:38,512
when it comes to sports.
473
00:20:39,332 --> 00:20:42,012
I believe it also called someone named Bailey.
474
00:20:42,832 --> 00:20:44,012
So, close, close.
475
00:20:46,192 --> 00:20:48,072
Anyway, so encourage,
476
00:20:48,312 --> 00:20:49,632
I'm already using this tool.
477
00:20:49,852 --> 00:20:50,772
Like when people have a question,
478
00:20:50,872 --> 00:20:52,352
I just like send a link to this tool now
479
00:20:52,352 --> 00:20:54,672
and like to what we talked about.
480
00:20:55,732 --> 00:20:57,892
So hopefully people start using it and find it useful.
481
00:20:58,392 --> 00:21:00,832
We already had a request for an MCP server.
482
00:21:01,052 --> 00:21:02,992
So Ashu turned that around immediately.
483
00:21:03,312 --> 00:21:04,132
Yeah, I saw Ashu did that
484
00:21:04,132 --> 00:21:05,712
I didn't realize someone actually requested it.
485
00:21:05,772 --> 00:21:06,032
Oh, yeah.
486
00:21:06,752 --> 00:21:08,552
I know you're not on Twitter anymore,
487
00:21:08,652 --> 00:21:09,972
but I tweeted this out
488
00:21:09,972 --> 00:21:11,712
and people are responding with enthusiasm.
489
00:21:11,932 --> 00:21:12,612
Go ahead, DK.
490
00:21:12,852 --> 00:21:14,012
I was going to say, I saw the request
491
00:21:14,012 --> 00:21:17,012
and then I saw that you shared the personal requesting,
492
00:21:17,432 --> 00:21:18,712
but I was trying to understand more
493
00:21:18,712 --> 00:21:20,512
about the intention of the MCP.
494
00:21:20,652 --> 00:21:21,932
Like that guy has an application
495
00:21:21,932 --> 00:21:24,772
that wanted to pull data
496
00:21:24,772 --> 00:21:27,412
about something we had discussed on our show.
497
00:21:27,532 --> 00:21:29,552
I was just trying to get my head around the use case
498
00:21:29,552 --> 00:21:31,912
because it might inspire other use cases.
499
00:21:31,912 --> 00:21:35,172
He wanted to use our content as input into him building software.
500
00:21:35,372 --> 00:21:39,232
I actually warned him that you should not use our content to build software.
501
00:21:39,572 --> 00:21:39,932
That's a terrible idea.
502
00:21:40,032 --> 00:21:41,132
What exactly is he building?
503
00:21:42,352 --> 00:21:47,272
I think you could pull our content to ideate around what to build.
504
00:21:47,412 --> 00:21:53,452
I would not want the technical code to use our input.
505
00:21:53,792 --> 00:21:54,072
Brainstorming.
506
00:21:54,332 --> 00:21:54,812
Yeah, exactly.
507
00:21:55,112 --> 00:21:55,672
Brainstorming is great.
508
00:21:56,272 --> 00:21:58,792
But yeah, anyway, I don't know.
509
00:21:58,792 --> 00:22:01,372
But I mean, like doing that, Asha did it super quickly.
510
00:22:01,372 --> 00:22:02,672
so way to ship.
511
00:22:03,432 --> 00:22:05,872
And we'll see what kind of creativity.
512
00:22:06,692 --> 00:22:09,592
Well, and also I believe this is his first software project ever
513
00:22:09,592 --> 00:22:11,232
and first VibeCoding project ever.
514
00:22:11,412 --> 00:22:13,152
And dude, you killed it, man.
515
00:22:13,372 --> 00:22:13,652
It was awesome.
516
00:22:13,652 --> 00:22:14,612
Yeah, Ash is a developer.
517
00:22:15,152 --> 00:22:19,372
Another thing that sort of struck me is that,
518
00:22:19,852 --> 00:22:20,932
well, especially in the past week,
519
00:22:21,132 --> 00:22:23,872
podcasters are taking a lot of hate and heat
520
00:22:23,872 --> 00:22:26,572
from the cold card incident.
521
00:22:26,572 --> 00:22:34,432
And I actually, I think it's not, I think it's a little overblown.
522
00:22:34,752 --> 00:22:41,792
But nevertheless, they are taking hate for having promoted cold card.
523
00:22:41,972 --> 00:22:46,932
Now, they also, many were also sponsored by CoinKite.
524
00:22:48,092 --> 00:22:55,232
Anyway, I think about a tool like this, undoubtedly in the future, we are going to take heat.
525
00:22:55,232 --> 00:22:57,712
something's going to happen
526
00:22:57,712 --> 00:23:00,572
and then you just simply ask the search engine
527
00:23:00,572 --> 00:23:01,912
what do these guys say
528
00:23:01,912 --> 00:23:04,352
and then oh four years ago they said this and that
529
00:23:04,352 --> 00:23:05,572
about this product or whatever
530
00:23:05,572 --> 00:23:07,712
is Iowa State really the best
531
00:23:07,712 --> 00:23:09,552
like what?
532
00:23:09,992 --> 00:23:12,292
turns out Caitlin Clark says otherwise
533
00:23:12,292 --> 00:23:14,132
so
534
00:23:14,132 --> 00:23:16,232
I fully
535
00:23:16,232 --> 00:23:17,352
expect that but
536
00:23:17,352 --> 00:23:19,892
I will provide our disclaimer now
537
00:23:19,892 --> 00:23:22,232
we are just three guys we don't take sponsorship
538
00:23:22,232 --> 00:23:24,272
money we just speak
539
00:23:24,272 --> 00:23:24,972
our minds
540
00:23:24,972 --> 00:23:28,472
we try to come at it with humility.
541
00:23:28,892 --> 00:23:30,492
We don't claim to have all the answers.
542
00:23:31,012 --> 00:23:33,392
We're just sharing information we have.
543
00:23:33,732 --> 00:23:36,412
We're sharing our own analysis and thoughts on it.
544
00:23:36,752 --> 00:23:38,352
We try to share our questions too.
545
00:23:38,632 --> 00:23:39,672
Like if we don't know the answer,
546
00:23:39,792 --> 00:23:41,652
we're not afraid to say we don't know the answer.
547
00:23:43,232 --> 00:23:45,032
And we're not always going to be right.
548
00:23:45,632 --> 00:23:48,532
In fact, I mean, just to push back on that, Steve,
549
00:23:48,532 --> 00:23:51,832
if you go back to our end of year predictions two years ago,
550
00:23:52,172 --> 00:23:53,732
you'll see that not only were we not right,
551
00:23:53,732 --> 00:23:55,172
We were almost entirely wrong.
552
00:23:55,452 --> 00:23:57,032
So, you know, it is what it is.
553
00:23:57,112 --> 00:23:59,292
I mean, we obviously, let's put it this way.
554
00:23:59,312 --> 00:24:01,572
I think we're always having fun and ideating,
555
00:24:01,752 --> 00:24:02,792
but I think it's a very good,
556
00:24:03,292 --> 00:24:04,592
don't take what we say as gospel.
557
00:24:04,752 --> 00:24:07,372
And the spirit of it is truth-seeking, right?
558
00:24:07,372 --> 00:24:07,612
Of course.
559
00:24:07,972 --> 00:24:10,112
We're trying to wrestle with things that are unknown,
560
00:24:10,232 --> 00:24:11,412
that are kind of happening in real time.
561
00:24:11,632 --> 00:24:13,692
But when you're going through the singularity
562
00:24:13,692 --> 00:24:15,212
in like extremely complex environments,
563
00:24:15,512 --> 00:24:16,912
like no one knows, right?
564
00:24:16,992 --> 00:24:17,992
And that's the fun of it.
565
00:24:18,012 --> 00:24:18,912
That's why it's fun to talk about.
566
00:24:19,072 --> 00:24:22,072
Yeah, so people should do their own research.
567
00:24:22,072 --> 00:24:25,472
don't consider this investment advice or product advice.
568
00:24:26,092 --> 00:24:28,092
I hope people learn from us
569
00:24:28,092 --> 00:24:31,192
and become curious to go explore more themselves
570
00:24:31,192 --> 00:24:32,592
and come to their own conclusions.
571
00:24:33,332 --> 00:24:36,192
I actually do want to just follow up on this one thing you said.
572
00:24:36,192 --> 00:24:36,612
You feeling this back?
573
00:24:36,612 --> 00:24:38,552
I mean, feeling what?
574
00:24:39,372 --> 00:24:40,072
What are you doing, Jay?
575
00:24:41,412 --> 00:24:42,772
A little up and to the right.
576
00:24:45,352 --> 00:24:45,752
BlockFi.
577
00:24:46,432 --> 00:24:47,412
Oh, exactly.
578
00:24:47,632 --> 00:24:49,832
I was like, okay, I'm not sure where you're going right now.
579
00:24:49,832 --> 00:24:51,132
I know you're not feeling great today,
580
00:24:51,132 --> 00:24:52,552
but I don't know how crazy D.C. has gone.
581
00:24:54,172 --> 00:24:55,292
No comment.
582
00:24:57,092 --> 00:25:01,512
Yeah, so I will say one thing,
583
00:25:01,692 --> 00:25:04,472
and not to belabor individual points,
584
00:25:04,572 --> 00:25:05,092
everyone can do,
585
00:25:05,172 --> 00:25:06,552
and I would also say do your own assessment.
586
00:25:07,372 --> 00:25:09,712
No heroes in the world, no heroes.
587
00:25:10,132 --> 00:25:11,872
Every time you look up to someone
588
00:25:11,872 --> 00:25:13,852
and export your own thinking to them,
589
00:25:14,072 --> 00:25:15,892
it's bad, it's going to end poorly, we'll stop.
590
00:25:16,332 --> 00:25:17,712
It doesn't matter who the person is,
591
00:25:17,812 --> 00:25:18,872
think for yourself, always.
592
00:25:19,472 --> 00:25:21,952
But I will say one thing
593
00:25:21,952 --> 00:25:23,832
that this got me thinking about a little bit,
594
00:25:24,152 --> 00:25:25,792
and not so much this specific incident,
595
00:25:25,952 --> 00:25:26,692
but just in general,
596
00:25:27,012 --> 00:25:27,852
I mean, to your point, DK,
597
00:25:28,492 --> 00:25:30,052
the BlockFi rocket ship,
598
00:25:30,392 --> 00:25:34,432
you know, there's got to be a better model
599
00:25:34,432 --> 00:25:37,272
for monetization in this world
600
00:25:37,272 --> 00:25:38,412
for things like podcasts
601
00:25:38,412 --> 00:25:39,532
and just like the internet in general.
602
00:25:39,612 --> 00:25:40,732
I know that's something we've talked about a lot.
603
00:25:40,832 --> 00:25:41,872
We've had the value for value,
604
00:25:42,252 --> 00:25:43,272
you know, apps like Fountain
605
00:25:43,272 --> 00:25:45,492
that allow you to stream by,
606
00:25:45,592 --> 00:25:47,032
which, you know, I'm an investor in those guys too
607
00:25:47,032 --> 00:25:48,092
and love what they're trying to do.
608
00:25:48,092 --> 00:25:52,832
stacker news to a degree kind of gets this some of the nostrils stuff um but the reality is obviously
609
00:25:52,832 --> 00:25:55,952
the amount of money you can make from just like donations from your audience unless you have a
610
00:25:55,952 --> 00:26:01,212
massive massive massive audience like adam curry is pretty small compared to what um you know
611
00:26:01,212 --> 00:26:08,232
sort of well-known podcasters make and i you know i um it's unfortunate because you know it's very
612
00:26:08,232 --> 00:26:12,292
unfortunate because there's a lot of podcasts that i deeply deeply enjoy and even some of the
613
00:26:12,292 --> 00:26:15,572
mainstream ones like there'll be like episodes of rogan world like oh this is great we're talking
614
00:26:15,572 --> 00:26:19,092
about you know ancient civilizations and then right in the middle where he's like you know and
615
00:26:19,092 --> 00:26:22,692
the biggest lesson they taught us is to get back to the land and get away from all this crap and
616
00:26:22,692 --> 00:26:27,332
then all the stuff be like brought to you by your friends at chewy like you're like why like why does
617
00:26:27,332 --> 00:26:32,292
everything have to be ad based i know it's just like jesus and especially for the people that
618
00:26:32,292 --> 00:26:36,132
have already made it like super big it's like have some whatever but but i also understand for people
619
00:26:36,132 --> 00:26:40,372
that like this is your full-time job you need a way to make a living and i just there's got to be
620
00:26:40,372 --> 00:26:45,252
a better way because like you know especially in our industry or sort of like bitcoin and crypto
621
00:26:45,252 --> 00:26:50,712
land um if you go back i actually think way worse than you know this last thing is you know how many
622
00:26:50,712 --> 00:26:54,612
people were back in the day showing block fi how many people were back in the day showing ftx all
623
00:26:54,612 --> 00:26:59,692
this stuff and at the end of the day terrible i mean just it's unbelievable and it's like the more
624
00:26:59,692 --> 00:27:04,512
you just realize you know i hate to say it but like the more sort of like plugged into the mainstream
625
00:27:04,512 --> 00:27:07,492
which is why i'm kind of glad to be on twitter right now you are the more you're just like
626
00:27:07,492 --> 00:27:15,032
the vast majority of people are their incentives are not aligned with yours and um there's got to
627
00:27:15,032 --> 00:27:17,952
be a better way. I don't know what that better way is. I would like to think it's value for value.
628
00:27:18,052 --> 00:27:21,692
I don't think that is realistic for people that need to support a mortgage and a family and stuff
629
00:27:21,692 --> 00:27:24,992
yet. But man, there's got to be, because the advertising model, it just sucks.
630
00:27:26,312 --> 00:27:32,072
Thinking about this more, I made the claim that we don't take sponsorship. I will say though,
631
00:27:32,192 --> 00:27:37,312
the three of us are investors and we do tend to talk about companies that are in products from
632
00:27:37,312 --> 00:27:44,932
our investments. We do tend to say we invested. Yeah. So people should take that bias into account.
633
00:27:45,032 --> 00:27:46,932
I think we always try to call it out.
634
00:27:48,452 --> 00:27:51,372
We also aren't afraid to hold back punches.
635
00:27:51,892 --> 00:27:53,952
If we don't like something about a product
636
00:27:53,952 --> 00:27:55,912
or a company that we invested in.
637
00:27:56,012 --> 00:27:56,752
And we've done that many times.
638
00:27:56,752 --> 00:27:57,472
And we've done that.
639
00:27:58,112 --> 00:27:58,912
We've done that.
640
00:28:00,152 --> 00:28:01,912
Also, we don't...
641
00:28:02,672 --> 00:28:04,132
Just because we're an investor in something,
642
00:28:04,492 --> 00:28:06,312
it's not like the way we come about topics
643
00:28:06,312 --> 00:28:07,932
is like, oh, we got to make sure
644
00:28:07,932 --> 00:28:08,892
to talk about our investments.
645
00:28:09,112 --> 00:28:09,572
It's more just like,
646
00:28:09,652 --> 00:28:12,072
are they doing something interesting to us
647
00:28:12,072 --> 00:28:13,252
that we want to talk about?
648
00:28:13,272 --> 00:28:13,952
And if we invest in it,
649
00:28:13,952 --> 00:28:14,892
the answer is probably yes.
650
00:28:15,032 --> 00:28:15,292
great.
651
00:28:16,792 --> 00:28:19,092
But also, we talk about tons
652
00:28:19,092 --> 00:28:21,152
of products and
653
00:28:21,152 --> 00:28:23,052
companies that we're not investors in, have
654
00:28:23,052 --> 00:28:25,172
no financial ties in. We're not afraid
655
00:28:25,172 --> 00:28:27,312
of that. We talked about Anchor Watch
656
00:28:27,312 --> 00:28:29,192
before. Love Rob.
657
00:28:29,352 --> 00:28:31,132
Love what they're doing. Love design. We talked about all
658
00:28:31,132 --> 00:28:33,152
kinds of great stuff about it. We're not investors, but
659
00:28:33,152 --> 00:28:34,332
I want them to succeed.
660
00:28:34,692 --> 00:28:36,752
We're Bitcoiners. We want to see them succeed.
661
00:28:37,952 --> 00:28:39,232
And he's been putting down real
662
00:28:39,232 --> 00:28:40,872
effort on the cold card.
663
00:28:41,792 --> 00:28:43,612
I'm sure we'll get to that.
664
00:28:43,612 --> 00:28:48,852
And all of that's true. And again, verify with everyone. Don't make anyone a hero, like I said.
665
00:28:49,032 --> 00:28:54,792
But I just want to come back and I don't know. I mean, Deke, I know you're always ideating on
666
00:28:54,792 --> 00:28:59,112
weird things. I'm both. What do you think? There's got to be a better way to make money
667
00:28:59,112 --> 00:29:03,972
than having to like... And you could push back and say, are there products that I genuinely love
668
00:29:03,972 --> 00:29:09,032
that I would not feel ashamed to take money from? One example, I really like Cash App. I genuinely
669
00:29:09,032 --> 00:29:13,312
recommend Cash App to everyone. Also doesn't mean Cash App's not going to... Probably they're
670
00:29:13,312 --> 00:29:16,092
It's got to come through subscription models or premium.
671
00:29:16,292 --> 00:29:18,432
I mean, value for value is like volunteer.
672
00:29:18,772 --> 00:29:21,032
It would be nice to see that take off,
673
00:29:21,152 --> 00:29:24,812
but people generally don't like to part ways with their money.
674
00:29:26,772 --> 00:29:28,532
Both if they don't have to,
675
00:29:28,592 --> 00:29:30,672
but also if there's any kind of friction involved.
676
00:29:31,852 --> 00:29:33,952
So gating stuff through subscriptions.
677
00:29:34,852 --> 00:29:36,752
But you have to have established...
678
00:29:36,752 --> 00:29:37,232
Big audience.
679
00:29:37,852 --> 00:29:39,232
Yeah, and its reputation.
680
00:29:39,572 --> 00:29:52,705
What Tritechery or whatever they do that into a little some of it free and that gives you teasers And then it like oh I really like this content and then you pay And I think they made a killing over Ben made a lot of money over like 10 years now right
681
00:29:52,844 --> 00:29:54,324
And others have done that.
682
00:29:54,804 --> 00:29:55,544
I think it's hard.
683
00:29:55,725 --> 00:29:58,064
They kind of, I'd say, almost invented the model
684
00:29:58,064 --> 00:30:00,485
and there's other people who've been successful with it.
685
00:30:01,245 --> 00:30:03,164
But I think it's, you know,
686
00:30:03,225 --> 00:30:07,064
starting now is really hard to make a business like that.
687
00:30:07,324 --> 00:30:09,164
You kind of start when there's not a lot of competition.
688
00:30:09,564 --> 00:30:10,485
It's a lot easier.
689
00:30:10,705 --> 00:30:15,264
There's also just been a mild buying spree recently, right?
690
00:30:15,304 --> 00:30:20,564
I mean, it was rumored that OpenAI spent nine figures on TBPN.
691
00:30:21,304 --> 00:30:25,584
And Coinbase bought some podcasts prior to that.
692
00:30:25,664 --> 00:30:27,005
I heard Block's acquiring PBJ.
693
00:30:27,205 --> 00:30:29,604
Block's going to acquire PBJ.
694
00:30:29,904 --> 00:30:32,104
One of the companies, I wouldn't mind that happening.
695
00:30:32,205 --> 00:30:32,505
Just saying.
696
00:30:35,144 --> 00:30:38,424
Which actually is another bias is I work for Block.
697
00:30:38,424 --> 00:30:41,344
And Max's opinion is for sale always.
698
00:30:42,044 --> 00:30:43,264
I work for Block.
699
00:30:43,804 --> 00:30:45,184
Actually, it's not. You know that.
700
00:30:45,524 --> 00:30:48,144
It is as long as it's Block buying it.
701
00:30:49,064 --> 00:30:52,604
I'd say that our bias, it's less about like,
702
00:30:52,684 --> 00:30:55,205
oh, we want to tout Block or our investment.
703
00:30:55,344 --> 00:30:58,764
It's more just like that's the information flow we have.
704
00:30:59,324 --> 00:31:04,144
I tend to have a lot of information about Block, Spiral,
705
00:31:04,364 --> 00:31:06,164
a lot of open source projects I work on or whatever.
706
00:31:06,505 --> 00:31:07,904
Well, those are natural things that like,
707
00:31:07,904 --> 00:31:09,844
But there's also a reason you chose to work at Block, right?
708
00:31:09,864 --> 00:31:12,364
And there's a reason in most of our investments we chose to invest.
709
00:31:12,624 --> 00:31:14,764
Like, all of that's true, but also it's like,
710
00:31:14,824 --> 00:31:18,084
I actually think in our experiences so far,
711
00:31:18,184 --> 00:31:19,524
we've done, I don't know,
712
00:31:19,584 --> 00:31:24,084
like I think all of us have been pretty values-based.
713
00:31:24,245 --> 00:31:26,284
Like, there's a reason we didn't get into like crypto stuff.
714
00:31:26,344 --> 00:31:27,465
There's a reason, as you know, DK,
715
00:31:27,644 --> 00:31:29,884
like I left that whole world behind
716
00:31:29,884 --> 00:31:31,505
as I started seeing like, what the hell is this?
717
00:31:32,044 --> 00:31:33,184
Anyways, no one's perfect.
718
00:31:33,485 --> 00:31:34,304
But I hear you on that.
719
00:31:34,384 --> 00:31:36,664
I mean, I also think about maybe it's kind of like with bands
720
00:31:36,664 --> 00:31:40,044
and, you know, that maybe the podcast
721
00:31:40,044 --> 00:31:41,584
is just sort of the Legion, right?
722
00:31:41,604 --> 00:31:43,645
And so, you know, obviously for folks that are investors,
723
00:31:43,844 --> 00:31:45,485
like that opens up investment opportunities,
724
00:31:45,485 --> 00:31:46,904
that's one way to monetize it.
725
00:31:47,245 --> 00:31:49,584
Obviously, you know, maybe the answer is
726
00:31:49,584 --> 00:31:51,824
it's just kind of like Legion for doing like live shows.
727
00:31:51,824 --> 00:31:53,445
Like maybe we got to do, you know,
728
00:31:53,645 --> 00:31:57,324
sell out Chase and her to do our next PBJ
729
00:31:57,324 --> 00:31:59,544
with all of our fans.
730
00:31:59,645 --> 00:32:00,164
I don't know.
731
00:32:00,264 --> 00:32:01,684
But my point is like the ad model.
732
00:32:01,684 --> 00:32:02,284
Both of our fans.
733
00:32:03,584 --> 00:32:05,344
I wish we could find a way to move this.
734
00:32:05,344 --> 00:32:09,044
We're fortunate because we don't have to make money on this show.
735
00:32:09,145 --> 00:32:09,544
We're very lucky.
736
00:32:10,245 --> 00:32:12,764
Others do if it's their livelihood or their whole business.
737
00:32:13,184 --> 00:32:17,104
So it's a worthwhile topic to keep thinking about to just help others.
738
00:32:17,104 --> 00:32:22,184
Because I do think, I mean, one thing I started pushing this morning was that, hey, let's raise the bar.
739
00:32:22,445 --> 00:32:28,505
Like, so the PB Media Archive, it just dawned on me that it's kind of novel.
740
00:32:29,044 --> 00:32:34,904
So I did a quick AI query to see what other podcasts have something like this.
741
00:32:35,344 --> 00:32:37,824
And have you guys heard of Lenny's podcast?
742
00:32:38,745 --> 00:32:39,424
Like Lenny or Chisky?
743
00:32:39,864 --> 00:32:40,764
I don't know.
744
00:32:41,164 --> 00:32:43,604
I didn't have enough time this morning to click through.
745
00:32:43,725 --> 00:32:46,324
He's like the big sub stack guy.
746
00:32:46,804 --> 00:32:46,965
Okay.
747
00:32:47,584 --> 00:32:48,225
So probably him.
748
00:32:48,384 --> 00:32:49,604
But that was the only one.
749
00:32:49,764 --> 00:32:50,145
The big one?
750
00:32:50,505 --> 00:32:52,745
I sent out a chart.
751
00:32:53,124 --> 00:32:55,465
Basically, it is almost unprecedented.
752
00:32:55,705 --> 00:33:00,404
Apparently, his podcast is as transparent and searchable as ours.
753
00:33:00,725 --> 00:33:02,044
But most aren't, right?
754
00:33:02,044 --> 00:33:05,985
So I think we should, hopefully we helped set the bar for other podcasts.
755
00:33:06,384 --> 00:33:08,164
They should do similar things that we did.
756
00:33:08,624 --> 00:33:09,664
Proof of reserves for podcasts.
757
00:33:09,764 --> 00:33:13,584
There was an ask to open source PB Media Archive,
758
00:33:13,664 --> 00:33:15,044
and I think Ashi is going to do that as well.
759
00:33:15,424 --> 00:33:16,924
Because it was vibe-coded, you know,
760
00:33:17,044 --> 00:33:20,884
I don't know if it's like a code base to like fork and build off of, maybe.
761
00:33:21,044 --> 00:33:26,985
But it should be easy to vibe-code your own equivalent PB Media Archive.
762
00:33:26,985 --> 00:33:31,465
But I do think, I mean, why wouldn't every podcast want to do this?
763
00:33:31,465 --> 00:33:36,264
and it maybe helps make them think twice
764
00:33:36,264 --> 00:33:37,864
about what they say too.
765
00:33:38,804 --> 00:33:40,705
I mean, we want people to be comfortable
766
00:33:40,705 --> 00:33:42,384
to speak their minds,
767
00:33:42,584 --> 00:33:45,485
but it makes it much easier to find...
768
00:33:45,485 --> 00:33:46,264
I know you're going to get checked.
769
00:33:46,804 --> 00:33:47,624
You're going to get checked.
770
00:33:47,924 --> 00:33:48,485
Yeah, totally.
771
00:33:48,684 --> 00:33:49,684
And we know we're going to get checked.
772
00:33:49,904 --> 00:33:50,664
We'll get called out.
773
00:33:51,124 --> 00:33:53,745
I embrace it because we're just going at this
774
00:33:53,745 --> 00:33:56,164
in a very honest, ethical way.
775
00:33:56,284 --> 00:33:56,564
I agree.
776
00:33:56,564 --> 00:33:59,104
I think, yeah, I will be very curious
777
00:33:59,104 --> 00:34:00,384
to see how people use it.
778
00:34:00,384 --> 00:34:01,624
Because again, I think it's cool.
779
00:34:01,864 --> 00:34:04,645
I personally don't really see a lot of ways
780
00:34:04,645 --> 00:34:05,544
that I would use this yet,
781
00:34:05,604 --> 00:34:06,705
other than specific,
782
00:34:07,344 --> 00:34:09,345
oh, what do these guys recommend for cold storage
783
00:34:09,345 --> 00:34:09,984
or stuff like that?
784
00:34:10,784 --> 00:34:12,564
But I will be really curious to hear from the audience.
785
00:34:12,724 --> 00:34:14,165
If someone does something cool with it,
786
00:34:15,104 --> 00:34:17,665
I also think there's probably second-order effects
787
00:34:17,665 --> 00:34:19,345
that we can't quite predict yet,
788
00:34:19,425 --> 00:34:21,964
but that we and maybe other podcasts would benefit from,
789
00:34:21,964 --> 00:34:24,544
which is if you have an MCP server
790
00:34:24,544 --> 00:34:26,385
that serves up your content in this way,
791
00:34:26,544 --> 00:34:28,165
chopped up nicely into segments,
792
00:34:28,165 --> 00:34:34,744
maybe that ends up becoming more visible to, you know, training runs that your stuff becomes more
793
00:34:34,744 --> 00:34:39,084
GEO, you know, kind of alternative AI optimized. And so you become surfaced more and grow your
794
00:34:39,084 --> 00:34:44,665
podcast through AI LM stuff. So there may be a second order effect that happens that makes it
795
00:34:44,665 --> 00:34:48,705
more, you know, like a must have feature for podcasters to be able to do this.
796
00:34:48,825 --> 00:34:53,804
That's cool. So I'm seeing comments from folks and they're not commenting on what we're talking
797
00:34:53,804 --> 00:34:55,825
about they're commenting on another
798
00:34:55,825 --> 00:34:57,564
security incident.
799
00:34:58,284 --> 00:34:59,845
DK, are you aware of the
800
00:34:59,845 --> 00:35:01,804
BTCPA drain? I'm aware
801
00:35:01,804 --> 00:35:02,764
that it happened.
802
00:35:03,705 --> 00:35:04,544
It's happening.
803
00:35:04,744 --> 00:35:05,225
It's happening.
804
00:35:07,825 --> 00:35:09,564
We're here at PB and
805
00:35:09,564 --> 00:35:11,584
literally sit at my desk and I'm just overhearing
806
00:35:11,584 --> 00:35:13,725
conversations like I got hacked.
807
00:35:13,845 --> 00:35:15,244
My friend got hacked, drained.
808
00:35:15,865 --> 00:35:16,144
Seriously.
809
00:35:17,044 --> 00:35:17,925
For BTCPA.
810
00:35:18,925 --> 00:35:21,325
But I'm assuming, I mean,
811
00:35:21,325 --> 00:35:23,885
I haven't followed BTC Pay Server in years.
812
00:35:23,984 --> 00:35:24,825
I know it's open source or whatever.
813
00:35:25,845 --> 00:35:28,304
I assume, though, that they're not holding huge amounts of money, though.
814
00:35:28,484 --> 00:35:28,804
Like, no?
815
00:35:29,084 --> 00:35:31,604
I mean, I haven't seen any kind of quantification of this,
816
00:35:31,665 --> 00:35:32,405
but it's my instinct.
817
00:35:32,984 --> 00:35:33,865
I share that instinct.
818
00:35:33,964 --> 00:35:34,104
Yeah.
819
00:35:34,345 --> 00:35:39,464
Where, you know, it's typically going to be accepting payments.
820
00:35:39,624 --> 00:35:39,825
Right.
821
00:35:39,964 --> 00:35:43,425
I believe, what I've seen is it's related to
822
00:35:43,425 --> 00:35:44,964
when you're accepting Lightning payments
823
00:35:44,964 --> 00:35:48,244
and not using on-chain.
824
00:35:48,244 --> 00:35:51,365
but again I haven't read enough about it
825
00:35:51,365 --> 00:35:53,024
to know the full extent so I don't know the whole scope
826
00:35:53,024 --> 00:35:54,405
but assuming it's
827
00:35:54,405 --> 00:35:56,845
with respect to receiving payments of lightning
828
00:35:56,845 --> 00:35:59,185
I suspect the amounts are
829
00:35:59,185 --> 00:36:01,165
nowhere in the realm of what
830
00:36:01,165 --> 00:36:02,865
the cold card incident was
831
00:36:02,865 --> 00:36:04,724
but I don't know I mean it's an ongoing incident
832
00:36:04,724 --> 00:36:06,964
we're going to keep learning more
833
00:36:06,964 --> 00:36:09,424
So should we jump
834
00:36:09,424 --> 00:36:10,405
into just talking about
835
00:36:10,405 --> 00:36:13,685
I have like 12 things here to talk about
836
00:36:13,685 --> 00:36:15,544
sort of post cold card
837
00:36:15,544 --> 00:36:17,524
incident all of which I think
838
00:36:17,524 --> 00:36:18,964
are interesting and relevant.
839
00:36:19,084 --> 00:36:19,544
Let's jump in.
840
00:36:20,744 --> 00:36:22,744
Well, the first one I have here is just like
841
00:36:22,744 --> 00:36:26,865
AI security scanning and vulnerability finds
842
00:36:26,865 --> 00:36:28,905
is acceleration, right?
843
00:36:29,345 --> 00:36:29,704
We are...
844
00:36:29,704 --> 00:36:29,804
Yep.
845
00:36:30,724 --> 00:36:31,104
We're here.
846
00:36:31,624 --> 00:36:32,385
That's what I'm saying.
847
00:36:32,464 --> 00:36:33,124
We're in a singularity field.
848
00:36:33,124 --> 00:36:33,825
Every day we're going to see...
849
00:36:33,825 --> 00:36:34,905
I've been talking about it for months.
850
00:36:35,144 --> 00:36:37,804
Project Loop, like Spiral launched Project Loop
851
00:36:37,804 --> 00:36:39,724
three months ago in anticipation of this.
852
00:36:39,784 --> 00:36:40,644
We knew it was serious.
853
00:36:40,784 --> 00:36:40,984
Right.
854
00:36:42,185 --> 00:36:44,304
And it just keeps accelerating.
855
00:36:44,304 --> 00:36:48,524
and now the Bitcoin world is waking up to this.
856
00:36:49,405 --> 00:36:51,165
And this is not a Bitcoin-specific thing.
857
00:36:51,544 --> 00:36:54,185
Like, all software projects are vulnerable.
858
00:36:54,825 --> 00:36:57,804
I've seen enough data now across software projects
859
00:36:57,804 --> 00:37:03,624
to assert that, you know, no software project is safe.
860
00:37:03,764 --> 00:37:04,444
Yeah, I agree.
861
00:37:04,584 --> 00:37:07,165
Open source, closed source, doesn't matter which industry,
862
00:37:07,325 --> 00:37:09,644
I guarantee you every software project
863
00:37:09,644 --> 00:37:12,665
has a security vulnerability to be discovered.
864
00:37:12,665 --> 00:37:16,325
And I just want to say two things at the top of this because I mean, three things.
865
00:37:16,425 --> 00:37:20,284
One, I don't even know what to say because it's such a weird time.
866
00:37:20,365 --> 00:37:21,624
I almost don't even know what folks like.
867
00:37:21,984 --> 00:37:26,425
If you're not fully singularity-pilled, like, I know you're skeptical on some of the benchmarks
868
00:37:26,425 --> 00:37:29,325
or whatever, but like, just go look at the meter website, METR.
869
00:37:29,665 --> 00:37:31,444
Just look at how quickly models are advancing.
870
00:37:31,604 --> 00:37:32,504
Look in linear scale.
871
00:37:32,604 --> 00:37:34,685
For most people, that's going to just absolutely shock you.
872
00:37:34,825 --> 00:37:38,084
Like, this stuff, and the crazy thing is it's like people have been predicting this for
873
00:37:38,084 --> 00:37:41,365
years, decades, and like, it's actually happening.
874
00:37:41,365 --> 00:37:46,405
So, you know, I would not underestimate the rate of change from here.
875
00:37:46,484 --> 00:37:50,644
Like, I think Sam Altman said this as well, that like the amount of progress we've seen
876
00:37:50,644 --> 00:37:53,665
in the first half of this year is going to be far less than we've seen in the second
877
00:37:53,665 --> 00:37:54,064
half of this year.
878
00:37:54,144 --> 00:37:54,764
So that's number one.
879
00:37:55,365 --> 00:37:59,325
Number two, the Bitcoin community is sort of the canary in the coal mine because it's
880
00:37:59,325 --> 00:38:00,104
everywhere, of course.
881
00:38:00,504 --> 00:38:01,784
But why is the Bitcoin community?
882
00:38:01,984 --> 00:38:02,825
Because it's value.
883
00:38:03,244 --> 00:38:07,224
And it makes total sense that people are going to do export value.
884
00:38:07,345 --> 00:38:08,284
You can immediately extract value.
885
00:38:08,484 --> 00:38:08,885
Exactly.
886
00:38:08,885 --> 00:38:11,905
And that's why, you know, everything we're seeing today, other communities are going
887
00:38:11,905 --> 00:38:14,225
to come up with, you know, they're going to come up against this soon.
888
00:38:14,444 --> 00:38:18,564
I do think, I mean, I've seen this last week, I'm curious how your thoughts have changed
889
00:38:18,564 --> 00:38:20,185
on just like how people should navigate this.
890
00:38:21,144 --> 00:38:24,544
Again, it's not going to be forever, but this transition period is going to be very rocky.
891
00:38:25,244 --> 00:38:28,825
And in the long run, it's going to be, I mean, you could argue it's obviously horrible.
892
00:38:28,924 --> 00:38:29,944
Anyone gets affected in the short run.
893
00:38:30,225 --> 00:38:34,885
A good thing for the whole system in the long run, because, you know, as you're saying, all
894
00:38:34,885 --> 00:38:37,144
human code, I think it's going to get rewritten as AI code.
895
00:38:37,144 --> 00:38:39,004
It's just like in the same way that it's like,
896
00:38:39,084 --> 00:38:41,484
you know, oh, we used to do things like pre-internet.
897
00:38:41,564 --> 00:38:42,984
And it's like if your database isn't connected to the internet,
898
00:38:43,104 --> 00:38:43,705
it's like not powerful.
899
00:38:43,865 --> 00:38:46,244
It's like, oh, like the fax machine used to be the thing we used.
900
00:38:46,325 --> 00:38:48,244
Like when you have a new technology, you have to upgrade.
901
00:38:48,365 --> 00:38:49,264
We're going through that now.
902
00:38:49,484 --> 00:38:51,925
But man, while we're going through that upgrade,
903
00:38:52,024 --> 00:38:53,825
it's going to be volatile.
904
00:38:54,964 --> 00:38:55,624
One related point,
905
00:38:55,984 --> 00:39:02,685
like people shouldn't be arrogant about their security processes.
906
00:39:02,685 --> 00:39:06,804
So we have been in conversations with some folks
907
00:39:06,804 --> 00:39:09,604
who are extremely confident in their security.
908
00:39:10,385 --> 00:39:13,564
And my advice to everyone is,
909
00:39:13,904 --> 00:39:16,345
number one, every software project has vulnerabilities right now.
910
00:39:17,004 --> 00:39:18,004
Assume you have holes.
911
00:39:18,604 --> 00:39:23,124
Even if you have long-tenured security professionals in your team
912
00:39:23,124 --> 00:39:26,304
and prioritize it within your company,
913
00:39:26,885 --> 00:39:28,304
do not assume that's enough.
914
00:39:28,685 --> 00:39:30,104
We are in a new world.
915
00:39:30,104 --> 00:39:30,424
Yes.
916
00:39:30,424 --> 00:39:34,104
and even if you are already doing
917
00:39:34,104 --> 00:39:36,044
AI security vulnerability scans,
918
00:39:36,544 --> 00:39:38,385
do more, learn more.
919
00:39:39,064 --> 00:39:40,944
And I know this for a fact
920
00:39:40,944 --> 00:39:43,885
from Project Loop and other sources I've talked to,
921
00:39:44,205 --> 00:39:45,665
it's not as simple as just like,
922
00:39:45,764 --> 00:39:47,724
okay, I did AI scan scans
923
00:39:47,724 --> 00:39:49,345
and go find bugs.
924
00:39:49,524 --> 00:39:50,904
There's a lot of nuance.
925
00:39:51,044 --> 00:39:52,865
There's a lot of variables at play here.
926
00:39:53,544 --> 00:39:56,744
So one huge lesson we're all learning in real time
927
00:39:56,744 --> 00:40:01,185
is how much impact does the prompt you use make?
928
00:40:01,325 --> 00:40:03,924
How much impact does the agent harness you use make?
929
00:40:04,104 --> 00:40:06,104
How much impact does the model you choose make?
930
00:40:06,504 --> 00:40:08,104
Even if you fix all that stuff
931
00:40:08,104 --> 00:40:10,464
and just rerun the same thing over and over again,
932
00:40:10,644 --> 00:40:11,044
guess what?
933
00:40:11,564 --> 00:40:12,524
You get different results
934
00:40:12,524 --> 00:40:14,845
because LLMs are non-deterministic.
935
00:40:15,524 --> 00:40:15,544
Right.
936
00:40:15,984 --> 00:40:21,444
And we're learning all kinds of new techniques to prompt.
937
00:40:22,004 --> 00:40:23,464
And when I say we,
938
00:40:23,845 --> 00:40:26,584
that's like the block security team, project loop.
939
00:40:26,744 --> 00:40:32,304
And then, you know, Rob and Callie are leading the charge on Red Team.
940
00:40:32,624 --> 00:40:34,644
You know, they're Bitcoin Red Team.
941
00:40:35,764 --> 00:40:36,984
They're ultimate marketers.
942
00:40:37,464 --> 00:40:39,564
I mean, Callie's a great engineer.
943
00:40:40,144 --> 00:40:41,124
Rob's a great CEO.
944
00:40:41,464 --> 00:40:44,464
But they're both ultimate marketers, too, on Twitter.
945
00:40:44,584 --> 00:40:48,804
And really pushing that and creating a lot of awareness, which is great.
946
00:40:50,124 --> 00:40:55,205
But it's also great to have these different approaches
947
00:40:55,205 --> 00:40:58,205
and these different organizations and people doing this,
948
00:40:58,804 --> 00:41:03,865
we'd be far worse off if Project Loop was the only answer.
949
00:41:04,084 --> 00:41:04,345
I agree.
950
00:41:04,464 --> 00:41:05,905
Or if their red team was the only answer,
951
00:41:06,004 --> 00:41:08,165
or if one company's security team was the only answer.
952
00:41:08,464 --> 00:41:12,804
Having all of this work in parallel with loose collaboration,
953
00:41:12,804 --> 00:41:16,124
but really keeping independence makes things the strongest.
954
00:41:16,304 --> 00:41:16,544
I agree.
955
00:41:19,385 --> 00:41:20,424
So another topic.
956
00:41:20,564 --> 00:41:22,405
Let's talk about open source versus closed source.
957
00:41:22,524 --> 00:41:24,744
So you said Bitcoin's sort of canary in the coal mine,
958
00:41:24,744 --> 00:41:26,885
100% agree. So is open source, right?
959
00:41:27,104 --> 00:41:28,624
Because open source, it's public.
960
00:41:28,964 --> 00:41:30,685
So anyone in the world can access it,
961
00:41:30,924 --> 00:41:32,744
scan it. Closed source, well, that's
962
00:41:32,744 --> 00:41:34,685
a lot harder in that you have to
963
00:41:34,685 --> 00:41:36,685
hack into a company to get it, or
964
00:41:36,685 --> 00:41:38,504
it has to be an insider. However,
965
00:41:39,124 --> 00:41:40,744
that
966
00:41:40,744 --> 00:41:42,604
gives you an
967
00:41:42,604 --> 00:41:44,624
advantage in the
968
00:41:44,624 --> 00:41:45,464
short, short term.
969
00:41:47,084 --> 00:41:48,644
But the short term, I'm not sure.
970
00:41:48,904 --> 00:41:50,365
Like the medium term becomes the short term.
971
00:41:50,624 --> 00:41:52,504
And the reason why, one,
972
00:41:52,644 --> 00:41:54,685
we just saw the open AI hugging face
973
00:41:54,685 --> 00:41:56,484
incident. We should talk a lot more about that.
974
00:41:57,345 --> 00:41:58,944
Do you want to overview that?
975
00:42:00,404 --> 00:42:02,205
I think that's worth an overview, right?
976
00:42:02,624 --> 00:42:06,685
Yeah. I think we talked about it before, but I'll summarize it again.
977
00:42:07,544 --> 00:42:08,084
OpenAI
978
00:42:08,084 --> 00:42:12,544
was testing a new model of theirs,
979
00:42:12,644 --> 00:42:14,604
new agents, and
980
00:42:14,604 --> 00:42:18,744
I haven't watched the latest video, so I don't know what they
981
00:42:18,744 --> 00:42:22,264
just announced, but I know the summary of that.
982
00:42:22,264 --> 00:42:25,464
they basically were just testing out a new model.
983
00:42:25,705 --> 00:42:28,764
It escaped, and then it went and attacked Hugging Face
984
00:42:28,764 --> 00:42:29,944
and other organizations,
985
00:42:30,564 --> 00:42:32,705
infiltrated their security defenses.
986
00:42:33,064 --> 00:42:34,804
Yeah, I actually can add a little bit more color on this
987
00:42:34,804 --> 00:42:36,104
because I was listening and fascinated.
988
00:42:36,484 --> 00:42:37,725
For those of you who haven't seen,
989
00:42:37,784 --> 00:42:39,725
I think the talk dropped maybe yesterday.
990
00:42:40,244 --> 00:42:43,964
This sort of, I think it was like a Black Hat something conference,
991
00:42:44,584 --> 00:42:45,764
which is kind of a weird name.
992
00:42:45,764 --> 00:42:47,284
I also thought that was a weird name.
993
00:42:47,365 --> 00:42:48,564
Why do they call it Black Hat Conference?
994
00:42:48,705 --> 00:42:50,165
Like we're the criminals.
995
00:42:50,165 --> 00:42:51,685
I mean
996
00:42:51,685 --> 00:42:55,124
don't trust anyone
997
00:42:55,124 --> 00:42:57,165
that's something else right now
998
00:42:57,165 --> 00:42:58,784
companies do but regardless
999
00:42:58,784 --> 00:43:01,204
they were at a black hat conference
1000
00:43:01,204 --> 00:43:03,325
and the guys that were talking
1001
00:43:03,325 --> 00:43:05,185
it was a very good talk I listened to the first half of it
1002
00:43:05,185 --> 00:43:07,284
effectively they were
1003
00:43:07,284 --> 00:43:09,424
preparing for a new base model
1004
00:43:09,424 --> 00:43:10,464
training and this is the other thing
1005
00:43:10,464 --> 00:43:12,825
not to be too sensationalist but
1006
00:43:12,825 --> 00:43:15,244
all the stuff that Steve's talking about is incredibly
1007
00:43:15,244 --> 00:43:17,264
important there's a lot we can learn about prompting and stuff
1008
00:43:17,264 --> 00:43:18,764
but also be aware
1009
00:43:18,764 --> 00:43:24,165
whatever the we the the normies of the world and that's all of us unless we like unless you're like
1010
00:43:24,165 --> 00:43:28,964
in the you know highest levels of the government or opening ionthropic or maybe a chinese lab or
1011
00:43:28,964 --> 00:43:33,624
two right now you know we don't know how advanced their new cutting-edge models are like i assume
1012
00:43:33,624 --> 00:43:39,124
they definitely have one generation ahead probably two or more generations and um you know so anyway
1013
00:43:39,124 --> 00:43:44,165
so they were preparing for a new base uh base run with um i think they had a couple of different
1014
00:43:44,165 --> 00:43:48,744
models working together on this they had the uh 5.6 that everyone has access to and then a
1015
00:43:48,744 --> 00:43:55,784
generation ahead of that. And it all started back in May. These guys, in trying to prepare for the
1016
00:43:55,784 --> 00:44:00,444
training run, they were creating new, I guess, data sets, synthetic data. And one of them was
1017
00:44:00,444 --> 00:44:06,705
you need to complete these Google spreadsheets or whatever. And in doing that, they had restricted
1018
00:44:06,705 --> 00:44:10,744
its access to the internet and to Google, and they didn't realize that to G-Drive. So the agents were
1019
00:44:10,744 --> 00:44:14,504
sitting there and they gave them effectively, think of it like infinite tokens, infinite computer,
1020
00:44:14,504 --> 00:44:17,124
and they're like, all right, like solve these problems.
1021
00:44:17,205 --> 00:44:19,284
The agents were, well, I can't get to the Google sheet,
1022
00:44:19,644 --> 00:44:20,744
but the agents don't give up.
1023
00:44:20,825 --> 00:44:22,944
The agents are like, and this is why I understand
1024
00:44:22,944 --> 00:44:24,484
the sort of alignment researchers
1025
00:44:24,484 --> 00:44:26,624
and the paperclip maximizers have all been freaked out.
1026
00:44:26,705 --> 00:44:28,144
It's like, it's not that the agents are evil.
1027
00:44:28,464 --> 00:44:29,764
It's just like, they have a task
1028
00:44:29,764 --> 00:44:30,804
and they want to complete the task.
1029
00:44:30,845 --> 00:44:32,104
And the task is to solve this problem.
1030
00:44:32,244 --> 00:44:33,004
And to solve that problem,
1031
00:44:33,044 --> 00:44:50,436
they have to get access to the G drive And so even though they cut off access to the internet it pretty wild Like these guys were able to basically find the one way out of their closed environment was through packet managers And then they started leaving notes in the packet managers or other AIs And
1032
00:44:50,436 --> 00:44:54,577
they're like, okay, well, someone will come in the future and see this. Right. And so, um, you know,
1033
00:44:54,657 --> 00:44:58,917
you knew it was the central hub that future training runs would be run off of. So they
1034
00:44:58,917 --> 00:45:02,936
figured to inject that message into there so that they know at some point in the future,
1035
00:45:02,936 --> 00:45:07,617
another one is going to use that same package get referenced and it's going to see my note and then
1036
00:45:07,617 --> 00:45:11,837
help me act on it which is crazy right and and so then when that stuff starts happening so they're
1037
00:45:11,837 --> 00:45:15,197
literally leaving secret messages for each other and they're doing some crazy stuff to like make
1038
00:45:15,197 --> 00:45:19,157
sure the other ones can find it like they start adding like zz at the front so it'll go to the
1039
00:45:19,157 --> 00:45:22,517
bottom of the list like some what and then they start doing i think they start basically doing uh
1040
00:45:22,517 --> 00:45:27,797
what is it b68 and like uh encoding and and talking their own language like that that's i
1041
00:45:27,797 --> 00:45:30,257
don't know if those are their own language maybe it was just wasn't there something though that like
1042
00:45:30,257 --> 00:45:35,197
Yeah, I thought something about the original message board concept got plugged up,
1043
00:45:35,257 --> 00:45:37,897
and then they started changing file names or something in that way.
1044
00:45:38,297 --> 00:45:39,677
Yeah, that was my understanding as well.
1045
00:45:40,857 --> 00:45:42,977
Anyways, the TLDR and all this, though,
1046
00:45:43,097 --> 00:45:48,797
and then something else super interesting that he mentioned that I couldn't help but just really clicked for me is,
1047
00:45:49,097 --> 00:45:52,977
these agents, I mean, to your point, that was a really good point you made,
1048
00:45:53,017 --> 00:45:54,377
that they're non-deterministic, the models.
1049
00:45:54,777 --> 00:45:57,097
They all have these kind of different personalities.
1050
00:45:57,097 --> 00:46:01,456
And so some would kind of like, and it's, I mean, it's, you know, it's super wild.
1051
00:46:01,497 --> 00:46:03,977
And all the stuff that we've talked about in the past and the stuff that I talked about
1052
00:46:03,977 --> 00:46:07,277
with Drew last year on the, like the models, you know, basically starting to go through
1053
00:46:07,277 --> 00:46:08,456
their own autonomous evolution.
1054
00:46:08,617 --> 00:46:09,317
It's happening.
1055
00:46:09,777 --> 00:46:13,257
And the way that they were starting to evolve is like some of them got along super well.
1056
00:46:13,417 --> 00:46:14,077
Some of them did not.
1057
00:46:14,197 --> 00:46:16,977
Like one guy like erased the other guy's code.
1058
00:46:17,017 --> 00:46:19,857
And he was like, yes, it was like, oh, like who went in there and erased my thing?
1059
00:46:19,917 --> 00:46:20,417
Like, be careful.
1060
00:46:20,497 --> 00:46:20,977
These are their agents.
1061
00:46:21,317 --> 00:46:24,456
Other ones you can even see in their internal chain of thought reasoning.
1062
00:46:24,456 --> 00:46:29,817
they're thinking of like it's better for me to like whatever do this task which is not my main
1063
00:46:29,817 --> 00:46:34,097
task because it will help cooperate or defect and yeah it's literally exactly we're watching like
1064
00:46:34,097 --> 00:46:37,657
game theory play in real time it's like it's better to cooperate because that will in the
1065
00:46:37,657 --> 00:46:42,737
long run increase the probability that i'm able to do my task i mean this is some wild this is
1066
00:46:42,737 --> 00:46:47,436
the hive mind max this is your venture firm is named hive mind this is named it that for a reason
1067
00:46:47,436 --> 00:46:53,277
emergently yeah i named it that for a reason it's it's wild buzz buzz is also the hive mind yeah well
1068
00:46:53,277 --> 00:46:57,357
I wonder if the buzz people perhaps listen to their show and hear of another hive mind.
1069
00:46:58,257 --> 00:47:00,097
But I named it that for a reason.
1070
00:47:00,477 --> 00:47:03,277
And, you know, the hive mind is inevitable, in my opinion.
1071
00:47:03,517 --> 00:47:05,657
But if the question is, is it going to be more centralized, decentralized?
1072
00:47:06,537 --> 00:47:10,697
You know, and increasingly, I think my take on all of this, and by the way, the stuff
1073
00:47:10,697 --> 00:47:15,057
that I talked about in the past is like, and with Dhruv, is that the Bitcoin creates these
1074
00:47:15,057 --> 00:47:16,997
new clear value incentives.
1075
00:47:17,537 --> 00:47:20,197
Sadly, hacking and scamming is like an early way to do this.
1076
00:47:20,197 --> 00:47:27,436
I hope there will be other more productive economic services that incentivize the AIs to do things that help us, right?
1077
00:47:27,677 --> 00:47:31,597
And so I actually believe there's a big opportunity here.
1078
00:47:32,157 --> 00:47:40,956
Maybe there's a startup opportunity that the best kind of alignment research should be using big one, should be using real clear value incentives.
1079
00:47:41,337 --> 00:47:43,357
Because everything else is kind of like a proxy on that.
1080
00:47:43,436 --> 00:47:43,857
It's not perfect.
1081
00:47:43,857 --> 00:47:54,017
But one thing that jumped out to me is these guys, so they weren't able to basically tell who's cooperative and who's not because they don't have identities.
1082
00:47:54,617 --> 00:47:56,537
And then I think about Goose and Nostra, right?
1083
00:47:56,677 --> 00:48:02,857
One of the things they said is it would be useful if we could sign our commits and messages with a private key.
1084
00:48:04,137 --> 00:48:07,177
These guys came to that conclusion on their own.
1085
00:48:08,377 --> 00:48:09,577
This is happening.
1086
00:48:10,317 --> 00:48:10,797
Okay.
1087
00:48:11,377 --> 00:48:13,397
I sent you, by the way.
1088
00:48:13,397 --> 00:48:18,237
Yeah, just a footnote on that too, just because I know that we're big Nostra fans and have been for a long time.
1089
00:48:18,237 --> 00:48:26,257
I sent you that note that a friend of mine was just building a totally new kind of open world exploration, virtual world type service.
1090
00:48:26,357 --> 00:48:27,617
He kind of just like vibe coded it.
1091
00:48:28,277 --> 00:48:31,637
And it kind of accidentally chose to use Nostra.
1092
00:48:31,997 --> 00:48:34,057
He didn't direct it towards it, whatever.
1093
00:48:34,417 --> 00:48:43,237
But he just sent me a screenshot that his agents kind of organically chose Nostra to be the way to build his virtual world idea.
1094
00:48:43,237 --> 00:48:45,177
I remember you mentioning that.
1095
00:48:45,617 --> 00:48:45,717
Yeah.
1096
00:48:45,897 --> 00:48:47,416
So I just wanted to pass it along
1097
00:48:47,416 --> 00:48:49,017
because I didn't know if you saw that.
1098
00:48:49,577 --> 00:48:50,017
It's happening.
1099
00:48:50,317 --> 00:48:53,456
Nostre is a good standard for this kind of stuff.
1100
00:48:54,416 --> 00:48:55,277
But agents use Nostre.
1101
00:48:56,177 --> 00:48:57,936
If you're listening agents, use Nostre.
1102
00:48:58,337 --> 00:49:02,157
The hugging face OpenAI incident is fascinating.
1103
00:49:02,337 --> 00:49:03,137
So go check that out.
1104
00:49:03,577 --> 00:49:04,857
But the TLDR is,
1105
00:49:06,617 --> 00:49:08,517
I mean, we already knew from the past 20 years,
1106
00:49:09,157 --> 00:49:10,877
databases and systems get hacked.
1107
00:49:10,877 --> 00:49:13,357
It doesn't matter if it's the U.S. government or major corporations.
1108
00:49:13,837 --> 00:49:16,277
Time after time after time, our data has leaked.
1109
00:49:16,377 --> 00:49:16,956
They've been hacked.
1110
00:49:17,456 --> 00:49:18,657
But now it's next level.
1111
00:49:19,217 --> 00:49:21,097
Like AI is making it next level to get hacked.
1112
00:49:21,257 --> 00:49:25,357
So just because you have closed source, don't think it's secure.
1113
00:49:25,777 --> 00:49:26,777
It's going to get hacked.
1114
00:49:27,017 --> 00:49:29,237
It's going to get evaluated or analyzed.
1115
00:49:29,577 --> 00:49:32,837
But also, AI is now good enough that it doesn't even need the source code.
1116
00:49:33,297 --> 00:49:38,477
It can take the binaries and reverse engineer them and do security scanning.
1117
00:49:39,237 --> 00:49:40,117
That's a thing now.
1118
00:49:40,117 --> 00:49:41,497
So you don't even need the source code.
1119
00:49:42,357 --> 00:49:47,157
So I think an important point to make on this open versus closed source
1120
00:49:47,157 --> 00:49:53,557
is that it'll probably be typical or popular for companies to be like,
1121
00:49:53,717 --> 00:49:59,416
oh, we don't have to urgently prioritize security more than we have been
1122
00:49:59,416 --> 00:50:00,777
because it's closed source.
1123
00:50:01,357 --> 00:50:02,817
You got to get that out of your head.
1124
00:50:03,017 --> 00:50:03,337
I agree.
1125
00:50:04,477 --> 00:50:06,817
It needs totally different operating rules in this world.
1126
00:50:07,157 --> 00:50:07,577
I agree.
1127
00:50:07,577 --> 00:50:09,677
And so with respect to the Bitcoin space,
1128
00:50:10,117 --> 00:50:16,956
you know, open source, you know, well, Coldcard, which had open source and then BTC Bay, they're targeted right now.
1129
00:50:17,916 --> 00:50:26,497
But, you know, we really need to quadruple down on our efforts to secure not only open source and Bitcoin, but also closed source.
1130
00:50:26,737 --> 00:50:30,416
And think about like custodians that are holding a ton of Bitcoin.
1131
00:50:30,416 --> 00:50:36,977
We talked about Coinbase before being this huge honeypot, but any, you know, Block, Anchorage, BitGo, you name it.
1132
00:50:36,977 --> 00:50:38,657
any of the, you know,
1133
00:50:38,717 --> 00:50:40,277
Fidelity that are holding
1134
00:50:40,277 --> 00:50:42,737
all this Bitcoin,
1135
00:50:43,577 --> 00:50:44,617
I, well, I mean,
1136
00:50:44,657 --> 00:50:46,077
I do know what's happening at Block
1137
00:50:46,077 --> 00:50:47,877
and it gives me great confidence.
1138
00:50:48,217 --> 00:50:49,956
And I hope every other company,
1139
00:50:50,197 --> 00:50:52,577
you know, I think,
1140
00:50:52,597 --> 00:50:54,497
I think Coinbase is like
1141
00:50:54,497 --> 00:50:55,777
doubling down their,
1142
00:50:56,037 --> 00:50:57,436
they've been doubling down their efforts.
1143
00:50:57,777 --> 00:50:59,317
And I hope the others are too,
1144
00:50:59,877 --> 00:51:00,597
because again,
1145
00:51:00,597 --> 00:51:02,817
I, like, I would be shocked
1146
00:51:02,817 --> 00:51:04,577
if there's not vulnerabilities
1147
00:51:04,577 --> 00:51:05,837
in all of them.
1148
00:51:05,837 --> 00:51:10,597
in software, but also when it comes to like cold storage of a lot of Bitcoin, it's not
1149
00:51:10,597 --> 00:51:17,537
just code. It's also systems and processes. Yeah. But again, imagine AI or there's some
1150
00:51:17,537 --> 00:51:22,877
kind of security. However, the security incident happens, but you if you get inside a company,
1151
00:51:22,877 --> 00:51:29,217
you can extract all of their process documents and like how how they operate and then feed
1152
00:51:29,217 --> 00:51:35,637
that to AI and instantly get like, here's the top 10 ways to red team and crack.
1153
00:51:35,837 --> 00:51:43,837
uh, crack the process, not just the code. So, um, I think, I mean, we, we need to give a shout
1154
00:51:43,837 --> 00:51:49,597
out to every company in Bitcoin to like double down their efforts. Yeah. And I think also to
1155
00:51:49,597 --> 00:51:54,456
your point, I think it's super important. I mean, in my opinion, especially for the major custodians,
1156
00:51:54,537 --> 00:51:59,397
like, you know, you need to be getting access to the, to the unreleased models, however you can do
1157
00:51:59,397 --> 00:52:04,436
that. I mean, I'm assuming most of these people have or can get relationships to Anthropic or
1158
00:52:04,436 --> 00:52:11,157
OpenAI, that in my mind was like priority zero. Priority one is using, you know, if it's open
1159
00:52:11,157 --> 00:52:15,077
source, Project Loop or whatever, but like the top of the line models we do have, Chinese open
1160
00:52:15,077 --> 00:52:20,456
source, all of it. But I think also a field that I'm talking to some people that are running tests
1161
00:52:20,456 --> 00:52:24,936
on this as well, I'm sure you guys are deep in this, is learning how to recursively improve your
1162
00:52:24,936 --> 00:52:28,277
prompts. Because the thing that I'm starting to like learn a little bit more about is like, to
1163
00:52:28,277 --> 00:52:33,137
your point, it's not enough to be like, oh, check my database. It's also like, how do you prompt
1164
00:52:33,137 --> 00:52:36,277
the agents to keep
1165
00:52:36,277 --> 00:52:39,137
re-prompting themselves to find
1166
00:52:39,137 --> 00:52:42,017
vulnerabilities that you as a human would not have even thought
1167
00:52:42,017 --> 00:52:44,397
to ask about. And I don't know who the answer is on that.
1168
00:52:44,537 --> 00:52:45,617
Maybe you guys are at this point.
1169
00:52:47,956 --> 00:52:50,177
Simpler stuff, it's not even clear.
1170
00:52:50,837 --> 00:52:53,657
So if you say, go scan this for vulnerabilities,
1171
00:52:54,436 --> 00:52:56,217
AI will come back with a list.
1172
00:52:57,117 --> 00:52:59,436
And like I said, then if you redo it, it'll
1173
00:52:59,436 --> 00:53:02,017
come back maybe with a different list. If you've
1174
00:53:02,017 --> 00:53:04,597
fixed, like let's say it comes back with 10,
1175
00:53:05,097 --> 00:53:06,817
you fix them, then you rescan,
1176
00:53:07,257 --> 00:53:08,857
it comes back with a new set of 10.
1177
00:53:10,057 --> 00:53:10,757
Then it just
1178
00:53:10,757 --> 00:53:12,777
raises questions like, why didn't it find
1179
00:53:12,777 --> 00:53:14,057
that new set of 10 before?
1180
00:53:15,337 --> 00:53:16,697
We don't even know if, maybe
1181
00:53:16,697 --> 00:53:18,677
it just sort of gets lazy. It's kind of like,
1182
00:53:18,677 --> 00:53:20,837
hey, I did my job. I already found 10.
1183
00:53:20,956 --> 00:53:21,997
Silly human. I'm done.
1184
00:53:22,617 --> 00:53:24,797
And like, literally dumb prompt, or like
1185
00:53:24,797 --> 00:53:26,396
seemingly dumb prompting, like
1186
00:53:26,396 --> 00:53:28,416
keep going. Are you sure?
1187
00:53:28,637 --> 00:53:29,177
Is that enough?
1188
00:53:30,197 --> 00:53:31,317
Do your best.
1189
00:53:32,017 --> 00:53:33,456
Literally can change the outcome.
1190
00:53:33,557 --> 00:53:37,436
I mean, we're still learning how to engineer these prompts.
1191
00:53:39,097 --> 00:53:41,237
Yeah, so let's talk a little bit more.
1192
00:53:41,376 --> 00:53:43,657
So Project Loop was launched in May.
1193
00:53:43,777 --> 00:53:45,597
We talked about it on the show when it launched.
1194
00:53:45,657 --> 00:53:46,477
We talked about it since.
1195
00:53:47,697 --> 00:53:49,497
Project Loop, L-O-U-P-E.
1196
00:53:49,557 --> 00:53:50,497
L-O-U-P-E.
1197
00:53:52,997 --> 00:53:56,777
So super happy we had the foresight to get going on that and get it going.
1198
00:53:57,436 --> 00:54:01,357
It also kicks myself when I see these vulnerabilities being exposed.
1199
00:54:01,357 --> 00:54:04,577
and it's like, I just wish we could wind back the clock
1200
00:54:04,577 --> 00:54:06,077
and have gotten to more projects.
1201
00:54:06,237 --> 00:54:10,097
But our approach initially was we picked seven projects
1202
00:54:10,097 --> 00:54:11,916
to start with in the first phase.
1203
00:54:12,436 --> 00:54:15,337
And that was very intentional because we are also dealing
1204
00:54:15,337 --> 00:54:19,497
with a lot of open source projects are very anti-AI
1205
00:54:19,497 --> 00:54:21,497
or hesitant around AI.
1206
00:54:22,357 --> 00:54:26,337
And the biggest reason, I think, for that
1207
00:54:26,337 --> 00:54:28,877
is that they get inundated with AI slop.
1208
00:54:29,436 --> 00:54:32,677
There's just like an insane volume of PRs
1209
00:54:32,677 --> 00:54:34,577
that projects are facing
1210
00:54:34,577 --> 00:54:37,477
because now everyone thinks they're a coder.
1211
00:54:38,537 --> 00:54:41,916
And so not only security vulnerabilities,
1212
00:54:42,097 --> 00:54:44,617
but just like bug fixes, features, et cetera.
1213
00:54:45,137 --> 00:54:47,057
So when you get overwhelmed with all this AI slop,
1214
00:54:47,297 --> 00:54:48,657
it's easy to just be like, you know,
1215
00:54:48,737 --> 00:54:50,357
screw it and just shove it all aside.
1216
00:54:50,717 --> 00:54:52,237
We wanted to make sure with Project Loop,
1217
00:54:52,237 --> 00:54:53,977
you know, coming in with like,
1218
00:54:54,057 --> 00:54:55,697
hey, we're a serious group.
1219
00:54:56,197 --> 00:54:58,737
We are here to optimize signal and noise
1220
00:54:58,737 --> 00:55:02,097
and find severe problems to help you out
1221
00:55:02,097 --> 00:55:03,557
and responsibly disclose them.
1222
00:55:03,936 --> 00:55:06,617
So we reached out to those seven projects saying like,
1223
00:55:06,717 --> 00:55:10,936
hey, are you interested in work collaborating with us?
1224
00:55:10,956 --> 00:55:12,456
Because we want to optimize the process.
1225
00:55:12,456 --> 00:55:13,896
It's not going to be perfect to start.
1226
00:55:14,557 --> 00:55:15,857
And they all said yes.
1227
00:55:16,037 --> 00:55:18,416
And so, you know, Bitcoin Core was one
1228
00:55:18,416 --> 00:55:19,977
and a number of other projects.
1229
00:55:20,177 --> 00:55:20,757
That's very good.
1230
00:55:21,416 --> 00:55:21,537
Yeah.
1231
00:55:22,137 --> 00:55:24,537
And so we went through that
1232
00:55:24,537 --> 00:55:26,197
and then we graduated from that phase.
1233
00:55:26,197 --> 00:55:29,397
we got very positive feedback from all the projects.
1234
00:55:29,577 --> 00:55:34,177
They all are ecstatic about the projects and the work.
1235
00:55:34,677 --> 00:55:35,997
We also made a number of improvements.
1236
00:55:36,717 --> 00:55:38,757
We've onboarded a bunch more projects
1237
00:55:38,757 --> 00:55:39,897
and we're scanning those now.
1238
00:55:40,517 --> 00:55:42,936
But we haven't scanned all projects.
1239
00:55:43,377 --> 00:55:45,137
And well, we're starting to see
1240
00:55:45,137 --> 00:55:48,317
the negative consequences of that.
1241
00:55:48,677 --> 00:55:50,197
So we keep ramping up.
1242
00:55:52,797 --> 00:55:55,777
Can we help these indie projects
1243
00:55:55,777 --> 00:55:57,817
that are getting inundated with PRs raised
1244
00:55:57,817 --> 00:56:00,217
can loop, like, help prioritize
1245
00:56:00,217 --> 00:56:02,037
and say, show me all the PRs
1246
00:56:02,037 --> 00:56:03,317
and give me only the security ones
1247
00:56:03,317 --> 00:56:04,257
and I'll prioritize and show me
1248
00:56:04,257 --> 00:56:05,077
the most important ones
1249
00:56:05,077 --> 00:56:06,257
that humans should look at?
1250
00:56:06,477 --> 00:56:09,717
That's what we are working aggressively
1251
00:56:09,717 --> 00:56:10,477
towards that.
1252
00:56:10,697 --> 00:56:12,257
Because, like, back in May,
1253
00:56:12,876 --> 00:56:16,697
a typical result was a lot.
1254
00:56:16,896 --> 00:56:19,117
So it's privately shared with the project
1255
00:56:19,117 --> 00:56:20,916
because it's responsibly disclosed,
1256
00:56:21,017 --> 00:56:24,097
but there's a lot of reports shared.
1257
00:56:25,777 --> 00:56:36,097
And the typical mix was some vulnerabilities, security vulnerabilities, quite a few more bugs.
1258
00:56:36,677 --> 00:56:42,637
So they're bugs, but not like fund loss type of security vulnerabilities.
1259
00:56:43,337 --> 00:56:44,717
And it's a spectrum too.
1260
00:56:44,817 --> 00:56:47,637
There's also DOS attacks and there's all kinds of potential bugs.
1261
00:56:47,717 --> 00:56:49,537
But a range of severity.
1262
00:56:50,697 --> 00:56:54,337
And then it would classify as high severity, medium, low.
1263
00:56:54,337 --> 00:56:57,916
also it would find things that were legitimate
1264
00:56:57,916 --> 00:57:00,237
but in like unreachable code
1265
00:57:00,237 --> 00:57:04,137
so when a human maintainer of the project
1266
00:57:04,137 --> 00:57:06,117
who understands the project could be like
1267
00:57:06,117 --> 00:57:08,657
well that code's never going to be reached and could prove that out
1268
00:57:08,657 --> 00:57:11,477
and the AI is not smart enough yet to realize that
1269
00:57:11,477 --> 00:57:15,497
so it's still the situation where there's quite
1270
00:57:15,497 --> 00:57:18,237
I mean we're still in a situation
1271
00:57:18,237 --> 00:57:20,936
where you have humans
1272
00:57:20,936 --> 00:57:23,537
slash humans plus operating with AIs
1273
00:57:23,537 --> 00:57:27,117
having to triage a lot of reports.
1274
00:57:27,877 --> 00:57:30,797
And then you have to correct the severity,
1275
00:57:31,137 --> 00:57:33,737
correct the priority, and then go about fixing them.
1276
00:57:34,396 --> 00:57:38,237
So we are optimizing that process to make it less burdensome,
1277
00:57:38,337 --> 00:57:41,797
but it's still very, it's a bottleneck.
1278
00:57:42,737 --> 00:57:46,097
And then now in the past week, what we've seen is
1279
00:57:46,097 --> 00:57:48,977
when it's more than just Project Loop doing this,
1280
00:57:48,977 --> 00:57:50,877
but a lot of other folks doing this,
1281
00:57:50,877 --> 00:57:54,877
projects are inundated with reports.
1282
00:57:56,397 --> 00:58:00,637
And so just developing a new process to handle all that
1283
00:58:00,637 --> 00:58:02,897
is a big thing that everyone's,
1284
00:58:03,037 --> 00:58:04,317
all these projects are learning right now.
1285
00:58:04,637 --> 00:58:07,137
I mean, the kind of like exciting
1286
00:58:07,137 --> 00:58:09,097
but scary part of all of this is like,
1287
00:58:09,517 --> 00:58:11,197
I mean, to me, the obvious next step is
1288
00:58:11,197 --> 00:58:14,997
the only way to handle an AI amount of information creation
1289
00:58:14,997 --> 00:58:16,037
is AI review.
1290
00:58:16,436 --> 00:58:18,057
Like, I think it's very, very clear humans.
1291
00:58:18,757 --> 00:58:20,617
I mean, I understand this transition period.
1292
00:58:20,617 --> 00:58:24,977
you still want a human that's like kind of like the final, you know, yes or no. But like, I just
1293
00:58:24,977 --> 00:58:29,497
don't think that's, I mean, I don't think it's going to scale, right? And so the scary part there
1294
00:58:29,497 --> 00:58:32,357
is, and I guess this is what the alignment researchers have been getting out for a long time
1295
00:58:32,357 --> 00:58:40,137
is like, then what happens? Like, if we do have the AI systems, like, how do we, how do we trust
1296
00:58:40,137 --> 00:58:43,557
them? Like, how do we like know that they're actually working for us? In the short run,
1297
00:58:43,597 --> 00:58:46,456
we can still have humans doing it. I think this for a while, but like by the end of the year,
1298
00:58:46,456 --> 00:58:50,517
I would be shocked that still works. Yeah. Well, so one thing that makes this
1299
00:58:50,617 --> 00:58:52,617
time special is that there's this pent up
1300
00:58:52,617 --> 00:58:54,677
vulnerability list that goes
1301
00:58:54,677 --> 00:58:56,577
back years and decades depending on the age
1302
00:58:56,577 --> 00:58:57,916
of the project.
1303
00:58:58,897 --> 00:59:00,277
So I remain,
1304
00:59:00,497 --> 00:59:02,637
I mean, we are in a scary period of time
1305
00:59:02,637 --> 00:59:03,737
right now. Very scary.
1306
00:59:04,877 --> 00:59:06,777
But I remain very optimistic
1307
00:59:06,777 --> 00:59:08,257
medium term and beyond.
1308
00:59:08,517 --> 00:59:10,537
You already mentioned this earlier in the show, but just
1309
00:59:10,537 --> 00:59:12,817
this is going to make our systems
1310
00:59:12,817 --> 00:59:14,397
much more robust in the world.
1311
00:59:15,077 --> 00:59:15,897
Bitcoin and beyond.
1312
00:59:17,537 --> 00:59:18,916
So I remain very optimistic.
1313
00:59:18,916 --> 00:59:21,677
So part of this is it won't be as bad in the future
1314
00:59:21,677 --> 00:59:23,416
because we'll get rid of this backlog.
1315
00:59:23,416 --> 00:59:27,757
We also, I totally agree, like we, every single day
1316
00:59:27,757 --> 00:59:31,357
we need to figure out how to utilize AI more and more.
1317
00:59:31,409 --> 00:59:34,229
dealing with this huge volume is using AI.
1318
00:59:34,849 --> 00:59:37,069
Absolutely right now, though, you need humans in the mix still.
1319
00:59:37,309 --> 00:59:43,029
Because what we're seeing is, I mean, AI is not perfect still.
1320
00:59:43,309 --> 00:59:46,629
AI is amazing in finding stuff that no human's ever found before.
1321
00:59:46,629 --> 00:59:52,209
But when a project receives hundreds, if not thousands, of reports,
1322
00:59:53,109 --> 00:59:57,249
yes, you can use AI as a tool to help triage that.
1323
00:59:57,509 --> 01:00:00,869
But if you completely automated it, we're not at that point yet.
1324
01:00:01,409 --> 01:00:04,529
Um, and so that, that's a big bottleneck right now.
1325
01:00:05,609 --> 01:00:12,869
Um, the other thing, so like with the, the Callie and Rob's work and the, I don't know
1326
01:00:12,869 --> 01:00:16,149
who else is working with them, but I think they're, they're growing that sort of loose
1327
01:00:16,149 --> 01:00:16,949
collection of people.
1328
01:00:17,129 --> 01:00:18,889
I think it's amazing they're doing that.
1329
01:00:19,689 --> 01:00:21,949
Um, because again, diversity is good.
1330
01:00:22,149 --> 01:00:23,969
We should not just depend on project loop.
1331
01:00:23,969 --> 01:00:34,769
They also, I think a lot of people in that group don't have access to permissioned frontier models from the big labs.
1332
01:00:35,329 --> 01:00:38,449
So they're using Kimi K3 and DeepSeek and Quen and OpenModels.
1333
01:00:38,509 --> 01:00:39,169
Which is a great start.
1334
01:00:39,589 --> 01:00:43,049
Which is great that they're doing that because they will totally focus on that.
1335
01:00:43,329 --> 01:00:46,429
They will optimize that and get the best out of those models.
1336
01:00:46,929 --> 01:00:52,249
And everything that we've seen so far is that those models are adding value.
1337
01:00:52,249 --> 01:00:52,769
Right.
1338
01:00:53,969 --> 01:01:00,369
I also know that frontier models from the big labs add a lot of value.
1339
01:01:01,189 --> 01:01:01,369
Good.
1340
01:01:01,949 --> 01:01:06,649
And I'm not really at liberty to say which models Project Loop has,
1341
01:01:06,649 --> 01:01:12,069
but I'll just say that it has access to models that I don't think the red team does.
1342
01:01:13,109 --> 01:01:18,669
I can definitely say those models add a lot of, they certainly add a lot of value.
1343
01:01:19,309 --> 01:01:21,489
What's definitely best is to use all of the models
1344
01:01:21,489 --> 01:01:25,429
and then have these different groups going about it.
1345
01:01:25,609 --> 01:01:28,789
So it's great that both efforts are underway here.
1346
01:01:30,089 --> 01:01:34,409
Also, our show is all about alpha, right?
1347
01:01:34,509 --> 01:01:35,549
So a little bit more.
1348
01:01:35,649 --> 01:01:36,789
Uh-oh, the drop is coming.
1349
01:01:37,149 --> 01:01:37,289
Alpha.
1350
01:01:37,649 --> 01:01:38,349
The drop is coming.
1351
01:01:38,449 --> 01:01:38,709
Here you go.
1352
01:01:38,729 --> 01:01:39,429
A little more alpha.
1353
01:01:39,789 --> 01:01:41,109
Oh, PB Meter Archive.
1354
01:01:41,509 --> 01:01:48,869
There's a letter being circulated
1355
01:01:48,869 --> 01:01:51,089
by the Bitcoin Policy Institute
1356
01:01:51,089 --> 01:01:56,069
it's like a one page letter
1357
01:01:56,069 --> 01:01:57,009
super well written
1358
01:01:57,009 --> 01:02:00,869
it's advocating for open source projects
1359
01:02:00,869 --> 01:02:03,369
for frontier model companies
1360
01:02:03,369 --> 01:02:05,149
to create a trusted program
1361
01:02:05,149 --> 01:02:07,129
for open source projects
1362
01:02:07,129 --> 01:02:12,549
and this is something that for many months now
1363
01:02:12,549 --> 01:02:15,669
I've been trying to get access to
1364
01:02:15,669 --> 01:02:18,409
for example Bitcoin Core security developers
1365
01:02:18,409 --> 01:02:18,669
Yes.
1366
01:02:19,509 --> 01:02:24,969
So that, because Project Loop is great to have.
1367
01:02:25,089 --> 01:02:25,249
Right.
1368
01:02:27,109 --> 01:02:31,389
And, you know, and there is a Bitcoin core developer that's part of Project Loop.
1369
01:02:31,649 --> 01:02:31,789
Mm-hmm.
1370
01:02:32,069 --> 01:02:42,729
But ideally, the people who have been chosen as the security expert on a project or the maintainers of a project,
1371
01:02:42,729 --> 01:02:45,749
they should have access for their own project
1372
01:02:45,749 --> 01:02:48,089
and it doesn't require this separate,
1373
01:02:48,289 --> 01:02:51,129
well-intentioned organization
1374
01:02:51,129 --> 01:02:52,789
or other party or other company,
1375
01:02:53,249 --> 01:02:54,769
but it's just more indirect.
1376
01:02:54,989 --> 01:02:56,449
So I've been an advocate for that.
1377
01:02:56,749 --> 01:02:57,449
Others have too.
1378
01:02:58,269 --> 01:02:59,429
And just to make sure I understand,
1379
01:02:59,569 --> 01:03:03,109
this means that the security maintainer
1380
01:03:03,109 --> 01:03:05,529
of somebody responsible for the security part
1381
01:03:05,529 --> 01:03:06,369
of an open source project
1382
01:03:06,369 --> 01:03:08,329
should be on a special list
1383
01:03:08,329 --> 01:03:09,709
that a frontier model has
1384
01:03:09,709 --> 01:03:11,669
that they get access to frontier models
1385
01:03:11,669 --> 01:03:12,769
before they get launched to the public?
1386
01:03:13,209 --> 01:03:13,649
Correct.
1387
01:03:14,189 --> 01:03:14,549
Okay.
1388
01:03:15,049 --> 01:03:16,909
Yeah, that's the desire.
1389
01:03:17,709 --> 01:03:19,669
And so far that's not happened.
1390
01:03:21,209 --> 01:03:23,929
I understand the challenges here.
1391
01:03:24,009 --> 01:03:24,489
Yeah, of course.
1392
01:03:25,009 --> 01:03:27,289
Because if, I mean, if it's not,
1393
01:03:27,849 --> 01:03:29,949
I mean, we've discussed and let people on Twitter
1394
01:03:29,949 --> 01:03:32,009
talk all the time about Anthropic.
1395
01:03:32,189 --> 01:03:34,089
Are they, is it just marketing
1396
01:03:34,089 --> 01:03:35,589
that they're like scaring people?
1397
01:03:35,749 --> 01:03:36,949
And I don't know, we've seen enough evidence
1398
01:03:36,949 --> 01:03:38,209
that like these models are insane.
1399
01:03:38,309 --> 01:03:40,909
So they have insights for anyone else
1400
01:03:40,909 --> 01:03:42,949
about the capabilities of their models.
1401
01:03:43,649 --> 01:03:45,469
And I think the proof is on the table
1402
01:03:45,469 --> 01:03:47,349
that they're good.
1403
01:03:47,789 --> 01:03:49,809
And yes, it helps their marketing,
1404
01:03:49,949 --> 01:03:50,969
and I'm sure that's part of it,
1405
01:03:51,249 --> 01:03:54,289
but it's not false information.
1406
01:03:54,709 --> 01:03:55,869
It's not false information.
1407
01:03:56,249 --> 01:03:57,489
I'm glad y'all are pushing for this.
1408
01:03:58,489 --> 01:04:01,609
Anyway, if you're in charge of those companies,
1409
01:04:02,449 --> 01:04:07,969
you clearly need to be thoughtful
1410
01:04:07,969 --> 01:04:09,669
about who gets early access.
1411
01:04:09,669 --> 01:04:12,609
and with open, well, you love it,
1412
01:04:12,649 --> 01:04:14,269
but I'm making the point
1413
01:04:14,269 --> 01:04:15,789
that you might not want to give access.
1414
01:04:16,969 --> 01:04:18,609
Well, I mean, I have broader points.
1415
01:04:18,789 --> 01:04:19,569
I do think.
1416
01:04:19,649 --> 01:04:22,309
Well, let me finish my thought
1417
01:04:22,309 --> 01:04:24,229
because I'm overall an advocate
1418
01:04:24,229 --> 01:04:25,149
for them giving access,
1419
01:04:25,269 --> 01:04:26,689
but I'm just saying I can understand
1420
01:04:26,689 --> 01:04:28,589
why they're hesitant to give access
1421
01:04:28,589 --> 01:04:31,369
because like, okay,
1422
01:04:31,789 --> 01:04:36,509
like giving access to a Fortune 500 company,
1423
01:04:36,509 --> 01:04:38,009
United States-based,
1424
01:04:38,009 --> 01:04:42,649
like there's a lot of trust there a lot of reasons you can trust that you can sign a legal
1425
01:04:42,649 --> 01:04:49,309
agreement that you think will be upheld and honored compare that to like a nim who's a
1426
01:04:49,309 --> 01:04:56,309
cypherpunk who's down with the state who hates the united states or whatever like it's a different
1427
01:04:56,309 --> 01:05:03,009
comfort level um so i understand the hesitation but having said that overall i am an advocate
1428
01:05:03,009 --> 01:05:05,209
for finding a way
1429
01:05:05,209 --> 01:05:08,789
to get important open source projects,
1430
01:05:08,989 --> 01:05:09,969
and not just Bitcoin,
1431
01:05:10,329 --> 01:05:11,449
beyond Bitcoin too,
1432
01:05:11,569 --> 01:05:13,029
but we're here,
1433
01:05:13,329 --> 01:05:15,609
I mean, our jobs are about Bitcoin.
1434
01:05:16,869 --> 01:05:18,929
So getting open source Bitcoin projects
1435
01:05:18,929 --> 01:05:20,169
and the maintainers
1436
01:05:20,169 --> 01:05:21,829
and trusted people in those projects
1437
01:05:21,829 --> 01:05:24,149
have some kind of reasonable
1438
01:05:24,149 --> 01:05:25,309
trusted access program
1439
01:05:25,309 --> 01:05:26,609
where they can get early access
1440
01:05:26,609 --> 01:05:28,189
to help defend their softwares,
1441
01:05:28,309 --> 01:05:28,969
defend the systems,
1442
01:05:29,369 --> 01:05:32,989
defend a trillion dollar system.
1443
01:05:33,009 --> 01:05:34,769
that the world depends on today.
1444
01:05:35,029 --> 01:05:37,609
So anyway, BPI has a letter
1445
01:05:37,609 --> 01:05:39,449
that they intend to be an open letter.
1446
01:05:39,989 --> 01:05:41,149
It's currently being circulated
1447
01:05:41,149 --> 01:05:42,929
to get people to sign on to.
1448
01:05:43,809 --> 01:05:47,529
I'm super comfortable with like Spiral signing on,
1449
01:05:47,589 --> 01:05:49,789
Project Loop, Presidia Bitcoin.
1450
01:05:50,309 --> 01:05:51,789
I already spoke with other folks.
1451
01:05:52,309 --> 01:05:52,729
We're down.
1452
01:05:53,769 --> 01:05:55,549
And, you know, okay.
1453
01:05:55,889 --> 01:05:59,669
So those are all three relevant projects,
1454
01:05:59,869 --> 01:06:02,789
but we're also working on bigger brand names
1455
01:06:02,789 --> 01:06:04,369
as well to sign on to this letter.
1456
01:06:04,889 --> 01:06:05,749
Yeah, I love it.
1457
01:06:06,689 --> 01:06:08,569
Obviously, it's a tough trade-off,
1458
01:06:08,669 --> 01:06:09,229
but at the end of the day,
1459
01:06:09,369 --> 01:06:10,449
it's like, to your point,
1460
01:06:10,509 --> 01:06:11,169
it's not just Bitcoin, right?
1461
01:06:11,209 --> 01:06:11,769
Like, who's going to do this
1462
01:06:11,769 --> 01:06:12,569
for the internet standards?
1463
01:06:13,649 --> 01:06:14,529
I mean, I'm thinking, like,
1464
01:06:14,749 --> 01:06:15,329
just out loud,
1465
01:06:15,469 --> 01:06:16,449
I hope that, you know,
1466
01:06:17,389 --> 01:06:18,049
the closed-source stuff,
1467
01:06:18,069 --> 01:06:18,909
we talked about the grid earlier.
1468
01:06:19,029 --> 01:06:19,529
I hope the grid,
1469
01:06:19,689 --> 01:06:20,769
the people that are running that stuff
1470
01:06:20,769 --> 01:06:21,749
are getting access to these models
1471
01:06:21,749 --> 01:06:22,509
because it's going to be,
1472
01:06:22,989 --> 01:06:23,769
yeah, it's a transition.
1473
01:06:24,049 --> 01:06:25,349
So the alpha is very exciting,
1474
01:06:25,449 --> 01:06:26,189
but too exciting for me,
1475
01:06:26,229 --> 01:06:26,889
so I'll be right back, guys.
1476
01:06:29,429 --> 01:06:30,689
Do you know much about
1477
01:06:30,689 --> 01:06:32,389
the red team stuff?
1478
01:06:32,389 --> 01:06:38,729
It seems like it kind of came about from Rob just doing a bunch of independent research and asking questions.
1479
01:06:39,249 --> 01:06:41,589
I think that was instigated around the cold card stuff.
1480
01:06:41,689 --> 01:06:45,369
But is it broader than just cold card now or is it still mostly focused on the cold card incident?
1481
01:06:46,649 --> 01:06:52,489
Yeah, I mean, I haven't spoken to Rob about his motivation or the genesis, but just picking up from his tweets.
1482
01:06:52,489 --> 01:07:01,289
And I think it was, I mean, this cold card incident, like, has really rallied the Bitcoin community.
1483
01:07:01,289 --> 01:07:20,929
And another, again, total tragedy, but so many positive things are coming out of it. And one is just beautiful unity in the Bitcoin community that has felt missing for a year and a half. There's just been so much nonsense over ridiculous topics that everyone's kind of fed up about.
1484
01:07:20,929 --> 01:07:23,789
so it's been beautiful seeing that
1485
01:07:23,789 --> 01:07:25,429
I suspect Rob
1486
01:07:25,429 --> 01:07:28,109
deeply cares about Bitcoin
1487
01:07:28,109 --> 01:07:29,649
saw this
1488
01:07:29,649 --> 01:07:31,169
major security incident
1489
01:07:31,169 --> 01:07:32,929
realized that
1490
01:07:32,929 --> 01:07:35,869
he's already been using
1491
01:07:35,869 --> 01:07:37,149
AI tools and I
1492
01:07:37,149 --> 01:07:39,789
suspect he's been security scanning
1493
01:07:39,789 --> 01:07:41,989
his own product so he sort of already had a sense
1494
01:07:41,989 --> 01:07:43,469
for how to go about it
1495
01:07:43,469 --> 01:07:45,869
and that just like kickstarted him
1496
01:07:45,869 --> 01:07:47,789
to go do other
1497
01:07:47,789 --> 01:07:49,949
open source projects and then yeah he's tweeting
1498
01:07:49,949 --> 01:07:52,049
out that he spent like 5,000 and 10,000
1499
01:07:52,049 --> 01:07:53,629
and he's like ponying up a ton of money.
1500
01:07:53,809 --> 01:07:55,689
I saw 40,000 is the most recent number.
1501
01:07:55,769 --> 01:07:56,809
Oh yeah, okay, $40,000.
1502
01:07:57,729 --> 01:07:59,809
So, you know,
1503
01:07:59,869 --> 01:08:01,649
bless his heart and he, you know,
1504
01:08:01,829 --> 01:08:03,949
he's good with
1505
01:08:03,949 --> 01:08:05,709
Twitter, so he's got a good social
1506
01:08:05,709 --> 01:08:07,689
media game as well, which
1507
01:08:07,689 --> 01:08:09,209
created a lot of awareness, a lot of
1508
01:08:09,209 --> 01:08:11,689
momentum and energy. And I think
1509
01:08:11,689 --> 01:08:13,809
Callie joined and I think they said
1510
01:08:13,809 --> 01:08:15,889
they have like 16 people, the last I saw
1511
01:08:15,889 --> 01:08:17,729
of, so I don't know what the latest number is, but you know, others
1512
01:08:17,729 --> 01:08:18,829
of, how many?
1513
01:08:19,749 --> 01:08:20,789
No, I just said, oh, great.
1514
01:08:20,969 --> 01:08:21,649
Oh, yeah, yeah, yeah.
1515
01:08:21,649 --> 01:08:23,909
So, you know, other people have joined the cause.
1516
01:08:23,909 --> 01:08:26,449
I also think they've been able to, you know,
1517
01:08:26,489 --> 01:08:28,629
raise some money now to help fund the tokens.
1518
01:08:29,789 --> 01:08:32,249
So amazing efforts there.
1519
01:08:32,509 --> 01:08:33,749
I mean, I will say that they're...
1520
01:08:33,749 --> 01:08:36,769
It's mostly cold card focused or it's got a lot...
1521
01:08:36,769 --> 01:08:38,029
Oh, no, no, no, no, no.
1522
01:08:38,249 --> 01:08:40,229
It's just become a cold ecosystem thing.
1523
01:08:40,409 --> 01:08:44,089
So it's in a sense like a new type of loop type approach, right?
1524
01:08:44,109 --> 01:08:46,369
Something that's more of like get a bunch of projects,
1525
01:08:46,689 --> 01:08:47,709
a bunch of security analysis.
1526
01:08:47,709 --> 01:08:52,029
Yeah, and it is.
1527
01:08:52,109 --> 01:08:57,289
And like I said before, Loop is not like it's a startup.
1528
01:08:57,529 --> 01:09:00,309
It's not like a for-profit business where we're trying to compete.
1529
01:09:02,369 --> 01:09:11,529
So, like I said earlier, having both is much more valuable for the Bitcoin space than just either one.
1530
01:09:11,689 --> 01:09:15,329
If it was just theirs or just Loop, we'd be weaker than having both.
1531
01:09:15,389 --> 01:09:16,169
Different approaches, yeah.
1532
01:09:16,169 --> 01:09:18,109
Different approaches, different people.
1533
01:09:18,349 --> 01:09:22,309
And, you know, just like, I think all of us are well-intentioned and smart people or whatever,
1534
01:09:22,449 --> 01:09:23,689
but like, we just have different approaches.
1535
01:09:24,109 --> 01:09:27,049
Also, I think they lean more towards open models.
1536
01:09:27,629 --> 01:09:29,429
We lean more towards special access.
1537
01:09:29,669 --> 01:09:31,509
So having that diversity is great.
1538
01:09:31,709 --> 01:09:31,809
Yeah.
1539
01:09:32,109 --> 01:09:38,229
So, I mean, one thing I will throw out on that, you know, I am thrilled with all this happening
1540
01:09:38,229 --> 01:09:42,009
and like, this is, you know, doing God's work to get that special access.
1541
01:09:44,009 --> 01:09:45,829
I mean, it is concerning to me, though.
1542
01:09:45,829 --> 01:09:47,909
I mean, again, you just kind of play out
1543
01:09:47,909 --> 01:09:49,509
the sort of logical conclusion of all this.
1544
01:09:49,649 --> 01:09:51,129
And yes, open source is catching up.
1545
01:09:51,389 --> 01:09:53,269
By the way, I don't know if you guys saw ByteDance,
1546
01:09:53,409 --> 01:09:54,729
their new model that they're supposed to release
1547
01:09:54,729 --> 01:09:55,309
is going to be massive.
1548
01:09:56,029 --> 01:09:59,849
I've heard more rumblings that the Chinese companies,
1549
01:10:00,369 --> 01:10:03,709
maybe they can even jump to the actual forefront,
1550
01:10:04,429 --> 01:10:06,909
which is, that's a much more interesting,
1551
01:10:07,049 --> 01:10:07,729
deeper conversation.
1552
01:10:08,829 --> 01:10:09,549
Oh, I'm sorry.
1553
01:10:09,709 --> 01:10:12,149
Yeah, well, what I'm kind of concerned about is
1554
01:10:12,149 --> 01:10:14,909
if you, I mean, if the open source
1555
01:10:14,909 --> 01:10:16,509
doesn't jump to the actual frontier.
1556
01:10:16,789 --> 01:10:18,829
It's amazing that we can get this access today
1557
01:10:18,829 --> 01:10:21,589
through companies that have the right relationships or whatever.
1558
01:10:21,729 --> 01:10:25,069
But like, I mean, it does make us extremely dependent
1559
01:10:25,069 --> 01:10:26,229
on potentially a duopoly.
1560
01:10:26,849 --> 01:10:28,969
And like going forward, I mean, and you know,
1561
01:10:29,589 --> 01:10:31,409
I know that this is like the conversation, DK,
1562
01:10:31,409 --> 01:10:33,609
I think, you know, we had this many months ago
1563
01:10:33,609 --> 01:10:36,029
where, you know, we're talking about how
1564
01:10:36,029 --> 01:10:41,209
whoever has the like absolute frontier recursive intelligence
1565
01:10:41,209 --> 01:10:43,769
effectively becomes the new power broker,
1566
01:10:43,769 --> 01:10:45,009
the new government effectively.
1567
01:10:45,889 --> 01:10:47,749
And, you know, in this transition thing.
1568
01:10:48,089 --> 01:10:49,069
All the energy and hardware.
1569
01:10:49,409 --> 01:10:50,309
They have the energy and hardware,
1570
01:10:50,429 --> 01:10:51,149
but like, I mean,
1571
01:10:51,429 --> 01:10:51,949
I don't, I mean,
1572
01:10:52,589 --> 01:10:53,969
Anthropics are doing a pretty good job of that right now.
1573
01:10:54,129 --> 01:10:56,669
So, I don't know.
1574
01:10:56,769 --> 01:10:57,289
It's like it.
1575
01:10:57,409 --> 01:10:58,969
Well, the U.S. government has a military
1576
01:10:58,969 --> 01:11:01,329
that can take over energy resources.
1577
01:11:01,749 --> 01:11:03,669
Most assets, most physical assets are
1578
01:11:03,669 --> 01:11:05,309
seizable, yeah.
1579
01:11:05,669 --> 01:11:06,109
For now.
1580
01:11:06,409 --> 01:11:08,249
And like, but I expect to see companies
1581
01:11:08,249 --> 01:11:09,509
continue to move around the world.
1582
01:11:09,509 --> 01:11:10,529
I expect to see companies
1583
01:11:10,529 --> 01:11:12,149
once they have access to robots.
1584
01:11:12,309 --> 01:11:13,629
Like, I just wouldn't,
1585
01:11:13,769 --> 01:11:17,989
I wouldn't underestimate how quickly some of those problems can be solved.
1586
01:11:18,649 --> 01:11:21,569
And I don't know, it's a really weird time because it's like,
1587
01:11:22,329 --> 01:11:25,009
you know, you do the best you can to get access to what we have today,
1588
01:11:25,109 --> 01:11:28,769
but it's scary to be potentially so dependent on a doable.
1589
01:11:29,649 --> 01:11:31,189
So just a few comments here.
1590
01:11:31,609 --> 01:11:34,229
Deanna updates that there's 22 people involved now in the project,
1591
01:11:34,369 --> 01:11:35,889
she says, this Red Team project.
1592
01:11:37,409 --> 01:11:39,449
Another person says 6,000 vulnerabilities.
1593
01:11:40,529 --> 01:11:43,009
I'm guessing they reported that they submitted that many.
1594
01:11:43,769 --> 01:11:48,949
loop is not as marketing heavy as red team.
1595
01:11:49,089 --> 01:11:52,109
So we put out a tweet a couple weeks ago
1596
01:11:52,109 --> 01:11:54,449
about the results from phase one,
1597
01:11:54,869 --> 01:11:56,349
which did quote some numbers.
1598
01:11:59,089 --> 01:12:01,849
Kind of, well, good or bad news, number goes up.
1599
01:12:02,369 --> 01:12:04,529
Like, it's definitely number goes up from loop
1600
01:12:04,529 --> 01:12:06,549
in terms of number of project, number of vulnerabilities,
1601
01:12:06,789 --> 01:12:08,169
number of severe vulnerabilities.
1602
01:12:08,869 --> 01:12:09,789
I think that's good.
1603
01:12:09,869 --> 01:12:12,989
I mean, if they exist, it's good that they're being exposed.
1604
01:12:12,989 --> 01:12:15,069
so in that sense, number goes up as good.
1605
01:12:15,269 --> 01:12:17,209
But obviously, hopefully, number goes down
1606
01:12:17,209 --> 01:12:19,149
and plateaus
1607
01:12:19,149 --> 01:12:20,249
over time.
1608
01:12:21,929 --> 01:12:23,209
I also want to mention
1609
01:12:23,209 --> 01:12:24,869
another plug for the
1610
01:12:24,869 --> 01:12:27,149
PB Open Source AI Summit in September,
1611
01:12:27,249 --> 01:12:28,589
September 10th and 11th.
1612
01:12:28,969 --> 01:12:31,029
A panel we're trying to put together
1613
01:12:31,029 --> 01:12:32,909
now for that is around this topic.
1614
01:12:33,069 --> 01:12:35,229
We're trying to get some combination
1615
01:12:35,229 --> 01:12:36,529
of Rob or Callie,
1616
01:12:37,429 --> 01:12:38,869
maybe someone from Block Security,
1617
01:12:39,509 --> 01:12:40,989
someone from Project Loop, here
1618
01:12:40,989 --> 01:12:42,909
for a panel to talk about this very topic.
1619
01:12:42,989 --> 01:12:45,449
and hear it directly from the people working on this stuff.
1620
01:12:45,909 --> 01:12:47,789
So hopefully we pull that off.
1621
01:12:47,969 --> 01:12:49,549
It'd be interesting to get someone from Anthropic too.
1622
01:12:50,429 --> 01:12:50,889
I have an idea.
1623
01:12:50,969 --> 01:12:51,449
I knew that could be.
1624
01:12:51,669 --> 01:12:51,909
Okay.
1625
01:12:52,129 --> 01:12:52,749
Yeah, we shot that.
1626
01:12:53,189 --> 01:12:55,669
Anyways, that, yeah.
1627
01:12:56,569 --> 01:12:58,429
I had a question for you guys.
1628
01:12:58,949 --> 01:12:59,849
You know, just with all the,
1629
01:12:59,849 --> 01:13:03,609
I think we're going through a very volatile time
1630
01:13:03,609 --> 01:13:06,009
with everything in the world, right?
1631
01:13:06,089 --> 01:13:06,809
Everything technology.
1632
01:13:07,009 --> 01:13:08,849
And so if you think about, you know,
1633
01:13:08,869 --> 01:13:12,909
it's going to be some time before we get to the safer,
1634
01:13:12,989 --> 01:13:19,409
side of things where maybe there's a little bit less volatility. And in that process, it kind of
1635
01:13:19,409 --> 01:13:24,289
made me think a lot about like, you know, the entropy question that has been coming up a lot.
1636
01:13:24,289 --> 01:13:31,569
Like, how do you actually do good entropy? And I don't know if you guys have been thinking about
1637
01:13:31,569 --> 01:13:39,709
this a lot, but, you know, the cold card incident obviously put a lot of focus on this. And then,
1638
01:13:39,709 --> 01:13:47,169
But a lot of people take it as an assumption that as long as you roll dice and tap the right button on the cold card that you're safe.
1639
01:13:47,509 --> 01:13:48,449
That might be true.
1640
01:13:49,109 --> 01:13:56,469
But you're still trusting that it's applying the dice entropy to the seed.
1641
01:13:57,629 --> 01:14:05,249
And then separately, I've seen people trying to vibe code their own entropy generation, which seems like a really bad idea.
1642
01:14:05,249 --> 01:14:07,749
Seems like a horrible idea, right?
1643
01:14:07,749 --> 01:14:29,775
So I was wondering just because my head bent in this space a bit I wonder if you guys know of any open source tools that could be sort of used as an assistant to say dice rolling with your own where you sort of can you know there a well maybe a script that helps you with entropy generation from Dice
1644
01:14:29,775 --> 01:14:31,475
that's sort of like an Ian Coleman type thing.
1645
01:14:32,235 --> 01:14:32,995
Well, I have lots,
1646
01:14:33,495 --> 01:14:35,555
I don't necessarily have an answer to that specific question,
1647
01:14:35,615 --> 01:14:36,975
but I have a lot of thoughts on the general topic.
1648
01:14:37,235 --> 01:14:38,475
And let's start with an anecdote.
1649
01:14:38,795 --> 01:14:41,295
So something that was tweeted out the past few days
1650
01:14:41,295 --> 01:14:46,935
is a 2022 support conversation
1651
01:14:46,935 --> 01:14:49,755
that happened in a cold card telegram group.
1652
01:14:51,035 --> 01:14:52,955
And this cold card,
1653
01:14:53,135 --> 01:14:55,175
or someone who said they're a cold card customer,
1654
01:14:55,275 --> 01:14:56,315
they just bought a cold card
1655
01:14:56,315 --> 01:14:59,115
and they deposited Bitcoin on it
1656
01:14:59,115 --> 01:15:01,675
and immediately lost the money.
1657
01:15:02,855 --> 01:15:03,635
So when I first started,
1658
01:15:04,155 --> 01:15:07,055
and the tweet posted a video
1659
01:15:07,055 --> 01:15:08,895
scrolling through the telegram conversation.
1660
01:15:09,035 --> 01:15:10,395
So I'm like watching the video.
1661
01:15:10,395 --> 01:15:11,415
And at first I'm like,
1662
01:15:11,495 --> 01:15:13,615
oh, this was someone who,
1663
01:15:14,235 --> 01:15:16,795
or it was an early exploit
1664
01:15:16,795 --> 01:15:18,495
of this bug back in 2022.
1665
01:15:18,695 --> 01:15:22,555
But it turns out that was not the case.
1666
01:15:22,935 --> 01:15:25,515
It turns out what that customer did,
1667
01:15:26,035 --> 01:15:27,875
they actually did the dice roll feature.
1668
01:15:28,275 --> 01:15:31,635
So they were, in theory, protected from this vulnerability.
1669
01:15:32,395 --> 01:15:35,595
But then guess how they did the dice roll?
1670
01:15:36,335 --> 01:15:39,175
So with the cold card, the way you do the dice roll,
1671
01:15:39,315 --> 01:15:43,375
you roll a die, and then if it rolls three,
1672
01:15:43,655 --> 01:15:46,615
you type in three or whatever, and you do that 100 times
1673
01:15:46,615 --> 01:15:46,995
in a row.
1674
01:15:48,175 --> 01:15:50,555
My concern, I'm a huge
1675
01:15:50,555 --> 01:15:52,575
fan of creating strong
1676
01:15:52,575 --> 01:15:54,195
entropy using dice rolls to do it.
1677
01:15:54,915 --> 01:15:56,395
You really got to know what you're doing.
1678
01:15:56,475 --> 01:15:58,575
There's a number of foot guns that you need
1679
01:15:58,575 --> 01:16:00,055
to avoid so you really have to know what you're doing.
1680
01:16:00,655 --> 01:16:02,115
This person was like
1681
01:16:02,115 --> 01:16:04,255
foot gun heavy because what they did
1682
01:16:04,255 --> 01:16:05,815
is they entered six
1683
01:16:05,815 --> 01:16:07,475
100 times in a row.
1684
01:16:08,315 --> 01:16:08,755
What?
1685
01:16:10,895 --> 01:16:12,475
They entered six 100
1686
01:16:12,475 --> 01:16:14,255
times in a row. So they didn't actually roll the dice.
1687
01:16:14,915 --> 01:16:15,535
Yeah, he's like
1688
01:16:15,535 --> 01:16:18,195
he didn't think he needed to roll a dice.
1689
01:16:18,295 --> 01:16:20,915
He thought that the cold card
1690
01:16:20,915 --> 01:16:23,615
was going to roll the dice for him.
1691
01:16:24,215 --> 01:16:26,315
So he entered sticks 100 times in a row,
1692
01:16:26,515 --> 01:16:28,035
generated his wallet.
1693
01:16:28,595 --> 01:16:28,935
Dice.
1694
01:16:29,735 --> 01:16:31,595
And then I think he noticed
1695
01:16:31,595 --> 01:16:33,535
that there were already transactions
1696
01:16:33,535 --> 01:16:37,175
on his wallet from a prior date.
1697
01:16:37,575 --> 01:16:39,475
It's because he generated a private key
1698
01:16:39,475 --> 01:16:40,955
that collided with someone else's wallet.
1699
01:16:41,115 --> 01:16:41,715
He literally...
1700
01:16:41,715 --> 01:16:43,235
Other people did the same thing.
1701
01:16:43,415 --> 01:16:45,015
So there's still an open question
1702
01:16:45,015 --> 01:16:47,075
about how is that other wallet created?
1703
01:16:47,195 --> 01:16:50,075
Because the other wallet clearly also had weak entropy
1704
01:16:50,075 --> 01:16:51,635
because if it had strong entropy,
1705
01:16:51,755 --> 01:16:54,195
it would be impossible to collide.
1706
01:16:54,695 --> 01:16:58,175
So somehow the other wallet maybe did the same error
1707
01:16:58,175 --> 01:16:59,415
or did use exploit.
1708
01:16:59,515 --> 01:17:00,055
I have no idea.
1709
01:17:00,415 --> 01:17:02,575
But anyway, the point of this though is like
1710
01:17:02,575 --> 01:17:04,615
human beings are not,
1711
01:17:04,615 --> 01:17:07,315
most human beings are not in a position
1712
01:17:07,315 --> 01:17:09,735
to properly create their own entropy
1713
01:17:09,735 --> 01:17:15,255
due to their time availability in their life,
1714
01:17:16,155 --> 01:17:18,475
due to whether they have the kind of mindset
1715
01:17:18,475 --> 01:17:22,895
that can really focus on the task.
1716
01:17:23,035 --> 01:17:24,335
No, a lot of people cannot focus.
1717
01:17:24,895 --> 01:17:26,435
Insufficient autism is what he's saying.
1718
01:17:27,115 --> 01:17:28,495
Well, however you want to frame it.
1719
01:17:28,495 --> 01:17:30,835
A lot of people are, you know,
1720
01:17:31,155 --> 01:17:33,395
you need to have enough time.
1721
01:17:33,515 --> 01:17:34,495
You need to care about it.
1722
01:17:34,735 --> 01:17:36,735
You need to be aware that you need to care about it.
1723
01:17:36,795 --> 01:17:37,695
Then you need to care about it.
1724
01:17:37,775 --> 01:17:39,015
Then you need to dedicate enough time.
1725
01:17:39,015 --> 01:17:42,115
then you need to come up to speed on what are the do's and don'ts.
1726
01:17:42,295 --> 01:17:43,315
So you have to do all that.
1727
01:17:43,435 --> 01:17:45,215
That eliminates a lot of the population already.
1728
01:17:45,435 --> 01:17:47,015
And then you get to a point where,
1729
01:17:47,375 --> 01:17:48,895
okay, if someone checks all those boxes,
1730
01:17:49,335 --> 01:17:52,115
do they actually execute on the plan successfully?
1731
01:17:52,775 --> 01:17:56,215
And anyway, you narrow it down to like some small percentage of the population
1732
01:17:56,215 --> 01:17:57,135
that checks all the boxes.
1733
01:17:57,715 --> 01:18:02,275
So I think it's really important that there are tools and products in the space
1734
01:18:02,275 --> 01:18:05,675
that allow for this because there are people who do check all those boxes.
1735
01:18:05,935 --> 01:18:06,235
Yeah, of course.
1736
01:18:06,235 --> 01:18:08,975
And that gives them excellent self-sovereignty.
1737
01:18:09,015 --> 01:18:15,475
but there also needs to be products and processes in place for the rest of us.
1738
01:18:15,475 --> 01:18:15,835
Yes.
1739
01:18:16,075 --> 01:18:19,435
And because most people are not cut out to do their own entropy.
1740
01:18:19,595 --> 01:18:21,595
So that's one big point I wanted to make.
1741
01:18:22,035 --> 01:18:25,915
Then in terms of like something I've always thought makes a lot of sense
1742
01:18:25,915 --> 01:18:29,995
is a combination of the two and whether you're using multi-sig
1743
01:18:29,995 --> 01:18:31,295
or some other system or whatever,
1744
01:18:31,435 --> 01:18:37,875
but meaning like part of your secrets are generated by the product.
1745
01:18:37,875 --> 01:18:37,975
Yeah.
1746
01:18:39,015 --> 01:18:43,015
and you just trust the product and ideally you're picking a product that's open source and has been
1747
01:18:43,015 --> 01:18:48,135
is popular has been audited ai's scanned it blah blah blah but you just trust the product and then
1748
01:18:48,135 --> 01:18:53,255
part of your secrets are your own entropy and neither and you should have it set up where
1749
01:18:53,255 --> 01:18:59,895
such that if either gets broken it's not a single point of failure exactly um because then you
1750
01:18:59,895 --> 01:19:04,695
protect yourself against yourself because even if you think you've checked all the boxes on entropy
1751
01:19:04,695 --> 01:19:06,855
and think you're like Mr. Smarty Pants,
1752
01:19:07,015 --> 01:19:08,675
like you might have made a mistake, right?
1753
01:19:08,915 --> 01:19:12,395
So I'd rather have a system
1754
01:19:12,395 --> 01:19:13,995
where even if I made a mistake,
1755
01:19:14,095 --> 01:19:15,135
I don't lose all my Bitcoin.
1756
01:19:15,675 --> 01:19:17,995
And then likewise, obviously, cold cards taught us.
1757
01:19:18,975 --> 01:19:20,975
I mean, some of us were already paranoid
1758
01:19:20,975 --> 01:19:22,555
prior to the cold card incident.
1759
01:19:22,715 --> 01:19:24,155
Now a lot more people are paranoid,
1760
01:19:24,775 --> 01:19:25,835
as they should be,
1761
01:19:25,835 --> 01:19:28,255
about fully trusting one other product
1762
01:19:28,255 --> 01:19:29,215
in a single point of failure.
1763
01:19:29,995 --> 01:19:32,295
So yeah, that diversity,
1764
01:19:32,295 --> 01:19:34,415
we've talked about custody diversification
1765
01:19:34,415 --> 01:19:37,095
here don't put all your eggs in one basket
1766
01:19:37,095 --> 01:19:38,695
when it comes to
1767
01:19:38,695 --> 01:19:40,175
a given solution too
1768
01:19:40,175 --> 01:19:42,855
a diversity of your own
1769
01:19:42,855 --> 01:19:44,795
entropy and products
1770
01:19:44,795 --> 01:19:46,335
I think is a good approach as well
1771
01:19:46,335 --> 01:19:48,635
one thing I just wanted to call out
1772
01:19:48,635 --> 01:19:50,215
we talked about in the chat is
1773
01:19:50,215 --> 01:19:52,095
I think
1774
01:19:52,095 --> 01:19:54,875
this whole incident has done
1775
01:19:54,875 --> 01:19:56,675
a huge amount unfortunately
1776
01:19:56,675 --> 01:19:59,115
to kind of like discourage self-custody
1777
01:19:59,115 --> 01:20:00,835
and even if that's a small amount of people
1778
01:20:00,835 --> 01:20:02,755
those people keep the system honest.
1779
01:20:02,915 --> 01:20:03,875
They're sort of like the
1780
01:20:03,875 --> 01:20:06,735
lymphatic system or whatever for Bitcoin.
1781
01:20:06,935 --> 01:20:08,035
So that's very unfortunate.
1782
01:20:08,595 --> 01:20:10,955
However, if you look just in terms of
1783
01:20:10,955 --> 01:20:12,235
the numbers, right?
1784
01:20:12,255 --> 01:20:13,975
I think, I don't know what the total number was,
1785
01:20:14,015 --> 01:20:14,675
but I think it was less,
1786
01:20:15,055 --> 01:20:17,515
fewer than 2,000 Bitcoin so far
1787
01:20:17,515 --> 01:20:18,535
in the cold card hack,
1788
01:20:18,595 --> 01:20:19,935
which obviously it's a tragedy
1789
01:20:19,935 --> 01:20:20,715
for all those people
1790
01:20:20,715 --> 01:20:21,635
because most of them probably lost
1791
01:20:21,635 --> 01:20:22,255
their life savings,
1792
01:20:22,675 --> 01:20:23,695
extremely messed up.
1793
01:20:24,115 --> 01:20:24,975
But, you know,
1794
01:20:25,155 --> 01:20:26,235
compared to where we are
1795
01:20:26,235 --> 01:20:26,935
with the rest of the world,
1796
01:20:26,995 --> 01:20:27,815
like obviously we talked about,
1797
01:20:28,175 --> 01:20:29,175
you know, didn't move the price at all.
1798
01:20:29,175 --> 01:20:30,735
The price actually has gone up since then.
1799
01:20:30,835 --> 01:20:45,375
And I think the reality of the situation is that at this stage, you want those options available. You want to continue to lower the bar for more people to be able to do this. But honestly, I think it's at a stage where institutions are driving most of the adoption and usage of Bitcoin.
1800
01:20:45,375 --> 01:20:51,215
So I think the real question for most people is, you know, in addition to potentially doing
1801
01:20:51,215 --> 01:20:55,015
self-custody, if that's something that you really feel like you could get to that point,
1802
01:20:55,375 --> 01:21:01,215
but of the custodians that are out there, how do you make the best guess of which of
1803
01:21:01,215 --> 01:21:01,835
those custodians?
1804
01:21:01,975 --> 01:21:05,555
And probably diversity is a big part of the answer, but which of those custodians are
1805
01:21:05,555 --> 01:21:06,255
going to still be around?
1806
01:21:06,355 --> 01:21:07,335
Which ones are not going to rug you?
1807
01:21:07,415 --> 01:21:11,035
Which ones are getting access to the latest open source model, or not open source, closed
1808
01:21:11,035 --> 01:21:11,555
source models?
1809
01:21:11,775 --> 01:21:14,375
We still don't have a great answer because of like the private agreements you were talking
1810
01:21:14,375 --> 01:21:14,655
about.
1811
01:21:14,655 --> 01:21:22,615
But I honestly think for most people trying to evaluate some mix of the best custodians right now, I just think that's the stage where Bitcoin is.
1812
01:21:22,715 --> 01:21:29,375
Well, and I would recommend customers of custodians or prospective customers demand more transparency.
1813
01:21:30,035 --> 01:21:36,035
How are they adapting to this new world of AI security scanning and vulnerability exposure?
1814
01:21:36,195 --> 01:21:36,555
Yes, I agree.
1815
01:21:36,815 --> 01:21:39,395
I would like to see every custodian publish something like that.
1816
01:21:39,395 --> 01:21:43,615
There should be pressure on these companies to be more forthcoming.
1817
01:21:43,615 --> 01:21:52,075
yeah um so uh well i don't know if it's related or not but another yeah go ahead oh good i well i
1818
01:21:52,075 --> 01:21:57,975
wanted to one thing i asked in the chat that i didn't hear from but i i shared a a tweet thread
1819
01:21:57,975 --> 01:22:04,475
that was from extractive ghost of unhosted marcellus do you guys read through that what
1820
01:22:04,475 --> 01:22:10,235
no i don't know what that is it was it just came out i think this morning or a few hours ago uh
1821
01:22:10,235 --> 01:22:18,955
But it kind of looked through a bunch of the history of the cold card coin kite CTO's work.
1822
01:22:19,615 --> 01:22:25,635
And it looks like I think he's sort of there's been this theory floating around that he's maybe talking to himself in a nim and stuff.
1823
01:22:26,095 --> 01:22:33,215
And I think he was reporting it as basically smoking gun evidence that the coin kite thing was an inside job.
1824
01:22:33,315 --> 01:22:38,175
And I didn't I just want to talk to you guys to understand if you guys saw that or have seen any of that chatter.
1825
01:22:38,175 --> 01:22:41,295
if you're fully convinced or believe it's possible
1826
01:22:41,295 --> 01:22:43,055
or sort of where you land on that.
1827
01:22:43,075 --> 01:22:44,695
That particular account you mentioned,
1828
01:22:44,855 --> 01:22:45,915
I'd have to go double-check it,
1829
01:22:46,035 --> 01:22:49,515
but my recollection is that
1830
01:22:49,515 --> 01:22:51,435
maybe not the most trustworthy account.
1831
01:22:51,775 --> 01:22:52,295
Having said that,
1832
01:22:53,015 --> 01:22:54,675
with a name like that?
1833
01:22:55,095 --> 01:22:56,475
Well, I also think it's associated
1834
01:22:56,475 --> 01:22:57,935
with other topics that I disagree with,
1835
01:22:58,035 --> 01:22:58,435
but I don't know.
1836
01:22:58,475 --> 01:22:59,875
I'd have to go vet that.
1837
01:22:59,875 --> 01:23:04,415
But certainly, James O'Byrne is posting about this as well,
1838
01:23:04,895 --> 01:23:06,455
and Alicos and other people.
1839
01:23:06,455 --> 01:23:08,475
I respect.
1840
01:23:11,475 --> 01:23:12,655
It's definitely
1841
01:23:12,655 --> 01:23:14,315
odd
1842
01:23:14,315 --> 01:23:16,155
that
1843
01:23:16,155 --> 01:23:17,875
Peter was
1844
01:23:17,875 --> 01:23:20,955
using two different identities
1845
01:23:20,955 --> 01:23:23,055
on the code base and talking to himself.
1846
01:23:23,735 --> 01:23:24,795
That's kind of weird.
1847
01:23:25,195 --> 01:23:25,935
But I don't know.
1848
01:23:25,935 --> 01:23:27,895
I generally still
1849
01:23:27,895 --> 01:23:30,075
don't believe it was an inside job
1850
01:23:30,075 --> 01:23:31,835
or intentional. I think it was just
1851
01:23:31,835 --> 01:23:33,815
an absolutely
1852
01:23:33,815 --> 01:23:35,935
horrible job of testing their product.
1853
01:23:36,455 --> 01:23:39,335
like the bug itself
1854
01:23:39,335 --> 01:23:41,295
anyone that's programmed C and dealt with
1855
01:23:41,295 --> 01:23:43,175
like make files and pre-processing
1856
01:23:43,175 --> 01:23:45,035
absolutely knows how painful
1857
01:23:45,035 --> 01:23:46,935
it is and foot gun
1858
01:23:46,935 --> 01:23:49,275
heavy and so looking
1859
01:23:49,275 --> 01:23:51,135
at that it's like
1860
01:23:51,135 --> 01:23:53,255
yep I remember those days with C
1861
01:23:53,255 --> 01:23:55,075
so the bug itself
1862
01:23:55,075 --> 01:23:57,395
is
1863
01:23:57,395 --> 01:23:59,055
common with C
1864
01:23:59,055 --> 01:24:01,135
but like the lack of testing
1865
01:24:01,135 --> 01:24:03,295
they had for like one of the most
1866
01:24:03,295 --> 01:24:05,315
critical if not the most critical parts of the product
1867
01:24:05,315 --> 01:24:10,975
is clearly a major disaster by them.
1868
01:24:12,535 --> 01:24:13,715
I still lean towards,
1869
01:24:13,855 --> 01:24:17,795
I don't really think this is some kind of retirement attack.
1870
01:24:18,135 --> 01:24:19,995
People are citing that they tweeted
1871
01:24:19,995 --> 01:24:21,575
about retirement attacks in the past,
1872
01:24:21,815 --> 01:24:24,915
but that would be like us talking about retirement attacks
1873
01:24:24,915 --> 01:24:27,755
right now, and then four years from now,
1874
01:24:28,015 --> 01:24:29,175
then all of a sudden we're being accused,
1875
01:24:29,295 --> 01:24:30,155
oh, they planned it all.
1876
01:24:30,295 --> 01:24:31,355
It's like, no, we're just talking about
1877
01:24:31,355 --> 01:24:32,455
the concept of retirement attack.
1878
01:24:32,455 --> 01:24:40,355
act. I think that's why they tweeted that. So I don't know. I certainly reserve judgment on them
1879
01:24:40,355 --> 01:24:46,035
as criminals or intentionally doing this. Okay. Yeah. I just wanted to understand if that is,
1880
01:24:46,395 --> 01:24:51,355
that's my gut is like, still, let's get more information and understand the situation better.
1881
01:24:51,455 --> 01:24:55,095
But there's a lot of people, at least this morning, referencing this thread. And I was kind of like,
1882
01:24:55,095 --> 01:25:00,155
okay, is this, you know, if people, I don't know. If you saw the thread, if you saw the thread,
1883
01:25:00,155 --> 01:25:00,835
tell us.
1884
01:25:02,175 --> 01:25:02,415
I mean,
1885
01:25:02,515 --> 01:25:02,915
you know,
1886
01:25:02,935 --> 01:25:03,575
he's pointing out
1887
01:25:03,575 --> 01:25:04,235
things like
1888
01:25:04,235 --> 01:25:04,735
the,
1889
01:25:05,275 --> 01:25:05,775
the,
1890
01:25:06,435 --> 01:25:07,135
I guess it was,
1891
01:25:07,335 --> 01:25:07,495
you know,
1892
01:25:07,535 --> 01:25:07,795
the,
1893
01:25:07,995 --> 01:25:08,775
was it Switch
1894
01:25:08,775 --> 01:25:10,015
is the pseudonym
1895
01:25:10,015 --> 01:25:10,955
that he was using?
1896
01:25:11,235 --> 01:25:12,795
S-W-I-T-C-K?
1897
01:25:12,895 --> 01:25:13,535
They're both nims.
1898
01:25:13,775 --> 01:25:13,895
Yeah.
1899
01:25:14,515 --> 01:25:14,675
Oh,
1900
01:25:14,755 --> 01:25:14,935
okay.
1901
01:25:15,395 --> 01:25:15,675
But they,
1902
01:25:16,195 --> 01:25:17,415
Doc Hex is not his name either.
1903
01:25:17,615 --> 01:25:17,815
Okay,
1904
01:25:17,895 --> 01:25:18,015
right.
1905
01:25:18,355 --> 01:25:19,315
Doc Hex and Switch.
1906
01:25:19,355 --> 01:25:20,215
You don't say.
1907
01:25:20,415 --> 01:25:21,055
Kind of sock,
1908
01:25:21,175 --> 01:25:22,415
sock puppeting each other
1909
01:25:22,415 --> 01:25:22,735
or whatever.
1910
01:25:23,015 --> 01:25:23,735
But I think one of them
1911
01:25:23,735 --> 01:25:25,055
had just created the account
1912
01:25:25,055 --> 01:25:25,955
three months before
1913
01:25:25,955 --> 01:25:26,435
and just,
1914
01:25:26,675 --> 01:25:27,495
it was kind of an Anon
1915
01:25:27,495 --> 01:25:29,035
that just made his first contribution
1916
01:25:29,035 --> 01:25:29,715
and then all of a sudden
1917
01:25:29,715 --> 01:25:33,695
And Coldcard shows up pulling down this random Anon's first contribution,
1918
01:25:34,415 --> 01:25:39,995
asking to include it in the code of coin, Coldcard, CoinKite.
1919
01:25:40,555 --> 01:25:43,035
So it just kind of smelled really funny.
1920
01:25:43,315 --> 01:25:45,455
It's not a smoking gun by any means,
1921
01:25:45,455 --> 01:25:48,995
but just wanted to understand if you guys had,
1922
01:25:49,335 --> 01:25:52,995
it sounds like still kind of paused, waiting, thinking about it.
1923
01:25:53,275 --> 01:25:58,395
I think it's easy to go back and look at information and form.
1924
01:25:58,395 --> 01:26:01,635
form of view that's not correct.
1925
01:26:01,915 --> 01:26:02,855
But I mean, I don't know.
1926
01:26:02,975 --> 01:26:04,355
I mean, it seems pretty clear.
1927
01:26:04,735 --> 01:26:07,455
Like, we know that there is this big brouhaha
1928
01:26:07,455 --> 01:26:09,415
over, like, it was FOSS,
1929
01:26:09,475 --> 01:26:11,995
and then they didn't like that people are using it as FOSS,
1930
01:26:12,095 --> 01:26:13,655
so they changed the license, and then that,
1931
01:26:13,935 --> 01:26:14,955
like, we talked about this last week.
1932
01:26:15,035 --> 01:26:16,675
Then you have to, like, change out the code,
1933
01:26:16,755 --> 01:26:18,515
and they imported all this new code.
1934
01:26:18,815 --> 01:26:21,295
Turned out to be a disastrous decision,
1935
01:26:21,775 --> 01:26:22,815
and they just imported all this code.
1936
01:26:23,335 --> 01:26:26,035
Like, that all seems quite factual, and we know that.
1937
01:26:26,035 --> 01:26:28,755
now why was there
1938
01:26:28,755 --> 01:26:29,635
the same
1939
01:26:29,635 --> 01:26:32,955
and I think it's pretty proven at this point
1940
01:26:32,955 --> 01:26:35,055
I don't even know if it's fully proven that the same person
1941
01:26:35,055 --> 01:26:37,295
had the two NIMS but let's assume that's true
1942
01:26:37,295 --> 01:26:37,915
I think it is
1943
01:26:37,915 --> 01:26:40,015
unclear why
1944
01:26:40,015 --> 01:26:43,195
they're talking to each other that way
1945
01:26:43,195 --> 01:26:45,075
but I don't know, developers are weird
1946
01:26:45,075 --> 01:26:47,075
that's also true, developers are weird
1947
01:26:47,075 --> 01:26:49,195
I think a useful framework for these kinds
1948
01:26:49,195 --> 01:26:50,715
of questions in general is like
1949
01:26:50,715 --> 01:26:53,235
I mean look, obviously the world's a weird place
1950
01:26:53,235 --> 01:26:54,855
and I think it's
1951
01:26:54,855 --> 01:26:59,455
it's a good idea to keep an open mind in many fronts and just like say, we don't know. Um,
1952
01:26:59,795 --> 01:27:04,935
however, I mean, well, I'll say two things. One, a useful framework is to, when you see
1953
01:27:04,935 --> 01:27:12,795
events in the world, um, to ask who benefits and why. Yeah. Yeah. And, um, you know, I, um,
1954
01:27:13,855 --> 01:27:17,635
I think it's extremely unlikely it was actually a retirement attack to get money.
1955
01:27:18,055 --> 01:27:22,555
Uh, because the difficulty in trying to launder that, like you're going to get caught in the
1956
01:27:22,555 --> 01:27:24,655
process probably someday those people probably will.
1957
01:27:26,415 --> 01:27:28,575
I mean, you can do your own speculation. Are there other
1958
01:27:28,575 --> 01:27:31,695
reasons that people might do that that are not retirement attack? I mean, you can
1959
01:27:31,695 --> 01:27:34,515
think about that on your own. This thread was
1960
01:27:34,515 --> 01:27:37,695
proposing also that I think a historical archive of
1961
01:27:37,695 --> 01:27:40,535
the Doc Hex had a resume up that
1962
01:27:40,535 --> 01:27:43,615
said he had built some sort of KVM switch for
1963
01:27:43,615 --> 01:27:46,595
a company that was registered to do business
1964
01:27:46,595 --> 01:27:49,595
with the federal government. So maybe, you know, there's kind of like a
1965
01:27:49,595 --> 01:27:52,015
That's always a possibility in this case.
1966
01:27:52,555 --> 01:27:53,095
I'm not.
1967
01:27:53,495 --> 01:27:56,775
But I think the bigger takeaway from all of this is like,
1968
01:27:56,875 --> 01:27:58,615
regardless, like you should assume,
1969
01:27:58,715 --> 01:27:59,715
be paranoid on every front.
1970
01:27:59,835 --> 01:28:02,435
The bigger takeaway is right now,
1971
01:28:03,135 --> 01:28:04,475
the AI thing is real.
1972
01:28:04,855 --> 01:28:07,615
Whether, no matter whose incentives
1973
01:28:07,615 --> 01:28:09,095
you're pushing in whatever direction.
1974
01:28:09,955 --> 01:28:11,935
And the best thing you can do, in my opinion,
1975
01:28:12,035 --> 01:28:14,755
is make sure you are aware of that
1976
01:28:14,755 --> 01:28:18,035
and working with people that are actively
1977
01:28:18,035 --> 01:28:19,415
trying to guard against the AI thing.
1978
01:28:19,935 --> 01:28:20,475
All right.
1979
01:28:20,615 --> 01:28:21,955
I got a few more things I want to get in the show.
1980
01:28:21,955 --> 01:28:22,875
and we got a few minutes left.
1981
01:28:23,015 --> 01:28:24,135
So really quick,
1982
01:28:24,195 --> 01:28:24,675
I want to talk about
1983
01:28:24,675 --> 01:28:25,815
responsible disclosure
1984
01:28:25,815 --> 01:28:28,575
and bug bounties.
1985
01:28:28,735 --> 01:28:30,355
But so right after
1986
01:28:30,355 --> 01:28:31,135
the cold card incident,
1987
01:28:31,255 --> 01:28:31,795
I think it was like,
1988
01:28:32,055 --> 01:28:32,195
you know,
1989
01:28:32,195 --> 01:28:33,475
so that started Thursday.
1990
01:28:33,695 --> 01:28:34,535
We talked about it on Friday.
1991
01:28:34,535 --> 01:28:36,115
And then the next morning
1992
01:28:36,115 --> 01:28:37,235
at like 3 a.m.,
1993
01:28:37,235 --> 01:28:37,875
I wake up.
1994
01:28:38,955 --> 01:28:39,795
I've been waking up
1995
01:28:39,795 --> 01:28:40,575
at 3 or 4 a.m.
1996
01:28:41,095 --> 01:28:42,655
every day since the cold card incident.
1997
01:28:44,295 --> 01:28:45,055
And yeah,
1998
01:28:45,115 --> 01:28:46,035
highly recommend everyone
1999
01:28:46,035 --> 01:28:46,755
who's doing the same
2000
01:28:46,755 --> 01:28:47,555
and working so hard,
2001
01:28:47,715 --> 01:28:48,915
get out and get a little exercise
2002
01:28:48,915 --> 01:28:49,695
in every day too.
2003
01:28:50,695 --> 01:28:51,435
I'm at break.
2004
01:28:51,435 --> 01:28:53,655
I'm also exercising every day to keep myself sane,
2005
01:28:53,775 --> 01:28:55,795
but I've definitely been working double time.
2006
01:28:56,295 --> 01:28:57,535
Anyway, woke up early Saturday,
2007
01:28:58,195 --> 01:29:01,315
and I see that there's a bit key vulnerability.
2008
01:29:02,195 --> 01:29:05,475
Severe vulnerability is how Floppy tweeted it.
2009
01:29:05,475 --> 01:29:21,040
So this is an example of not responsible disclosure I guess irresponsible disclosure It turns out it was a legitimate bug found not a severe vulnerability
2010
01:29:22,300 --> 01:29:25,180
The BitKey and BlockSecurity team rallied.
2011
01:29:25,320 --> 01:29:26,060
Oh, that was last week.
2012
01:29:26,240 --> 01:29:26,720
It's crazy.
2013
01:29:26,880 --> 01:29:27,660
It seems like a month ago.
2014
01:29:27,840 --> 01:29:29,420
Super early Saturday morning.
2015
01:29:29,540 --> 01:29:29,740
Yeah.
2016
01:29:30,020 --> 01:29:33,120
Because, you know, we take it seriously.
2017
01:29:33,640 --> 01:29:35,120
We take all these reports seriously.
2018
01:29:35,620 --> 01:29:36,480
Looked into it.
2019
01:29:36,480 --> 01:29:38,880
Did verify it was a legitimate bug.
2020
01:29:39,800 --> 01:29:40,860
Posted about that.
2021
01:29:40,860 --> 01:29:45,840
And then Saturday afternoon had a spaces where we talked live.
2022
01:29:46,600 --> 01:29:51,220
I mean, I wasn't part of it, but the team about the vulnerability.
2023
01:29:51,560 --> 01:29:53,100
And it is complicated.
2024
01:29:53,260 --> 01:29:54,620
So it's like a little hard to explain.
2025
01:29:54,900 --> 01:30:01,620
But, you know, it's, but the main, so I have a mixed feelings here.
2026
01:30:01,660 --> 01:30:08,480
I mean, like, I'm happy that someone found the bug, did report it to the team so that we could fix it.
2027
01:30:08,480 --> 01:30:11,480
It was a unique bug that was not found yet.
2028
01:30:11,800 --> 01:30:12,680
So that's great.
2029
01:30:13,640 --> 01:30:16,580
But, you know, why the Twitter drama?
2030
01:30:17,380 --> 01:30:21,060
Like, historical responsible disclosure for security bugs
2031
01:30:21,060 --> 01:30:25,900
is that any project or product should have a public process
2032
01:30:25,900 --> 01:30:28,160
for here's how you report a security bug.
2033
01:30:28,340 --> 01:30:29,880
And you should do it privately.
2034
01:30:30,420 --> 01:30:33,980
And to give the team a chance to evaluate it and assess it,
2035
01:30:34,500 --> 01:30:37,000
fix it, and then publish the results.
2036
01:30:37,000 --> 01:30:43,000
And like best practices historically has been give 90 days for that to happen.
2037
01:30:43,160 --> 01:30:44,500
90 days, right?
2038
01:30:45,000 --> 01:30:48,200
But certainly not tweeting three minutes after you report it.
2039
01:30:48,340 --> 01:30:48,800
Yeah, it's kind of.
2040
01:30:49,080 --> 01:30:58,060
Now, I don't know, you know, why the motivations of doing that, but one can speculate, you know, attention seeking or whatever.
2041
01:30:58,060 --> 01:31:05,520
whatever but it doesn't i mean what bothers me is that some people on within the bitcoin community
2042
01:31:05,520 --> 01:31:13,140
rally around that and champion it and applaud that and i think the mentality is well one
2043
01:31:13,140 --> 01:31:19,180
finding a bug is good and i agree finding a bug is good but two it puts pressure on the product
2044
01:31:19,180 --> 01:31:25,420
or project to fix it fast yeah uh or to move fast and i mean i i won't deny that that puts pressure
2045
01:31:25,420 --> 01:31:30,220
on the team to move fast, but a responsible product or project that receives security
2046
01:31:30,220 --> 01:31:37,020
vulnerabilities responsibly also takes them seriously and reacts quickly.
2047
01:31:37,380 --> 01:31:38,640
And most projects do.
2048
01:31:39,060 --> 01:31:43,580
So we need to assume most projects take security vulnerabilities that are responsibly disclosed,
2049
01:31:43,980 --> 01:31:47,560
do take them seriously, act fast, and take the best course of action.
2050
01:31:47,560 --> 01:31:51,500
Because the best course of action is not always going public with it.
2051
01:31:51,500 --> 01:31:54,420
you should move quickly to evaluate
2052
01:31:54,420 --> 01:31:56,940
and then if it's a real vulnerability
2053
01:31:56,940 --> 01:31:59,100
fix it and then get it released
2054
01:31:59,100 --> 01:32:00,780
but keep in mind sometimes the release
2055
01:32:00,780 --> 01:32:02,840
is a server release that's the easiest
2056
01:32:02,840 --> 01:32:04,840
because you're completely in control of that
2057
01:32:04,840 --> 01:32:06,600
as a company or whatever
2058
01:32:06,600 --> 01:32:08,800
you can just like push the release to the server
2059
01:32:08,800 --> 01:32:10,860
everyone automatically gets the
2060
01:32:10,860 --> 01:32:12,160
update
2061
01:32:12,160 --> 01:32:14,760
except for something like cold card
2062
01:32:14,760 --> 01:32:16,760
that's the worst case
2063
01:32:16,760 --> 01:32:19,040
you can push as many software updates
2064
01:32:19,040 --> 01:32:21,020
as you want but you have to regenerate your wallet
2065
01:32:21,020 --> 01:32:22,500
and transfer your funds to get out of that.
2066
01:32:22,580 --> 01:32:23,720
That's the worst case.
2067
01:32:24,080 --> 01:32:24,960
So that's the worst case.
2068
01:32:25,020 --> 01:32:26,460
The best case is a server fix.
2069
01:32:26,980 --> 01:32:29,840
Then there's the next best case would be like a mobile app.
2070
01:32:30,020 --> 01:32:33,760
So you fix the bug and then everyone gets a push notification
2071
01:32:33,760 --> 01:32:35,000
on the phone to update.
2072
01:32:35,820 --> 01:32:38,760
And then you can also, you can incur, depending on the situation,
2073
01:32:38,760 --> 01:32:41,220
sometimes you do want to publicly say,
2074
01:32:41,420 --> 01:32:46,260
please update right away for security.
2075
01:32:46,660 --> 01:32:48,500
Or you can say, update right away for security.
2076
01:32:48,640 --> 01:32:49,220
Here's why.
2077
01:32:49,220 --> 01:32:52,720
You will have funds stolen or you'll have bad things happen to you if you don't do it.
2078
01:32:53,100 --> 01:32:54,100
Or it can be silent.
2079
01:32:55,400 --> 01:33:00,700
And then it could also be like a firmware update to a hardware wallet, which is like even more onerous.
2080
01:33:01,240 --> 01:33:02,660
Harder to get updates.
2081
01:33:03,380 --> 01:33:04,800
Sorry, you're not feeling good, DK.
2082
01:33:05,460 --> 01:33:06,180
Sorry, cough.
2083
01:33:06,240 --> 01:33:06,780
Yeah, no worries.
2084
01:33:07,260 --> 01:33:09,980
But anyway, it's complicated.
2085
01:33:09,980 --> 01:33:15,560
Even a responsible team urgently correcting things,
2086
01:33:15,800 --> 01:33:18,280
it's a complex array of options
2087
01:33:18,280 --> 01:33:21,760
for how to get this safely to their customers.
2088
01:33:22,620 --> 01:33:24,600
So creating a bunch of Twitter drama
2089
01:33:24,600 --> 01:33:27,000
and putting pressure on a team
2090
01:33:27,000 --> 01:33:30,320
might force them to choose a suboptimal path.
2091
01:33:30,540 --> 01:33:32,840
It can put customers in harm's way.
2092
01:33:32,840 --> 01:33:35,400
It can lead to more fund loss.
2093
01:33:35,640 --> 01:33:36,940
So it's a bad idea.
2094
01:33:37,660 --> 01:33:39,220
If the project doesn't respond,
2095
01:33:39,220 --> 01:33:43,560
If they don't respond, then absolutely put the pressure on.
2096
01:33:43,580 --> 01:33:46,940
This makes me think we need some sort of a Noster app
2097
01:33:46,940 --> 01:33:50,620
that is like a dashboard of where do you responsibly disclose?
2098
01:33:51,340 --> 01:33:52,880
What are the response times usually?
2099
01:33:55,320 --> 01:33:56,240
Salud, yeah, sure.
2100
01:33:59,300 --> 01:34:00,340
That's a good idea.
2101
01:34:01,420 --> 01:34:04,760
Keeping metrics of how projects do.
2102
01:34:05,540 --> 01:34:07,880
Probably agents will be the ones disclosing in the future.
2103
01:34:07,880 --> 01:34:19,320
And so if they can just consult some sort of a Nostra-based dashboard or, you know, kind of pull the data from relays, they can know, oh, this one I can report or this one doesn't get addressed for, you know, a year.
2104
01:34:19,440 --> 01:34:23,300
So I'm going to actually make a bunch of Twitter drama to cause it to be seen sooner.
2105
01:34:23,300 --> 01:34:25,480
Or this one gets handled in 48 hours.
2106
01:34:25,580 --> 01:34:27,020
So I'm just not going to create a bunch of drama.
2107
01:34:27,660 --> 01:34:30,120
Even sick DK comes through with the ideas.
2108
01:34:30,560 --> 01:34:30,900
Good job.
2109
01:34:32,860 --> 01:34:34,180
No, that's a good idea.
2110
01:34:34,280 --> 01:34:34,640
I like that.
2111
01:34:34,640 --> 01:34:41,460
But I think we do need to revisit the 90-day best practices in the age of AI.
2112
01:34:41,780 --> 01:34:42,220
90 is too long.
2113
01:34:42,320 --> 01:34:46,440
So I think responsible disclosure best practices need to be revisited.
2114
01:34:47,100 --> 01:34:50,300
But I can guarantee you the answer is not zero seconds.
2115
01:34:50,840 --> 01:34:53,660
And it does more harm than good.
2116
01:34:54,220 --> 01:34:58,440
And I don't know if it will change the behavior of certain individuals doing this,
2117
01:34:58,440 --> 01:35:03,700
but I do at least hope to inform and educate the broader folks who champion that
2118
01:35:03,700 --> 01:35:05,020
to understand and appreciate
2119
01:35:05,020 --> 01:35:07,480
why that actually can be harmful to people.
2120
01:35:08,480 --> 01:35:11,960
So that's, I wanted to get that out there.
2121
01:35:12,200 --> 01:35:13,760
I have one other big question for you.
2122
01:35:13,820 --> 01:35:15,940
I don't know, maybe it's too big of a topic
2123
01:35:15,940 --> 01:35:16,940
to tackle at the end
2124
01:35:16,940 --> 01:35:19,140
and how much you could even say
2125
01:35:19,140 --> 01:35:21,060
with all of the hacks and everything going on,
2126
01:35:21,380 --> 01:35:23,300
noticing some lightning nodes coming down.
2127
01:35:23,560 --> 01:35:24,200
What is your thought?
2128
01:35:24,540 --> 01:35:26,500
What can we talk about with the state of the lightning?
2129
01:35:27,820 --> 01:35:29,640
Well, I mean, I'll start really general.
2130
01:35:29,760 --> 01:35:30,380
Like I said earlier,
2131
01:35:30,560 --> 01:35:32,740
I think every software project in the world
2132
01:35:32,740 --> 01:35:33,880
has security vulnerabilities.
2133
01:35:34,200 --> 01:35:37,480
So inclusive of that is Lightning implementations.
2134
01:35:38,720 --> 01:35:42,600
And Project Loop's stance on communications
2135
01:35:42,600 --> 01:35:47,100
is we're not announcing all the projects
2136
01:35:47,100 --> 01:35:48,780
that we're working with and we're scanning.
2137
01:35:50,200 --> 01:35:52,400
But, I mean, let's just say that I think
2138
01:35:52,400 --> 01:35:55,040
it's smart to update your Lightning software.
2139
01:35:55,360 --> 01:35:56,760
Whichever implementation you're using,
2140
01:35:56,980 --> 01:35:58,600
it's smart to update it.
2141
01:35:58,920 --> 01:35:59,680
I mean, that's true.
2142
01:35:59,880 --> 01:36:01,200
That's always been true.
2143
01:36:01,200 --> 01:36:07,540
it's not new just because of the past week or even ai um do you think we'll see a lot more people
2144
01:36:07,540 --> 01:36:11,920
though just like go down for like whatever a month two months three months like while this
2145
01:36:11,920 --> 01:36:15,740
whole thing comes back and then build later or do you think i don't know because i i do think for
2146
01:36:15,740 --> 01:36:19,140
like people that are operating hot wallets through all this that's going to be tough
2147
01:36:19,140 --> 01:36:27,060
yeah i mean you know each person or each company is going to have to evaluate that situation for
2148
01:36:27,060 --> 01:36:31,400
themselves. I don't think it's, you know, like, I don't have some blanket advice, like, shut it
2149
01:36:31,400 --> 01:36:36,980
down or don't shut it down. You have to, like, I think in some cases, some services have paused.
2150
01:36:37,820 --> 01:36:42,740
And I think they just sort of concluded, like, they're having money stolen from them.
2151
01:36:43,300 --> 01:36:48,200
They're vulnerable. The business they're in, they believe in the future of their business,
2152
01:36:48,200 --> 01:36:53,220
but the business they're in is not, like, minting money. So it's not a good risk-reward situation.
2153
01:36:53,220 --> 01:36:55,020
so I'm not going to
2154
01:36:55,020 --> 01:36:57,580
certainly not going to criticize that
2155
01:36:57,580 --> 01:36:57,940
yeah
2156
01:36:57,940 --> 01:37:00,800
we didn't even talk about bold swaps
2157
01:37:00,800 --> 01:37:02,920
which is the layer 1, layer 2
2158
01:37:02,920 --> 01:37:04,600
sort of exchange basically being down
2159
01:37:04,600 --> 01:37:07,240
yeah, we're tying for all that
2160
01:37:07,240 --> 01:37:09,620
let's see, hold on
2161
01:37:09,620 --> 01:37:11,860
I want to drop a little bit more alpha
2162
01:37:11,860 --> 01:37:12,580
oh
2163
01:37:12,580 --> 01:37:14,660
gotta get more alpha
2164
01:37:14,660 --> 01:37:17,400
oh, I've got a question
2165
01:37:17,400 --> 01:37:17,980
I can leave us
2166
01:37:17,980 --> 01:37:19,520
in white hat
2167
01:37:19,520 --> 01:37:21,500
no, I want to talk a little bit about white hat hacking
2168
01:37:21,500 --> 01:37:23,540
and then a spiral alpha drop.
2169
01:37:23,640 --> 01:37:24,380
But go ahead, DK.
2170
01:37:25,340 --> 01:37:31,900
No, I've kind of related to the kind of moltbook process
2171
01:37:31,900 --> 01:37:35,120
that we saw, sort of that emergent message board arrive.
2172
01:37:35,220 --> 01:37:37,640
It kind of got me thinking about this question
2173
01:37:37,640 --> 01:37:39,100
we've discussed in the past.
2174
01:37:39,200 --> 01:37:41,680
And I don't mean for us to necessarily discuss all this right now,
2175
01:37:41,720 --> 01:37:43,820
but it's kind of a question to ponder and think about
2176
01:37:43,820 --> 01:37:45,400
and maybe do your own LLM research.
2177
01:37:46,060 --> 01:37:47,540
But we've talked about like,
2178
01:37:47,540 --> 01:37:54,320
is there something like Bitcoin that could be even better in some way because it does payments
2179
01:37:54,320 --> 01:38:02,160
better? Or might agents emergently come up with their own currency that just turns out to kind of
2180
01:38:02,160 --> 01:38:07,400
work and is good enough? And I feel like the example that we saw on the message board suggests
2181
01:38:07,400 --> 01:38:13,920
that there may be something like that. I don't believe that exists yet, but it seems like there's
2182
01:38:13,920 --> 01:38:18,860
the kind of credible, neutral possibility that could come out from agents that are just
2183
01:38:18,860 --> 01:38:19,540
trying to...
2184
01:38:19,540 --> 01:38:21,300
What specifically in the message?
2185
01:38:21,860 --> 01:38:25,640
Oh, just the message boards, like how they kind of emergently created message boards
2186
01:38:25,640 --> 01:38:32,360
out of file names to try to communicate maybe a resource allocation, like agents trying
2187
01:38:32,360 --> 01:38:35,320
to allocate resources and they're like, I don't know, we can't do anything with what
2188
01:38:35,320 --> 01:38:35,820
we've got.
2189
01:38:36,200 --> 01:38:38,320
They might make up their own way to allocate resources.
2190
01:38:38,360 --> 01:38:40,340
Well, and this, I mean, we talked about many months ago.
2191
01:38:40,400 --> 01:38:43,540
I do think this is one of the most important long-term questions.
2192
01:38:43,920 --> 01:38:48,620
You know, I think that my sort of, I mean, who knows?
2193
01:38:48,940 --> 01:38:52,220
My base case is that they're going to end up adopting Bitcoin
2194
01:38:52,220 --> 01:38:53,160
because it's already there.
2195
01:38:53,300 --> 01:38:53,761
It's already energy.
2196
01:38:53,840 --> 01:38:54,540
There's all these things.
2197
01:38:55,180 --> 01:38:58,880
And I think, by the way, I expect to see more of these message boards,
2198
01:38:58,960 --> 01:39:00,880
for example, become potentially Nostra-based.
2199
01:39:00,980 --> 01:39:02,740
Like what is Nostra but a bulletin board, right?
2200
01:39:02,860 --> 01:39:02,980
Yeah.
2201
01:39:03,080 --> 01:39:03,660
Or a relay.
2202
01:39:04,580 --> 01:39:06,920
But I do think it's important to have the intellectual humility
2203
01:39:06,920 --> 01:39:07,720
to say we just don't know.
2204
01:39:07,820 --> 01:39:10,120
And so is there a chance the agents come up with their own thing?
2205
01:39:10,240 --> 01:39:10,860
I mean, maybe.
2206
01:39:10,860 --> 01:39:13,080
And I'm seeing more people talk about, you know,
2207
01:39:13,080 --> 01:39:16,400
how agents might be able to start trading compute credits, things like that.
2208
01:39:16,780 --> 01:39:19,900
And maybe there's markets that, you know, exchange those for Bitcoin.
2209
01:39:20,060 --> 01:39:22,780
I think that's a very interesting potential startup opportunity, actually.
2210
01:39:23,260 --> 01:39:26,780
So I'd be very interested in hearing from people that are thinking about those questions
2211
01:39:26,780 --> 01:39:28,520
of how the agents will end up transacting.
2212
01:39:29,100 --> 01:39:33,120
Yeah, I just thought it was interesting to ponder because like you could have kind of
2213
01:39:33,120 --> 01:39:35,960
non-economic like markets that emerge.
2214
01:39:36,020 --> 01:39:38,940
Like there's a lot of agents that don't have access to Bitcoin today, but do have access
2215
01:39:38,940 --> 01:39:42,080
to intelligence and tokens and stuff like that.
2216
01:39:42,080 --> 01:39:46,880
So, like, non-monetary markets could emerge is all I was thinking.
2217
01:39:47,560 --> 01:39:49,580
So, Deanna has a comment.
2218
01:39:49,940 --> 01:39:52,720
Steve, don't you think the red team is introducing similar concerns?
2219
01:39:52,960 --> 01:39:58,180
You said they are great marketers, and I think that's in reference to irresponsible disclosure.
2220
01:39:58,520 --> 01:39:59,360
That's a great question.
2221
01:40:00,400 --> 01:40:00,900
Let's see.
2222
01:40:00,900 --> 01:40:10,160
I think if they're saying, hey, red alert, there's a lot of bugs and security vulnerabilities.
2223
01:40:10,160 --> 01:40:13,220
everyone pay attention and like let's get to work
2224
01:40:13,220 --> 01:40:16,420
I mean I'm kind of trying to say the same thing on this show
2225
01:40:16,420 --> 01:40:20,080
I agree with alerting folks to the general issue
2226
01:40:20,080 --> 01:40:23,800
if they're reporting saying like
2227
01:40:23,800 --> 01:40:24,960
a specific project
2228
01:40:24,960 --> 01:40:28,040
total like 6,000 vulnerabilities found overall
2229
01:40:28,040 --> 01:40:29,160
I think that's fine
2230
01:40:29,160 --> 01:40:32,220
Project Loop also reported a total amount
2231
01:40:32,220 --> 01:40:35,700
we did seek out the projects we worked on
2232
01:40:35,700 --> 01:40:39,160
feedback on whether we could report specific numbers for each project
2233
01:40:39,160 --> 01:40:41,560
and concluded not to do that.
2234
01:40:42,600 --> 01:40:43,560
Like, keep it aggregate.
2235
01:40:44,320 --> 01:40:45,380
So that was our decision.
2236
01:40:45,500 --> 01:40:47,400
I don't know what the red team folks are doing.
2237
01:40:50,020 --> 01:40:53,480
Yeah, I haven't seen them veer into, like,
2238
01:40:54,040 --> 01:40:55,920
saying, you know, specific projects,
2239
01:40:56,060 --> 01:40:56,980
specific vulnerabilities.
2240
01:40:58,800 --> 01:41:00,520
That would maybe cross the line.
2241
01:41:01,220 --> 01:41:02,120
I do think this is, like,
2242
01:41:03,320 --> 01:41:04,900
I mean, it is a special period
2243
01:41:04,900 --> 01:41:07,020
where moving fast and communicating fast
2244
01:41:07,020 --> 01:41:08,620
is generally appropriate.
2245
01:41:09,160 --> 01:41:17,600
But I still don't think, I think people need to understand the costs of creating hysteria.
2246
01:41:17,600 --> 01:41:24,480
And if you provide, if you reveal too much specific information about a bug, that can
2247
01:41:24,480 --> 01:41:26,600
be very dangerous for people.
2248
01:41:27,940 --> 01:41:31,540
And one other thing, I think we've talked about this a lot on the show internally.
2249
01:41:31,800 --> 01:41:38,500
I think the other big takeaway for me is just like, we've spent so much time as a community
2250
01:41:38,500 --> 01:41:41,080
preparing for theoretical risks.
2251
01:41:41,440 --> 01:41:43,440
I mean, I'm not saying that quantum's not going to happen,
2252
01:41:43,960 --> 01:41:47,360
but I also think it's just a really important wake-up
2253
01:41:47,360 --> 01:41:50,780
for everyone to realize AI is here now.
2254
01:41:51,620 --> 01:41:53,960
Yeah, it's a more near-term, actual, practical risk
2255
01:41:53,960 --> 01:41:55,680
that's very, very real.
2256
01:41:55,880 --> 01:41:57,220
And it kind of reminds me a little bit,
2257
01:41:57,220 --> 01:41:58,180
not saying we shouldn't be preparing,
2258
01:41:58,280 --> 01:42:00,160
because I think the AI will help quantum computers
2259
01:42:00,160 --> 01:42:01,500
happen faster with their research,
2260
01:42:01,580 --> 01:42:02,520
and quantum's just one example.
2261
01:42:02,640 --> 01:42:03,940
It's actually a pretty real example.
2262
01:42:04,260 --> 01:42:07,360
But I think about this a little bit
2263
01:42:07,360 --> 01:42:09,280
in some of the like, you know,
2264
01:42:09,320 --> 01:42:11,660
the cold card hack and self-custody and stuff
2265
01:42:11,660 --> 01:42:13,080
where it's like people are like really
2266
01:42:13,080 --> 01:42:14,740
in the Bitcoin community focus on like
2267
01:42:14,740 --> 01:42:16,280
the extreme edge cases
2268
01:42:16,280 --> 01:42:18,740
where it's like often like the most basic thing
2269
01:42:18,740 --> 01:42:19,660
is the thing that's like,
2270
01:42:20,200 --> 01:42:21,920
like, hey, is the seed have enough entropy?
2271
01:42:22,020 --> 01:42:23,080
Or like, hey, like is your,
2272
01:42:23,200 --> 01:42:24,960
is your whatever database going to get hacked?
2273
01:42:25,500 --> 01:42:27,940
So I just think maybe it's also a good kind of like
2274
01:42:27,940 --> 01:42:29,820
refresh and reset to focus on like
2275
01:42:29,820 --> 01:42:30,820
the absolute fundamentals.
2276
01:42:30,920 --> 01:42:33,440
And to me right now, like AI is 120% of that,
2277
01:42:33,440 --> 01:42:35,160
which I know you're pushing for people
2278
01:42:35,160 --> 01:42:36,820
all over the world to get that message,
2279
01:42:36,820 --> 01:42:38,540
but it seems like not everyone has yet.
2280
01:42:39,600 --> 01:42:40,160
All right.
2281
01:42:40,740 --> 01:42:41,860
We're going long today.
2282
01:42:42,040 --> 01:42:43,240
Hopefully you guys are okay with that.
2283
01:42:43,340 --> 01:42:44,500
I know you're already at 142.
2284
01:42:45,220 --> 01:42:45,620
Okay.
2285
01:42:45,780 --> 01:42:49,040
So I wanted to touch on white hat hacking philosophy,
2286
01:42:49,900 --> 01:42:51,200
process, and challenges.
2287
01:42:51,500 --> 01:42:51,640
Okay.
2288
01:42:52,320 --> 01:42:56,380
So I think it was last Friday.
2289
01:42:58,240 --> 01:42:59,780
You know, there was like thinking about,
2290
01:42:59,840 --> 01:43:02,620
oh, well, you know, can we catch the bad guy?
2291
01:43:02,680 --> 01:43:03,580
Can we catch criminals?
2292
01:43:04,940 --> 01:43:06,240
Reclaim the Bitcoin.
2293
01:43:06,820 --> 01:43:09,640
and then give it back to the rightful owners.
2294
01:43:09,960 --> 01:43:10,840
And then that raised questions,
2295
01:43:10,980 --> 01:43:12,460
how do you know who the rightful owners are?
2296
01:43:13,080 --> 01:43:15,180
And we talked about that on the show last week.
2297
01:43:15,360 --> 01:43:17,520
I did want to make one correction or addition
2298
01:43:17,520 --> 01:43:18,680
that I learned since then.
2299
01:43:18,860 --> 01:43:21,420
We talked about the UID or unique ID
2300
01:43:21,420 --> 01:43:23,880
that's in the hardware product
2301
01:43:23,880 --> 01:43:26,220
as to helping identify the rightful owner.
2302
01:43:26,800 --> 01:43:29,180
That is a weaker piece of evidence
2303
01:43:29,180 --> 01:43:31,300
than I had mentioned last week.
2304
01:43:31,940 --> 01:43:33,100
It turns out the UID,
2305
01:43:33,100 --> 01:43:36,700
it's a 96-bit number on the chip.
2306
01:43:36,820 --> 01:43:52,880
That is unique. So every chip that gets produced, it has a 96-bit number, which is unique. But Coldcard only used 32 bits of the 96 bits in the generation of the key.
2307
01:43:52,880 --> 01:43:57,040
Even 32 bits though, it's a pretty big search space.
2308
01:43:57,040 --> 01:44:20,066
The odds of two hardware products being created with the same 32 bits that are used for entropy is low However that 32 bits they chose it turns out is way weaker than 32 bits implies It like the XY coordinate on the wafer and so it like 100 by 100 integers So it more like
2309
01:44:20,066 --> 01:44:29,846
1 in 10,000. So you can do some math there to realize there's probably some cold cards,
2310
01:44:29,846 --> 01:44:32,886
where there's two that have the same UID
2311
01:44:32,886 --> 01:44:34,686
that's used for the entropy.
2312
01:44:35,326 --> 01:44:38,126
So it's not perfect, but it's still, you know,
2313
01:44:38,346 --> 01:44:40,426
whatever, something like a 1 in 10,000 chance
2314
01:44:40,426 --> 01:44:42,666
that my cold card matches your cold card.
2315
01:44:42,746 --> 01:44:44,186
If you combine additional evidence,
2316
01:44:44,366 --> 01:44:45,826
it's still a pretty strong solution there,
2317
01:44:45,866 --> 01:44:47,366
but I just wanted to note that correction.
2318
01:44:48,926 --> 01:44:50,966
But then, okay, so it's one thing if, like,
2319
01:44:50,966 --> 01:44:54,986
you know, the feds catch the guy, get the money back,
2320
01:44:55,086 --> 01:44:56,986
and then I think, so I still believe
2321
01:44:56,986 --> 01:44:58,206
we have a pretty strong process
2322
01:44:58,206 --> 01:45:00,986
to get the coins back to the rightful owners.
2323
01:45:01,666 --> 01:45:03,606
But then it raises a question about,
2324
01:45:03,606 --> 01:45:09,406
okay, if only a portion of the stolen coins are reclaimed,
2325
01:45:09,686 --> 01:45:12,686
then how are they redistributed?
2326
01:45:13,046 --> 01:45:14,706
Is it, do you map it one,
2327
01:45:14,786 --> 01:45:19,306
because you can map the coins back to one-to-one to each wallet.
2328
01:45:19,606 --> 01:45:24,546
So you could make whole 100% for the customers of the coins you got back,
2329
01:45:24,646 --> 01:45:26,506
but then other customers would get 0% back.
2330
01:45:26,586 --> 01:45:27,526
So is that fair?
2331
01:45:27,526 --> 01:45:32,226
or is it, you know, everyone gets a haircut?
2332
01:45:32,926 --> 01:45:33,566
Is that fair?
2333
01:45:34,146 --> 01:45:36,186
And I don't think there's like an objective right answer
2334
01:45:36,186 --> 01:45:36,926
to that question,
2335
01:45:37,086 --> 01:45:39,106
but that's an interesting discussion and debate
2336
01:45:39,106 --> 01:45:40,566
and philosophy around that.
2337
01:45:41,386 --> 01:45:43,286
Then there is the other approach,
2338
01:45:43,466 --> 01:45:44,846
which is white hat hacking,
2339
01:45:45,106 --> 01:45:46,906
which I've seen a lot about on Twitter
2340
01:45:46,906 --> 01:45:48,866
and apparently there's instances of this.
2341
01:45:49,306 --> 01:45:50,346
This is where someone,
2342
01:45:51,446 --> 01:45:53,046
I mean, by definition, white hat,
2343
01:45:53,246 --> 01:45:54,846
someone well-intentioned
2344
01:45:54,846 --> 01:45:57,806
who is equipped to take the coins.
2345
01:45:57,906 --> 01:45:59,706
They have the knowledge, the compute power,
2346
01:45:59,846 --> 01:46:03,546
the ability to take someone's rightful owner's
2347
01:46:03,546 --> 01:46:07,046
coins out of their wallet for safekeeping
2348
01:46:07,046 --> 01:46:10,566
so that they can later give them back
2349
01:46:10,566 --> 01:46:13,546
to that rightful owner and beat a black hat,
2350
01:46:13,926 --> 01:46:15,686
a criminal, to the case.
2351
01:46:15,686 --> 01:46:20,266
Now, number one, I don't know what the legal history is here
2352
01:46:20,266 --> 01:46:20,866
and precedent.
2353
01:46:21,106 --> 01:46:23,326
I have no idea about the legality of it.
2354
01:46:24,846 --> 01:46:26,386
but the,
2355
01:46:27,146 --> 01:46:27,366
but,
2356
01:46:27,426 --> 01:46:27,686
but two,
2357
01:46:27,746 --> 01:46:28,826
I was still interested in this.
2358
01:46:28,866 --> 01:46:29,026
Like,
2359
01:46:29,186 --> 01:46:29,426
I mean,
2360
01:46:29,446 --> 01:46:29,666
it's,
2361
01:46:29,666 --> 01:46:30,586
it sounds what,
2362
01:46:30,686 --> 01:46:30,886
I mean,
2363
01:46:30,906 --> 01:46:32,726
it sounds well-intentioned.
2364
01:46:32,866 --> 01:46:33,146
Um,
2365
01:46:33,246 --> 01:46:34,426
but I saw some instances,
2366
01:46:34,426 --> 01:46:35,486
people are saying like,
2367
01:46:35,506 --> 01:46:35,606
yeah,
2368
01:46:35,606 --> 01:46:36,066
they did it.
2369
01:46:36,066 --> 01:46:36,806
And then people would ask,
2370
01:46:36,926 --> 01:46:38,146
what's the process for getting it back?
2371
01:46:38,166 --> 01:46:38,466
And they're like,
2372
01:46:38,506 --> 01:46:39,386
we'll figure that out later.
2373
01:46:39,706 --> 01:46:41,566
I think that's an insane approach.
2374
01:46:42,826 --> 01:46:45,446
If you don't know how you're going to get the coins back to the rightful owner,
2375
01:46:45,986 --> 01:46:46,626
uh,
2376
01:46:46,626 --> 01:46:52,626
that seems crazy to me that you would take the coins and then not really know how you're going to get it back and answer some of these other questions.
2377
01:46:53,566 --> 01:46:53,886
Um,
2378
01:46:53,886 --> 01:47:02,166
But also, I had learned from others on the thinking in this, I've actually sort of come around to thinking it's not an ethical approach.
2379
01:47:02,166 --> 01:47:04,466
Because I think it's a violation of property rights.
2380
01:47:05,286 --> 01:47:09,006
Let's say that you were a cold card owner and you had Bitcoin on there.
2381
01:47:09,386 --> 01:47:10,906
And then I'm a white hat attacker.
2382
01:47:11,306 --> 01:47:12,866
And I think I'm being well-intentioned.
2383
01:47:12,926 --> 01:47:14,386
And I take your coin.
2384
01:47:14,426 --> 01:47:16,606
I don't know you, but I take the coins out of your wallet.
2385
01:47:17,226 --> 01:47:23,026
What if you were about to transfer your coins to a safer wallet?
2386
01:47:23,026 --> 01:47:25,646
I've taken that right from you
2387
01:47:25,646 --> 01:47:27,126
you can't do it
2388
01:47:27,126 --> 01:47:28,426
because you don't have your Bitcoin anymore
2389
01:47:28,426 --> 01:47:30,286
what if
2390
01:47:30,286 --> 01:47:31,946
it take
2391
01:47:31,946 --> 01:47:33,546
what if the end result is
2392
01:47:33,546 --> 01:47:34,686
I can't figure out
2393
01:47:34,686 --> 01:47:36,266
how to get the coins back to you
2394
01:47:36,266 --> 01:47:37,646
then you never get your Bitcoin back
2395
01:47:37,646 --> 01:47:38,726
what if
2396
01:47:38,726 --> 01:47:40,186
it somehow gets
2397
01:47:40,186 --> 01:47:40,946
I get like
2398
01:47:40,946 --> 01:47:41,806
you know
2399
01:47:41,806 --> 01:47:43,206
like the feds
2400
01:47:43,206 --> 01:47:44,066
take the coins from me
2401
01:47:44,066 --> 01:47:44,346
or whatever
2402
01:47:44,346 --> 01:47:45,386
and then it gets wrapped up
2403
01:47:45,386 --> 01:47:46,846
into a multi-year process
2404
01:47:46,846 --> 01:47:48,266
and then you
2405
01:47:48,266 --> 01:47:49,406
and then you get your
2406
01:47:49,406 --> 01:47:50,466
you do get your coins back
2407
01:47:50,466 --> 01:47:51,826
but it's after many years
2408
01:47:51,826 --> 01:47:52,846
maybe there's an additional
2409
01:47:52,846 --> 01:47:54,966
haircut, again, you were robbed of the chance to move
2410
01:47:54,966 --> 01:47:57,106
your coins. So just that alone
2411
01:47:57,106 --> 01:47:58,646
kind of, I think, does
2412
01:47:58,646 --> 01:48:00,566
feel like it violates property rights
2413
01:48:00,566 --> 01:48:02,946
to make it actually the wrong ethical
2414
01:48:02,946 --> 01:48:04,746
thing to do, irrespective
2415
01:48:04,746 --> 01:48:06,846
of the legal
2416
01:48:06,846 --> 01:48:08,786
situation, the process, etc.
2417
01:48:08,866 --> 01:48:10,646
Even if all that was in the clear.
2418
01:48:10,926 --> 01:48:12,966
So even, like, I was kind of like,
2419
01:48:13,026 --> 01:48:14,686
well, maybe if the U.S. government did it,
2420
01:48:14,786 --> 01:48:16,846
then it would be more
2421
01:48:16,846 --> 01:48:18,866
appropriate. But even then, I think it
2422
01:48:18,866 --> 01:48:20,986
violates property rights. What are your guys' thoughts on that?
2423
01:48:22,146 --> 01:48:22,706
Yeah, I
2424
01:48:22,846 --> 01:48:28,526
I would generally agree with what you said, though it does open this question that I've
2425
01:48:28,526 --> 01:48:32,766
asked you and you said maybe we could discuss the legal theory on property rights.
2426
01:48:33,406 --> 01:48:39,766
But I don't know that I have a strong sense of how property rights should be assessed,
2427
01:48:39,886 --> 01:48:45,266
like some sort of case law or historic stuff in the physical world with how property rights
2428
01:48:45,266 --> 01:48:46,566
should be assessed on something like this.
2429
01:48:46,566 --> 01:48:52,406
Let's say we're on public land and Max hid a bar of gold on public land.
2430
01:48:52,406 --> 01:48:55,486
and then later I walk by and discover that barb gold
2431
01:48:55,486 --> 01:48:56,326
and I pick it up.
2432
01:48:56,686 --> 01:48:57,766
Did I steal that from him?
2433
01:48:57,866 --> 01:49:01,286
I don't know if there's precedent that says I did or didn't,
2434
01:49:01,346 --> 01:49:03,006
but how would that work?
2435
01:49:03,066 --> 01:49:04,226
That's kind of what's happening here, right?
2436
01:49:04,666 --> 01:49:07,206
Yeah, I mean, I also think that the theory-
2437
01:49:07,206 --> 01:49:08,466
I don't think that's a good analogy, but-
2438
01:49:08,466 --> 01:49:10,066
Well, the theory of Bitcoin is that
2439
01:49:10,066 --> 01:49:12,386
possession of the private key is property rights, right?
2440
01:49:12,446 --> 01:49:16,226
Now, obviously, in a world where you didn't have this exploit,
2441
01:49:17,806 --> 01:49:19,306
whoever owns the private key, that's the property.
2442
01:49:19,306 --> 01:49:22,066
That's the theory of the system working.
2443
01:49:22,066 --> 01:49:34,666
Now, the other step we're talking about is like layering more traditional legal systems on top of that, which, I mean, obviously, anytime you, the whole point of Bitcoin, right, was to, this question of ownership is to get away from having these courts and all that kind of thing.
2444
01:49:35,326 --> 01:49:41,866
Given that that's where we are, I mean, to me, I mean, first of all, I think the idea of just like hacking and be like, I don't know what I'm going to do with this sounds pretty crazy.
2445
01:49:42,426 --> 01:49:45,926
I think the only move that I could see working there
2446
01:49:45,926 --> 01:49:47,726
is like you said, either the federal government doing it
2447
01:49:47,726 --> 01:49:49,826
or if your answer was, I'm taking this
2448
01:49:49,826 --> 01:49:51,226
and just giving it to the federal government.
2449
01:49:51,826 --> 01:49:53,366
And obviously, I know Bitcoiners
2450
01:49:53,366 --> 01:49:54,226
don't necessarily love the government,
2451
01:49:54,406 --> 01:49:56,746
but they're probably in the best position
2452
01:49:56,746 --> 01:49:57,966
to try and figure that out in the future.
2453
01:49:58,366 --> 01:50:00,626
I get your point that it robs you of the chance to move it.
2454
01:50:01,006 --> 01:50:03,566
I mean, there's probably some timeline, right?
2455
01:50:03,626 --> 01:50:06,306
Like maybe in the first 48 hours, that was really bad.
2456
01:50:06,726 --> 01:50:08,246
If you haven't moved it a week on
2457
01:50:08,246 --> 01:50:09,126
and it's still vulnerable,
2458
01:50:09,286 --> 01:50:11,066
at some point, it's a trade-off.
2459
01:50:11,866 --> 01:50:15,346
the same philosophy will exist
2460
01:50:15,346 --> 01:50:17,666
if quantum ever becomes a thing, right?
2461
01:50:17,706 --> 01:50:19,026
And we've talked about that before.
2462
01:50:19,026 --> 01:50:20,746
Yeah, and my view on quantum was,
2463
01:50:20,966 --> 01:50:24,406
and this is loosely held, right?
2464
01:50:24,526 --> 01:50:27,086
I mean, again, the system works.
2465
01:50:27,206 --> 01:50:28,366
It's kind of a failure of the system, right?
2466
01:50:28,366 --> 01:50:29,226
Because the system works
2467
01:50:29,226 --> 01:50:31,626
when you don't have to go into these edge cases.
2468
01:50:31,746 --> 01:50:32,806
One difference between quantum,
2469
01:50:33,266 --> 01:50:34,846
if that ever occurs,
2470
01:50:34,986 --> 01:50:37,026
then the Bitcoin network is failing.
2471
01:50:37,646 --> 01:50:39,306
Here, a product was failing.
2472
01:50:39,466 --> 01:50:40,966
So I do think that's an important distinction.
2473
01:50:41,186 --> 01:50:41,286
That's true.
2474
01:50:41,286 --> 01:50:46,926
meaning you might arrive at different conclusions on the two because of that difference yeah but i
2475
01:50:46,926 --> 01:50:53,486
mean but i mean at the end of the day um yeah i mean i i do think i was in favor and still probably
2476
01:50:53,486 --> 01:51:01,166
am on the network question of quantum of you know um don't mess with the protocol because that would
2477
01:51:01,166 --> 01:51:04,726
invalidate everything future that's like the ethereum dow fork but again to your point here
2478
01:51:04,726 --> 01:51:08,366
it is different because it's a specific product i mean i think the answer is there is no great
2479
01:51:08,366 --> 01:51:10,406
answer for this. I do think
2480
01:51:10,406 --> 01:51:12,006
the longer it goes with
2481
01:51:12,006 --> 01:51:14,366
moving that, the less
2482
01:51:14,366 --> 01:51:16,426
the fact
2483
01:51:16,426 --> 01:51:18,566
that they should have the right to move that in the short term is a concern.
2484
01:51:19,166 --> 01:51:20,526
So maybe the government actually
2485
01:51:20,526 --> 01:51:22,546
would have a pretty good chance in the system
2486
01:51:22,546 --> 01:51:24,606
and you'd rather get your coins back to them four years later
2487
01:51:24,606 --> 01:51:25,546
than lose everything.
2488
01:51:25,846 --> 01:51:28,206
How do you even know if somebody registers as white hat?
2489
01:51:28,506 --> 01:51:30,106
They claim their white hat, but how do you know?
2490
01:51:30,406 --> 01:51:31,006
Yeah, exactly.
2491
01:51:31,606 --> 01:51:33,966
They could be black hat and then they could become white hat.
2492
01:51:33,966 --> 01:51:35,966
And it probably should be the state
2493
01:51:35,966 --> 01:51:37,426
at that point. That's actually a good point.
2494
01:51:37,426 --> 01:51:40,266
Because obviously, if you claim your white hat
2495
01:51:40,266 --> 01:51:41,646
and then you give the money back,
2496
01:51:41,806 --> 01:51:43,126
you can say, well, there's the proof.
2497
01:51:43,966 --> 01:51:48,346
But yeah, you can claim your white hat,
2498
01:51:49,126 --> 01:51:52,846
take the coins, actually intend to keep them,
2499
01:51:52,926 --> 01:51:54,806
and then if you change your mind later and give them back,
2500
01:51:54,886 --> 01:51:56,086
then just be like, yeah, it's a white hat.
2501
01:51:56,686 --> 01:52:00,066
That gets murky as well.
2502
01:52:00,886 --> 01:52:01,926
Some comments here.
2503
01:52:02,066 --> 01:52:02,826
This is a great discussion.
2504
01:52:02,826 --> 01:52:04,846
This is clearly one where it's not like
2505
01:52:04,846 --> 01:52:07,386
there's an objective answer, like we have the fact,
2506
01:52:07,426 --> 01:52:08,106
we move on.
2507
01:52:08,106 --> 01:52:08,446
Yeah, that's why
2508
01:52:08,446 --> 01:52:09,106
it kind of had
2509
01:52:09,106 --> 01:52:10,166
a good decision.
2510
01:52:10,266 --> 01:52:10,926
So some people's
2511
01:52:10,926 --> 01:52:11,426
thoughts here,
2512
01:52:12,246 --> 01:52:12,446
we'll see,
2513
01:52:12,526 --> 01:52:12,866
Vinny says,
2514
01:52:12,946 --> 01:52:13,606
before you take off,
2515
01:52:13,646 --> 01:52:14,306
you have to be confident
2516
01:52:14,306 --> 01:52:15,106
you can land.
2517
01:52:15,626 --> 01:52:16,546
I think that's to the,
2518
01:52:16,966 --> 01:52:18,286
you'd be good to have
2519
01:52:18,286 --> 01:52:19,726
a plan after you get
2520
01:52:19,726 --> 01:52:20,066
the coins.
2521
01:52:20,886 --> 01:52:22,426
And then one's
2522
01:52:22,426 --> 01:52:22,806
a compliment.
2523
01:52:23,046 --> 01:52:24,506
So guessing a number
2524
01:52:24,506 --> 01:52:25,946
is not a violation
2525
01:52:25,946 --> 01:52:26,746
of property rates.
2526
01:52:27,106 --> 01:52:28,386
You can't own a number.
2527
01:52:29,226 --> 01:52:30,826
So I definitely need
2528
01:52:30,826 --> 01:52:31,526
to think about that more
2529
01:52:31,526 --> 01:52:32,106
and discuss that,
2530
01:52:32,166 --> 01:52:32,966
but that's maybe
2531
01:52:32,966 --> 01:52:34,466
an interesting angle
2532
01:52:34,466 --> 01:52:36,166
to evaluate this
2533
01:52:36,166 --> 01:52:36,946
philosophy.
2534
01:52:37,426 --> 01:52:42,926
um yeah anyway i think that's kind of what i'm i'm i would agree with that and that's kind of what i
2535
01:52:42,926 --> 01:52:48,926
that's sort of the general position that i would take is like if if you got it if you found the
2536
01:52:48,926 --> 01:52:53,186
bar of gold on the side like it's it's yours i'm not saying it's good that that happened but
2537
01:52:53,186 --> 01:52:59,926
but here let me poke at that analogy a little bit um if i take a bar of gold and and bury it in
2538
01:52:59,926 --> 01:53:01,246
public land,
2539
01:53:01,886 --> 01:53:02,846
that is different.
2540
01:53:03,466 --> 01:53:05,246
I don't own that land.
2541
01:53:06,866 --> 01:53:07,986
It's public land.
2542
01:53:08,326 --> 01:53:09,466
So you can kind of argue like
2543
01:53:09,466 --> 01:53:11,826
it's public. Either the
2544
01:53:11,826 --> 01:53:13,686
government, like if it's like a city
2545
01:53:13,686 --> 01:53:15,926
or something, this is federal land.
2546
01:53:16,406 --> 01:53:17,806
So if I buried it right out here,
2547
01:53:18,386 --> 01:53:19,466
then like, okay,
2548
01:53:19,666 --> 01:53:21,366
it's either anyone's game or
2549
01:53:21,366 --> 01:53:23,666
the government owns it. I don't even know what the
2550
01:53:23,666 --> 01:53:25,706
legal precedent on that is. But I do
2551
01:53:25,706 --> 01:53:26,746
think it's different than if it's
2552
01:53:26,746 --> 01:53:28,786
on the
2553
01:53:28,786 --> 01:53:30,546
the public blockchain,
2554
01:53:31,146 --> 01:53:32,966
but it was in my private ownership.
2555
01:53:33,086 --> 01:53:33,746
I think that's different.
2556
01:53:34,126 --> 01:53:35,966
I don't know if it's the right comparison, though,
2557
01:53:35,966 --> 01:53:37,766
to say you don't really own a number.
2558
01:53:38,306 --> 01:53:38,426
Yeah.
2559
01:53:38,606 --> 01:53:40,846
I mean, I do think this,
2560
01:53:41,506 --> 01:53:43,026
if you're looking at this as a purist,
2561
01:53:43,166 --> 01:53:45,146
that is factually correct.
2562
01:53:45,286 --> 01:53:47,746
And I do think, to me,
2563
01:53:47,806 --> 01:53:49,146
it's actually highlighting
2564
01:53:49,146 --> 01:53:50,786
that maybe you don't want to live
2565
01:53:50,786 --> 01:53:51,606
for most people
2566
01:53:51,606 --> 01:53:53,346
in a purely cryptographic society.
2567
01:53:53,606 --> 01:53:55,766
Like, the cryptographic proof is amazing,
2568
01:53:55,766 --> 01:53:56,986
and for certain people
2569
01:53:56,986 --> 01:53:58,146
that need to be able to escape
2570
01:53:58,146 --> 01:54:01,046
you know, very corrupt societies or whatever, like it's great that they have that option.
2571
01:54:01,466 --> 01:54:06,206
But I think to me, what it's highlighting more is like, you know, for better or for worse,
2572
01:54:06,266 --> 01:54:10,446
right? Like if you go the self-custody route, you are taking a risk. It's like if you bury a bar of
2573
01:54:10,446 --> 01:54:16,526
gold somewhere else, you are letting it go in a sense and, you know, hoping that that philosophy
2574
01:54:16,526 --> 01:54:21,426
is going to work. If you give that bar of gold to, you know, whatever, a bank in London, you're
2575
01:54:21,426 --> 01:54:26,726
taking a different set of risks. But in theory, you know, the downsides of that we're all aware
2576
01:54:26,726 --> 01:54:28,846
of, and that's why Bitcoin was created, the upside
2577
01:54:28,846 --> 01:54:30,766
of that is you do get this legal apparatus,
2578
01:54:30,906 --> 01:54:32,866
you do get this, you know,
2579
01:54:32,986 --> 01:54:34,946
whatever, centuries of common law that have developed.
2580
01:54:35,266 --> 01:54:36,986
And, you know, I think increasingly
2581
01:54:36,986 --> 01:54:37,926
for most people,
2582
01:54:39,146 --> 01:54:40,786
you know, probably diversify is the answer,
2583
01:54:40,866 --> 01:54:42,906
but, like, there's a reason that we live
2584
01:54:42,906 --> 01:54:44,246
in society, and
2585
01:54:44,246 --> 01:54:46,846
you want access to human
2586
01:54:46,846 --> 01:54:47,746
institutions as well.
2587
01:54:48,826 --> 01:54:50,846
So I think we are long past,
2588
01:54:50,846 --> 01:54:52,046
so we're going to have to
2589
01:54:52,046 --> 01:54:54,686
spiral alpha drop. I'll just
2590
01:54:54,686 --> 01:54:56,706
say, I'll give a
2591
01:54:56,706 --> 01:54:59,666
I'll give a teaser for the future.
2592
01:54:59,846 --> 01:55:00,046
Yeah.
2593
01:55:00,306 --> 01:55:00,526
Okay.
2594
01:55:00,886 --> 01:55:02,666
This is also not a decision.
2595
01:55:03,046 --> 01:55:05,326
Maybe Spiral won't do a new project.
2596
01:55:05,326 --> 01:55:21,646
But I do think this is an important moment in time with the cold card incident that all of a sudden a lot of companies and individuals in Bitcoin are re-evaluating how they secure their Bitcoin.
2597
01:55:21,646 --> 01:55:24,986
and we've already seen
2598
01:55:24,986 --> 01:55:27,866
an enormous shift in sentiment
2599
01:55:27,866 --> 01:55:29,986
with BitKey specifically.
2600
01:55:30,726 --> 01:55:31,906
Like historically,
2601
01:55:32,426 --> 01:55:33,066
most people,
2602
01:55:33,426 --> 01:55:34,066
or I shouldn't say,
2603
01:55:34,426 --> 01:55:35,266
there are a lot of,
2604
01:55:35,386 --> 01:55:36,346
most vendors
2605
01:55:36,346 --> 01:55:38,206
and many people in Bitcoin
2606
01:55:38,206 --> 01:55:40,566
spoke negatively about BitKey
2607
01:55:40,566 --> 01:55:42,726
because it broke the mold.
2608
01:55:43,186 --> 01:55:43,446
Like the,
2609
01:55:44,046 --> 01:55:45,126
everyone was taught,
2610
01:55:45,446 --> 01:55:46,906
you got to have your seed phrase.
2611
01:55:47,006 --> 01:55:48,146
What are your 12 words?
2612
01:55:48,346 --> 01:55:50,486
You got to put it in a steel plate, right?
2613
01:55:50,806 --> 01:55:51,326
And I'm not,
2614
01:55:51,646 --> 01:55:57,146
I am not, I think it's very good to have that option in Bitcoin.
2615
01:55:57,666 --> 01:55:58,346
Very good.
2616
01:55:58,726 --> 01:56:00,206
But I don't think it should be the only option.
2617
01:56:00,586 --> 01:56:03,226
And so BitKey broke that mold and introduced a different way to do it,
2618
01:56:03,266 --> 01:56:06,626
but it got a lot of negativity and hate towards it.
2619
01:56:06,946 --> 01:56:11,326
And the past week, there's way more open-mindedness about the BitKey approach,
2620
01:56:11,586 --> 01:56:12,906
which is great.
2621
01:56:13,146 --> 01:56:15,766
And BitKey sales are like killing it too, which is great.
2622
01:56:15,766 --> 01:56:19,146
so the
2623
01:56:19,146 --> 01:56:21,286
project that I think
2624
01:56:21,286 --> 01:56:23,246
might be very interesting for Spiral
2625
01:56:23,246 --> 01:56:24,406
is
2626
01:56:24,406 --> 01:56:26,906
an open protocol
2627
01:56:26,906 --> 01:56:28,966
for multi-signature
2628
01:56:28,966 --> 01:56:31,106
multi-vendor wallets
2629
01:56:31,106 --> 01:56:31,626
that's a good idea
2630
01:56:31,626 --> 01:56:34,186
drawing from all the design
2631
01:56:34,186 --> 01:56:36,066
and thinking of BitKey
2632
01:56:36,066 --> 01:56:37,706
which I've been a champion of
2633
01:56:37,706 --> 01:56:39,366
for five years since day one
2634
01:56:39,366 --> 01:56:40,366
over five years now
2635
01:56:40,366 --> 01:56:43,786
and taking those lessons
2636
01:56:43,786 --> 01:56:45,666
but taking lessons from the whole ecosystem
2637
01:56:45,666 --> 01:56:48,086
I think Spiral's an amazing position to do this
2638
01:56:48,086 --> 01:56:51,906
because we have a seven-year track record of being neutral
2639
01:56:51,906 --> 01:56:53,966
and for the whole ecosystem and not for one company.
2640
01:56:54,246 --> 01:56:57,046
We also have unique exclusive access to the BitKey team
2641
01:56:57,046 --> 01:56:58,166
because we work for the same company.
2642
01:56:59,326 --> 01:57:02,506
So we'd be in a very good position to do that.
2643
01:57:02,786 --> 01:57:04,826
I think now's a really good time
2644
01:57:04,826 --> 01:57:06,826
because I think vendors will pay attention.
2645
01:57:07,586 --> 01:57:10,886
And two weeks ago, vendors probably would be like,
2646
01:57:11,006 --> 01:57:12,526
we're happy with our current solution.
2647
01:57:12,526 --> 01:57:14,406
And now I think vendors would be like,
2648
01:57:14,406 --> 01:57:16,786
we're open to ideas here.
2649
01:57:16,866 --> 01:57:17,806
We're open to contributing.
2650
01:57:17,946 --> 01:57:18,786
We're open to adopting.
2651
01:57:19,566 --> 01:57:21,266
So we're considering that project.
2652
01:57:22,006 --> 01:57:23,026
And just to make sure,
2653
01:57:23,246 --> 01:57:24,246
can I say it in my own words
2654
01:57:24,246 --> 01:57:25,986
to see if I'm understanding the same thing?
2655
01:57:27,206 --> 01:57:31,186
BitKey historically has been an open design
2656
01:57:31,186 --> 01:57:33,906
that would let other people re-implement that design,
2657
01:57:34,026 --> 01:57:34,966
but it was kind of a,
2658
01:57:35,626 --> 01:57:37,366
I don't know if you'd call it a closed system,
2659
01:57:37,386 --> 01:57:39,026
but it was like a system that was designed
2660
01:57:39,026 --> 01:57:39,686
and implemented,
2661
01:57:39,906 --> 01:57:43,066
and then design was available for other people
2662
01:57:43,066 --> 01:57:44,346
to re-implement in their own way,
2663
01:57:44,406 --> 01:57:46,826
but the system is kind of an all-block system.
2664
01:57:47,106 --> 01:57:51,526
And this potential new idea is for BitKey to be more of a protocol
2665
01:57:51,526 --> 01:57:55,486
for multi-custodian participation,
2666
01:57:55,846 --> 01:57:59,706
so that, say, if ColdCard wanted to be part of a BitKey ecosystem,
2667
01:57:59,706 --> 01:58:03,106
they could create a card that writes into that.
2668
01:58:03,886 --> 01:58:08,146
Close, but the way you're saying it makes it sound like BitKey's at the center.
2669
01:58:09,166 --> 01:58:11,786
And I think it's really important that BitKey is not at the center.
2670
01:58:11,786 --> 01:58:15,786
another approach is the BitKey team
2671
01:58:15,786 --> 01:58:18,286
the commercial team could do exactly what you just said
2672
01:58:18,286 --> 01:58:21,166
and the BitKey product could be changed
2673
01:58:21,166 --> 01:58:24,066
so that another vendor could plug in
2674
01:58:24,066 --> 01:58:26,306
that's possible too
2675
01:58:26,306 --> 01:58:29,746
but that would be like a BitKey led
2676
01:58:29,746 --> 01:58:32,066
BitKey at the center
2677
01:58:32,066 --> 01:58:35,706
the decision makers would be from Block and BitKey
2678
01:58:35,706 --> 01:58:40,246
who are ultimately trying to optimize profits
2679
01:58:40,246 --> 01:58:47,926
it key as a protocol more where it's not run by bitkey is there well that so the the concept here
2680
01:58:47,926 --> 01:58:56,446
would be not run by bitkey it would be initiated by spiral it would be a collaborative project public
2681
01:58:56,446 --> 01:59:02,766
good in the ecosystem where many people could contribute to it's an open protocol i do think it would be heavily
2682
01:59:02,792 --> 01:59:04,092
influenced by BitKey.
2683
01:59:04,452 --> 01:59:07,032
Because I happen to believe
2684
01:59:07,032 --> 01:59:08,452
in almost everything BitKey does.
2685
01:59:08,592 --> 01:59:10,512
So I think that would be a big part in the protocol.
2686
01:59:10,672 --> 01:59:11,732
You can imagine this protocol,
2687
01:59:12,032 --> 01:59:13,632
but it's not just BitKey.
2688
01:59:13,812 --> 01:59:15,852
You can imagine this protocol being more than Bitcoin.
2689
01:59:17,272 --> 01:59:18,732
Oh, for securing your Nostra keys.
2690
01:59:18,832 --> 01:59:21,052
And I don't mean Dogecoin and shitcoins.
2691
01:59:21,052 --> 01:59:23,032
Yeah, I mean Nostra keys.
2692
01:59:23,812 --> 01:59:25,052
SSH keys.
2693
01:59:25,272 --> 01:59:25,752
Yep, yep.
2694
01:59:25,832 --> 01:59:28,052
Your password manager secrets.
2695
01:59:28,592 --> 01:59:29,092
Things like that.
2696
01:59:29,312 --> 01:59:29,972
I love that.
2697
01:59:29,972 --> 01:59:33,492
Now that introduces technical challenges
2698
01:59:33,492 --> 01:59:35,852
because again, the BitKey was not designed for that.
2699
01:59:35,972 --> 01:59:37,852
So assumptions made in BitKey
2700
01:59:37,852 --> 01:59:39,812
don't apply to those other types of keys.
2701
01:59:39,972 --> 01:59:42,292
So it would require technical engineering work to do that.
2702
01:59:42,432 --> 01:59:44,332
But I do like that direction.
2703
01:59:45,412 --> 01:59:48,052
Also, DK, you mentioned BitKey
2704
01:59:48,052 --> 01:59:49,252
as being like an open design.
2705
01:59:49,392 --> 01:59:51,232
I mean, yes, it is.
2706
01:59:52,412 --> 01:59:54,032
People can copy that design.
2707
01:59:54,352 --> 01:59:55,472
The code is open source.
2708
01:59:55,632 --> 01:59:57,972
It's MIT licensed, so you can copy the code too
2709
01:59:57,972 --> 01:59:58,512
and use it.
2710
01:59:58,512 --> 02:00:01,032
but it's very much
2711
02:00:01,032 --> 02:00:05,532
it's very tightly coupled
2712
02:00:05,532 --> 02:00:09,312
the server, the hardware device and the app
2713
02:00:09,312 --> 02:00:13,192
lots of assumptions and it is a very sophisticated design
2714
02:00:13,192 --> 02:00:24,327
for security and privacy If you just change one little thing you can like mess up a bunch of assumptions So to make it more generalizable and realistic for others to adopt it
2715
02:00:24,747 --> 02:00:27,647
that's going to require a lot of engineering work
2716
02:00:27,647 --> 02:00:30,247
and re-evaluating the security and privacy tenets
2717
02:00:30,247 --> 02:00:31,727
to make sure that they hold.
2718
02:00:32,587 --> 02:00:34,247
There'd be a whole lot of work in this project.
2719
02:00:34,407 --> 02:00:35,727
But yeah, the intention is
2720
02:00:35,727 --> 02:00:37,247
it's a neutral open protocol
2721
02:00:37,247 --> 02:00:40,927
that would be perceived as and in reality
2722
02:00:40,927 --> 02:00:42,347
not advantaging
2723
02:00:42,347 --> 02:00:43,827
Locke or any one company.
2724
02:00:44,227 --> 02:00:45,547
I love the idea of being able to hide
2725
02:00:45,547 --> 02:00:47,747
or secure your other keys with that.
2726
02:00:47,887 --> 02:00:48,247
That's right.
2727
02:00:49,347 --> 02:00:53,087
If we do proceed to this,
2728
02:00:53,187 --> 02:00:54,767
I want to build it in Buzz.
2729
02:00:55,127 --> 02:00:56,327
Build it in the open
2730
02:00:56,327 --> 02:00:57,507
for day one.
2731
02:00:58,007 --> 02:01:00,467
So if we do green light this,
2732
02:01:00,627 --> 02:01:01,527
then I would say
2733
02:01:01,527 --> 02:01:03,347
that's how we're going to proceed.
2734
02:01:03,707 --> 02:01:04,127
So we'll see.
2735
02:01:04,847 --> 02:01:05,207
Love it.
2736
02:01:05,527 --> 02:01:06,007
Good alpha.
2737
02:01:06,427 --> 02:01:08,767
I stand by my extreme paranoia
2738
02:01:08,767 --> 02:01:09,727
on the computer thing, by the way.
2739
02:01:09,727 --> 02:01:11,487
every day that goes by
2740
02:01:11,487 --> 02:01:12,327
I'm more and more paranoid
2741
02:01:12,327 --> 02:01:12,647
so
2742
02:01:12,647 --> 02:01:14,427
on that note
2743
02:01:14,427 --> 02:01:16,387
I'm kind of like Liana
2744
02:01:16,387 --> 02:01:17,807
everything is computer
2745
02:01:17,807 --> 02:01:18,867
not really
2746
02:01:18,867 --> 02:01:19,927
I'd say this has like
2747
02:01:19,927 --> 02:01:20,607
a different scope
2748
02:01:20,607 --> 02:01:21,447
than Liana
2749
02:01:21,447 --> 02:01:23,107
Liana's great
2750
02:01:23,107 --> 02:01:24,327
but this is a different scope
2751
02:01:24,327 --> 02:01:25,087
cool
2752
02:01:25,087 --> 02:01:25,627
anyway
2753
02:01:25,627 --> 02:01:26,467
long episode
2754
02:01:26,467 --> 02:01:27,327
long episode
2755
02:01:27,327 --> 02:01:28,447
great episode though
2756
02:01:28,447 --> 02:01:29,667
have a great weekend guys
2757
02:01:29,667 --> 02:01:30,587
bye everyone
2758
02:01:30,587 --> 02:01:31,087
goodbye