Breaking down the Coldcard Incident | Josh | The Bitcoin Nova #104

1
00:00:00,000 --> 00:00:04,480
On July 30th, over 500 Bitcoin addresses got drained in 41 minutes.
2
00:00:04,900 --> 00:00:08,640
Over $100 million gone from people who did everything right.
3
00:00:08,960 --> 00:00:12,380
I'm sitting down today with Josh back for his third time on the show.
4
00:00:12,620 --> 00:00:19,860
He's tracked the story from the beginning and we're going to break it all down for you to how we got here today and what you can do to protect yourself.
5
00:00:20,220 --> 00:00:23,960
We are here today to discuss the cold card incident.
6
00:00:23,960 --> 00:00:27,540
And, you know, sorry for anybody who's been affected.
7
00:00:27,1000 --> 00:00:31,800
And we're here today to talk about what happened.
8
00:00:32,320 --> 00:00:35,140
And Secure Sovereign, this is your third time here on the show.
9
00:00:35,220 --> 00:00:38,080
So thank you for being here and helping bring some clarity to the space.
10
00:00:38,720 --> 00:00:48,320
So if you could go through, you know, how this all went down, how it started, what, you know, what happened for people that, I mean, I'm sure people know.
11
00:00:48,320 --> 00:00:54,000
but just knowing what led to this is really important to understand.
12
00:00:54,960 --> 00:00:57,740
Sure. So we will definitely get into that.
13
00:00:57,940 --> 00:01:00,940
I hope to go into great depth, and thank you for having me.
14
00:01:02,280 --> 00:01:05,940
So I guess what I'll start, I'll start with, you know, who this happened to.
15
00:01:07,060 --> 00:01:11,100
These were not, and other people have made this point, these were not careless people.
16
00:01:12,040 --> 00:01:15,180
These were Bitcoiners who largely did everything right.
17
00:01:15,180 --> 00:01:18,240
They ran their own nodes. In many cases, they held their own keys.
18
00:01:18,320 --> 00:01:42,020
They bought a device. The entire community told them was the most secure option available. They followed setup guides written by people they trusted. Many of them had been in Bitcoin long enough to understand the stakes. And so, you know, that's, I think, the worst part about this is that it happened to people who did everything right.
19
00:01:42,020 --> 00:01:51,740
and the median loss is about 0.4 bitcoin per address roughly 26,500 dollars as of
20
00:01:51,740 --> 00:02:01,280
roughly today's price 110 addresses lost more than one bitcoin as of when this was written
21
00:02:01,280 --> 00:02:09,500
single largest loss at 29.9 almost 30 bitcoin so these are not whales these are plebs who
22
00:02:09,500 --> 00:02:15,000
you know like i said did self-custody right got wiped out many of which within the first hour
23
00:02:15,000 --> 00:02:21,660
and now we're up to over 100 million gone and it's still climbing so it's it's kind of it may
24
00:02:21,660 --> 00:02:30,320
could you frame this as the worst disaster in bitcoin self-custody ever yeah i mean i have
25
00:02:30,320 --> 00:02:36,440
chills and it's it's shaken the community it's shaken i mean i don't know how many people outside
26
00:02:36,440 --> 00:02:40,400
of the Bitcoin community understand what's been going on.
27
00:02:40,400 --> 00:02:45,920
And it really, it happened a week ago, almost,
28
00:02:46,580 --> 00:02:48,540
from when I first found out about it.
29
00:02:49,300 --> 00:02:52,900
And I think everyone was just scrambling
30
00:02:52,900 --> 00:02:54,820
because anybody that had a cold card,
31
00:02:54,940 --> 00:02:56,540
we were advised, at least the community,
32
00:02:57,100 --> 00:02:59,900
was putting out notifications, emails I was getting,
33
00:03:00,020 --> 00:03:01,120
saying if you have a cold card,
34
00:03:01,220 --> 00:03:04,120
remove all of your funds and send them elsewhere.
35
00:03:04,120 --> 00:03:11,660
and I've helped some people do that by the way and fortunately nobody I know has lost any funds but
36
00:03:11,660 --> 00:03:17,360
I've seen reports all day like that and like that's the thing too I've seen people that have
37
00:03:17,360 --> 00:03:21,680
been helping others who they helped initially set up with the cold card and then they felt like it
38
00:03:21,680 --> 00:03:28,160
was their duty to help make sure that they didn't get affected and I've just seen like people feeling
39
00:03:28,160 --> 00:03:31,500
exhausted people feeling tired because they're helping all their friends and they're reaching
40
00:03:31,500 --> 00:03:36,560
out and doing everything they can to make sure that they get to everybody. So it's nice to see
41
00:03:36,560 --> 00:03:41,640
that, that people are trying to make sure that other people are affected. So thank you for doing
42
00:03:41,640 --> 00:03:48,500
that for people you know. And it's just, you're also building solutions, which, you know, I can
43
00:03:48,500 --> 00:03:52,160
share the link for people too that want to check out your work, but you have... We can get into
44
00:03:52,160 --> 00:04:00,940
that later. But this isn't about me. This is really a tragic event. And I'm kind of looking at it from
45
00:04:00,940 --> 00:04:03,540
kind of a governance perspective.
46
00:04:03,540 --> 00:04:05,460
I was already doing a lot of governance research
47
00:04:05,460 --> 00:04:07,600
on Bitcoin and the source code
48
00:04:07,600 --> 00:04:09,640
and the review of the source code
49
00:04:09,640 --> 00:04:11,980
in Bitcoin Core in particular.
50
00:04:12,660 --> 00:04:16,200
And, you know, I think a lot of this carries over
51
00:04:16,200 --> 00:04:17,860
the kind of work, the kind of perspective
52
00:04:17,860 --> 00:04:19,940
that I was applying to Bitcoin Core.
53
00:04:20,160 --> 00:04:22,500
It carried over very easily.
54
00:04:23,180 --> 00:04:25,060
So I've put together an article,
55
00:04:25,280 --> 00:04:28,340
which you read, that I posted on my website,
56
00:04:28,520 --> 00:04:30,180
secsov.com.
57
00:04:30,180 --> 00:04:36,160
called Don't Trust Verify, where I go into, it's a bit of a postmortem, like a technical
58
00:04:36,160 --> 00:04:43,540
postmortem about what happened, but it's mostly about the sequence of events and how we got to
59
00:04:43,540 --> 00:04:49,520
where we're at in terms of people trusting this device, nobody checking, and then how it became
60
00:04:49,520 --> 00:04:55,180
one of the most popular devices, considered to be one of the most secure devices, when in fact,
61
00:04:55,180 --> 00:05:00,700
the opposite was true. And I think that is going to require a deep dive. The wounds are still fresh.
62
00:05:00,880 --> 00:05:06,300
And, you know, I considered that when I published the article. But for those who weren't affected
63
00:05:06,300 --> 00:05:18,863
and weren in the middle of it I think it a good time to start talking about this because it doesn just apply to wallets It applies to the whole ecosystem So I just go ahead and dive into this
64
00:05:19,023 --> 00:05:22,063
The story, as far as I'm concerned, starts in 2012.
65
00:05:23,203 --> 00:05:30,943
CoinKite was founded in Toronto by NVK, that's his pseudonym, his real name is Rodolfo Novak.
66
00:05:30,943 --> 00:05:37,663
it's a Canadian company that makes cold card and other bitcoin hardware wallets
67
00:05:37,663 --> 00:05:42,383
nbk is its founder and public face and a long time bitcoin personality
68
00:05:42,383 --> 00:05:47,043
known for his strong opinions on security and open source
69
00:05:47,043 --> 00:05:57,643
and so that he's doing that and then in 2014 cold card launched and at that time
70
00:05:57,643 --> 00:06:05,223
The firmware was built on Trezor's cryptographic library, which was under the GPL v3 license.
71
00:06:05,363 --> 00:06:07,043
GPL is the general public license.
72
00:06:07,263 --> 00:06:09,883
It's a very permissive open source license.
73
00:06:11,183 --> 00:06:18,023
Trezor is made by a Czech company named Satoshi Labs.
74
00:06:18,883 --> 00:06:23,943
And that was open source from day one.
75
00:06:23,943 --> 00:06:31,763
using the gpl library meant cold card inherited the gpl license obligation so that's how these
76
00:06:31,763 --> 00:06:38,923
licenses work if you use the code you're permitted to use the code you can even commercially profit
77
00:06:38,923 --> 00:06:44,063
from using that code but you have to in most cases you have to publish your changes in the
78
00:06:44,063 --> 00:06:52,663
your your code as well so basically anyone who forks or distributes it must
79
00:06:52,663 --> 00:06:55,943
also released their changes under the same terms,
80
00:06:56,163 --> 00:06:57,683
under the same GPL license.
81
00:06:58,563 --> 00:07:02,363
In 2019, onward, promotion starts.
82
00:07:02,883 --> 00:07:07,263
Stefan Levera, he is obviously the host
83
00:07:07,263 --> 00:07:08,543
of the Stefan Levera podcast,
84
00:07:08,803 --> 00:07:10,963
one of Bitcoin's most listened to shows,
85
00:07:10,963 --> 00:07:17,743
runs an ad, first major documented cold card episode.
86
00:07:18,543 --> 00:07:22,123
There's also an instance in 2019,
87
00:07:22,663 --> 00:07:36,143
There's a hardware wallet interface that was written by, partially written by Ava Chow, who's a Bitcoin maintainer, and Bitcoin Core's official tooling for hardware wallet support.
88
00:07:36,503 --> 00:07:41,383
And they add first-class cold card support in 2019.
89
00:07:42,043 --> 00:07:48,423
Now, this is before, I'll point out, this is before they switched the firmware over to the broken firmware.
90
00:07:48,623 --> 00:07:50,243
So that happened in 2021.
91
00:07:50,243 --> 00:07:52,403
at this time
92
00:07:52,403 --> 00:07:54,963
like I said Coldcard is still using
93
00:07:54,963 --> 00:07:56,663
the Trezor firmware which is secure
94
00:07:56,663 --> 00:07:58,943
firmware, Stefan Levero
95
00:07:58,943 --> 00:08:00,483
Ava Chow
96
00:08:00,483 --> 00:08:01,463
are working on
97
00:08:01,463 --> 00:08:04,083
Stefan posts this
98
00:08:04,083 --> 00:08:06,783
podcast promoting
99
00:08:06,783 --> 00:08:08,163
the Coldcard
100
00:08:08,163 --> 00:08:09,483
Ava Chow
101
00:08:09,483 --> 00:08:11,903
starts working on making
102
00:08:11,903 --> 00:08:14,563
first class support for Coldcard
103
00:08:14,563 --> 00:08:16,643
in the ecosystem and then
104
00:08:16,643 --> 00:08:18,543
also we have the
105
00:08:18,543 --> 00:08:19,703
rabbit hole recap
106
00:08:19,703 --> 00:08:23,923
which is a popular weekly podcast co-hosted by Matt O'Dell and Marty Bent,
107
00:08:24,643 --> 00:08:28,403
two of Bitcoin's most trusted self-custody educators,
108
00:08:29,083 --> 00:08:31,923
listing CoinKite as a direct sponsor.
109
00:08:32,343 --> 00:08:34,843
So this just gets to a discussion point,
110
00:08:34,983 --> 00:08:39,443
which is NBK's own words from Baltic Honey Badger 2019,
111
00:08:39,863 --> 00:08:44,463
quote, you can prove to yourself that the device is safe, end quote.
112
00:08:45,603 --> 00:08:49,323
And then from SLP 101, the Stefan Levera podcast,
113
00:08:49,323 --> 00:08:56,223
he says, quote, everything is open source. You can go and buy the parts yourself,
114
00:08:57,063 --> 00:09:04,083
compile the code yourself, and load it. And then he goes on to say, that proves that we're using
115
00:09:04,083 --> 00:09:11,223
this thing in an open source manner, end quote. From a different episode in February 2021, he says,
116
00:09:11,223 --> 00:09:17,643
quote, a fully open source device, all the code is verifiable, end quote. At that time,
117
00:09:17,643 --> 00:09:19,163
these statements were accurate.
118
00:09:19,663 --> 00:09:21,203
Because it hadn't switched over.
119
00:09:21,343 --> 00:09:22,963
It was still on Trezor's code.
120
00:09:23,563 --> 00:09:26,803
Right, with the very permissive open source license.
121
00:09:27,863 --> 00:09:31,323
So the problem is what came after.
122
00:09:31,523 --> 00:09:32,703
The license changed.
123
00:09:33,183 --> 00:09:35,463
The open source ethos disappeared.
124
00:09:35,643 --> 00:09:37,083
Now, this is an important point,
125
00:09:37,743 --> 00:09:41,843
but the verifiability marketing continued.
126
00:09:42,943 --> 00:09:45,403
So don't trust Verify on the packaging.
127
00:09:46,223 --> 00:09:46,983
It was also in there.
128
00:09:46,983 --> 00:09:52,983
Can you, before you continue with that, tell people why it changed so that they understand the history of their...
129
00:09:54,683 --> 00:09:55,963
Sure.
130
00:09:56,323 --> 00:09:58,983
So, yeah.
131
00:09:59,743 --> 00:10:01,123
So, I was about to get to that.
132
00:10:01,283 --> 00:10:01,783
Okay, okay.
133
00:10:02,003 --> 00:10:03,203
Continue that as part of it.
134
00:10:03,903 --> 00:10:05,423
We'll definitely build up to that.
135
00:10:05,523 --> 00:10:07,823
So, we've got to go through 2020, and then we'll get to 2021.
136
00:10:08,083 --> 00:10:09,143
That's where all the fun starts.
137
00:10:09,703 --> 00:10:09,843
Okay.
138
00:10:09,843 --> 00:10:18,703
So through that, the seed generation path was never independently audited, not once.
139
00:10:19,703 --> 00:10:31,046
Now to be fair Trezor code had been somewhat audited Wallet and other prominent security researchers had looked at Trezor code
140
00:10:31,046 --> 00:10:36,446
So there wasn't really much need to look at CoinKite's code because Trezor was pretty well vetted.
141
00:10:37,886 --> 00:10:45,566
That said, we also have in September 2020, Michael Flaxman, respected Bitcoin security researcher,
142
00:10:45,566 --> 00:10:50,286
publishes a definitive multi-sig self-custody reference.
143
00:10:50,526 --> 00:10:52,566
Cold Card is the primary signing device.
144
00:10:53,066 --> 00:10:54,366
Prior multi-sig guides,
145
00:10:55,166 --> 00:10:58,886
they shut themselves down and redirected users here.
146
00:10:59,226 --> 00:11:03,406
So this was the sort of de facto choice guide for people
147
00:11:03,406 --> 00:11:04,926
in many cases at that time.
148
00:11:06,026 --> 00:11:08,526
On July 2020, Foundation Devices,
149
00:11:08,666 --> 00:11:10,086
a U.S. hardware wallet company,
150
00:11:10,086 --> 00:11:12,946
announces Passport, a competing device
151
00:11:12,946 --> 00:11:15,446
on Cold Card's GPL firmware.
152
00:11:15,566 --> 00:11:18,126
which was legally permitted under the license.
153
00:11:19,246 --> 00:11:22,406
Now, on July 30th, so that was July 28th
154
00:11:22,406 --> 00:11:25,186
when Foundation Devices announced that,
155
00:11:25,446 --> 00:11:27,086
July 30th, two days later,
156
00:11:27,206 --> 00:11:30,086
NVK publicly regrets choosing the GPL,
157
00:11:30,746 --> 00:11:32,826
saying Coldcard now has a clone.
158
00:11:34,006 --> 00:11:36,786
And then he basically begins planning
159
00:11:36,786 --> 00:11:38,586
to change the license at that point.
160
00:11:40,426 --> 00:11:42,666
In November of 2020,
161
00:11:43,546 --> 00:11:47,126
CoinKite adds MIT plus Commons Clause license.
162
00:11:47,926 --> 00:11:54,786
The Commons Clause essentially prevents the code under that license from being used for
163
00:11:54,786 --> 00:11:55,806
commercial purposes.
164
00:11:56,546 --> 00:12:02,086
So while it was technically an open source license, most people in the open source community
165
00:12:02,086 --> 00:12:08,406
would not consider that to be an open source license because while the source code is available,
166
00:12:08,406 --> 00:12:13,866
you can't actually build or profit off of the codes,
167
00:12:14,066 --> 00:12:17,626
which would generate the kind of scrutiny
168
00:12:17,626 --> 00:12:23,726
that would lead people to check it out
169
00:12:23,726 --> 00:12:26,086
and perhaps discover this bug.
170
00:12:26,226 --> 00:12:27,826
So that while the source code was available,
171
00:12:28,666 --> 00:12:31,306
there wasn't any incentive for anybody to check.
172
00:12:31,586 --> 00:12:32,426
And so nobody did.
173
00:12:32,426 --> 00:12:44,426
um so uh it actually gets worse than that um so nvk's explanation in the 2025 bitcoin magazine
174
00:12:44,426 --> 00:12:51,386
interview was he says quote uh somebody went out there and without mentioning us without anything
175
00:12:51,386 --> 00:12:59,246
just took the code didn't even bother to change anything contribute back zero contributions back
176
00:12:59,246 --> 00:13:01,826
and started a competing company.
177
00:13:02,346 --> 00:13:04,126
So we're like, you know what?
178
00:13:04,426 --> 00:13:05,046
F you.
179
00:13:05,486 --> 00:13:08,006
And we changed the license, end quote.
180
00:13:08,786 --> 00:13:10,726
He was referring to Foundation Devices.
181
00:13:11,526 --> 00:13:15,346
Foundation Devices had in fact disclosed the fork publicly from day one
182
00:13:15,346 --> 00:13:20,546
and was legally exercising their rights NVK had granted by choosing the GPL.
183
00:13:20,546 --> 00:13:23,586
The code is still publicly readable,
184
00:13:23,946 --> 00:13:27,486
but no longer open source by any standard definition, as I said.
185
00:13:27,486 --> 00:13:36,386
And the Commons Clause FAQ itself states that the software using it is not open source.
186
00:13:37,086 --> 00:13:47,046
So I think the arc here is that this thing from the very beginning was sold as open source.
187
00:13:47,046 --> 00:14:00,066
And then by 2021, January 8, 2021, the license is formally changed, and CoinKite starts calling the firmware verifiable instead of open source.
188
00:14:00,346 --> 00:14:05,446
And that distinction becomes very murky, as I'll get into in a minute.
189
00:14:05,446 --> 00:14:16,126
But the discussion point is foundation devices forked this GPL firmware legally.
190
00:14:16,426 --> 00:14:18,606
NVK regretted it publicly two days later.
191
00:14:18,766 --> 00:14:22,066
Licensed migration produced a defective commit.
192
00:14:22,506 --> 00:14:25,746
So just consider the chain of events here.
193
00:14:26,246 --> 00:14:28,246
Trezor is under the GPL.
194
00:14:28,246 --> 00:14:37,126
and cold cards firmware in order to adhere to the gpl must uh be under the gpl so in order to get
195
00:14:37,126 --> 00:14:43,266
out from under the gpl it had to get rid of all of trezor's code okay and then in the getting rid
196
00:14:43,266 --> 00:14:51,426
of all of trezor's code which had the cryptographic library that um was vetted they pulled all of that
197
00:14:51,426 --> 00:15:03,106
out and replaced it with a, uh, another library called libngu and libngu had, I think at the time,
198
00:15:03,586 --> 00:15:09,246
uh, six followers on GitHub. It was a very small repository and that there's, there currently
199
00:15:09,246 --> 00:15:16,006
actually is some debate about whether or not, uh, that, that repository was in fact created and
200
00:15:16,006 --> 00:15:21,006
maintained by CoinKite itself under a pseudonym
201
00:15:21,966 --> 00:15:24,046
that is in fact their CTO,
202
00:15:24,046 --> 00:15:26,126
but I don't wanna say that definitively,
203
00:15:26,126 --> 00:15:27,206
that hasn't been proven,
204
00:15:27,206 --> 00:15:30,986
but there is some evidence to indicate that.
205
00:15:30,986 --> 00:15:43,629
So I you know this is I mean I don know what to say honestly I mean if you think about the fact pattern here you got NVK basically retaliating against foundation devices
206
00:15:43,629 --> 00:15:45,769
for using their license legally,
207
00:15:46,369 --> 00:15:49,729
and then in the process of retaliating,
208
00:15:50,449 --> 00:15:53,928
you know, cutting themselves off of the knees
209
00:15:53,928 --> 00:15:56,109
from their strongest security,
210
00:15:56,428 --> 00:16:00,029
and then replacing it with security that,
211
00:16:00,149 --> 00:16:01,928
well, isn't security at all.
212
00:16:01,928 --> 00:16:22,909
It's in fact very insecure and as we have seen has resulted in over $100 million worth of Bitcoin being stolen from average people, which is, you know, there are no words to describe the sadness and anger that I and others feel about that.
213
00:16:23,069 --> 00:16:27,509
Honestly, it's just unconscionable on so many levels.
214
00:16:27,509 --> 00:16:43,428
I think that just thinking about the alliances that form in life and how maybe people think with NVK's situation, they knew him for so long.
215
00:16:43,509 --> 00:16:45,249
They've maybe worked with him for a long time.
216
00:16:45,369 --> 00:16:46,829
It's like these people have been in the space.
217
00:16:46,949 --> 00:16:51,509
So then they thought, oh, he's mad that somebody copied him.
218
00:16:51,729 --> 00:16:54,569
Which he himself did with Trezor.
219
00:16:54,569 --> 00:16:57,089
Himself did already.
220
00:16:57,509 --> 00:17:08,809
But when it's like somebody you consider a friend, then people think, well, I need to stand by this person because it becomes this, like, then you look like you're siding, like you're taking sides with somebody, even though you like.
221
00:17:09,089 --> 00:17:13,948
So he was visibly angry online, you know, speaking his mind about it.
222
00:17:14,029 --> 00:17:14,849
People knew this.
223
00:17:15,789 --> 00:17:27,329
And then it feels like anybody who's associates with him instead of calling him out just takes his side and then like just trusts him and doesn't verify.
224
00:17:27,508 --> 00:17:29,948
when this changes.
225
00:17:30,948 --> 00:17:32,849
Right. And people have made the comment.
226
00:17:32,968 --> 00:17:36,029
I've seen comments about how when they would promote other wallets,
227
00:17:36,129 --> 00:17:38,429
they would get bullied by cold card users.
228
00:17:39,549 --> 00:17:47,129
NVK, he would routinely put down other wallets.
229
00:17:47,129 --> 00:17:50,208
He actually bought Seedsigner domain.
230
00:17:50,809 --> 00:17:56,029
I think he bought Seedsigner.org and then he redirected it somewhere else.
231
00:17:56,029 --> 00:18:16,289
So they couldn't even use the domain. They apparently asked him, you know, can we please have this domain? Well, I assume that they offered him money for it. He refused to give it to them. And there are many posts where he's trash talking using a Raspberry Pi, which is the seed signer method for their signing devices.
232
00:18:16,289 --> 00:18:22,609
and so he's out there trashing the alternative wallets
233
00:18:22,609 --> 00:18:27,708
and he's, you know, not only trashing them
234
00:18:27,708 --> 00:18:30,769
but literally sabotaging their operations in some cases
235
00:18:30,769 --> 00:18:38,008
and basically just being very hostile to the ecosystem.
236
00:18:38,008 --> 00:18:42,349
And honestly, you know, it's actually a pretty small ecosystem.
237
00:18:42,649 --> 00:18:43,468
I mean, if you think about it,
238
00:18:43,468 --> 00:18:48,549
there aren't many wallets in this ecosystem, hardware wallets. So, I mean, there's like
239
00:18:48,549 --> 00:18:54,488
three or four other people that he can trash talk with any significance. And he was.
240
00:18:54,968 --> 00:18:59,849
That's crazy to think that people just didn't have their alarm bells going off,
241
00:18:59,849 --> 00:19:05,289
like going off their intuition wasn't speaking to them as much because people think
242
00:19:05,289 --> 00:19:11,508
in such a small space with not that many wallets out there, this space is usually very like open
243
00:19:11,508 --> 00:19:16,748
source friendly helping each other building off each other like that's what bitcoin is and should
244
00:19:16,748 --> 00:19:25,349
be and when you see that somebody that again did the same thing that like hypocritically
245
00:19:25,349 --> 00:19:33,508
um you know just angry at somebody for doing exactly what what he did as a company to start
246
00:19:33,508 --> 00:19:39,049
it you know it's like what if legend what if trezor never had their code open source for anyone
247
00:19:39,049 --> 00:19:44,309
to borrow you know he would have never had anything you know so it's like and then openly
248
00:19:44,309 --> 00:19:50,948
and vocally criticizing everybody else it just makes you almost like people start to believe it
249
00:19:50,948 --> 00:19:55,049
because there isn't anyone else really in the space acting that way i don't think like i mean
250
00:19:55,049 --> 00:20:02,248
not to say that they're not um i just mean that nobody as prominent as he was and right not with
251
00:20:02,248 --> 00:20:08,909
the language not the same language yeah yeah and then it was amazing to me to think like when
252
00:20:08,909 --> 00:20:16,968
this happened because again like you said there was people that would talk down to you from you
253
00:20:16,968 --> 00:20:23,149
know from cold card that if you weren't using it like him like and mbk but even when you would talk
254
00:20:23,149 --> 00:20:28,389
to some of your friends online and you talk about how to store your bitcoin and how to secure it it
255
00:20:28,389 --> 00:20:38,008
was mainly the hardcore bitcoiners that were saying they used the cold card and if you if you
256
00:20:38,008 --> 00:20:39,809
were to mention other walls, they kind of were
257
00:20:39,809 --> 00:20:42,049
the same way, like the same, like,
258
00:20:42,089 --> 00:20:43,708
hey, no, cold card is the way to go.
259
00:20:44,149 --> 00:20:46,029
And then it almost made people
260
00:20:46,029 --> 00:20:47,769
feel like, in my case,
261
00:20:47,849 --> 00:20:49,369
I never...
262
00:20:50,011 --> 00:20:59,471
decided to go with cold card. And I felt like sometimes almost like the people that were users
263
00:20:59,471 --> 00:21:04,551
or the people that did support or like, I mean, it's like, obviously you believed it too. You'd
264
00:21:04,551 --> 00:21:09,671
think like, who's going to, all these people support it. You're going to just trust it,
265
00:21:09,671 --> 00:21:15,891
but I felt bad kind of being open about the fact that I'm not using cold card because I feel like
266
00:21:15,891 --> 00:21:21,411
You almost would be judged if you weren't a cold card user because it made you look like you weren't that serious.
267
00:21:22,831 --> 00:21:24,071
The irony of that.
268
00:21:24,811 --> 00:21:25,211
Yeah.
269
00:21:26,651 --> 00:21:29,691
So I think that people, they call it that purity test.
270
00:21:29,691 --> 00:21:41,711
And a lot of people are kind of, I've seen across the Bitcoin ecosystem, people talking about maybe this will end those purity tests and people will stop making others feel bad for not doing one thing or the other.
271
00:21:41,711 --> 00:21:56,751
But I also think that it's caused this turn where everybody's talking about, well, you have to do the dice method, roll the dice and roll the dice 200 times and create your own seed phrase.
272
00:21:56,911 --> 00:21:58,291
And okay, you can do that.
273
00:21:59,531 --> 00:22:06,971
But that's what I think I was investigating because so many people got so scared.
274
00:22:06,971 --> 00:22:12,991
they're like is this going to happen to every other wallet and every other cphrase generation
275
00:22:12,991 --> 00:22:16,631
is this something that's an issue in other companies is this something that we should
276
00:22:16,631 --> 00:22:21,731
worry about anywhere we store our funds that isn't multi-sig because a lot of people have just
277
00:22:21,731 --> 00:22:26,831
been saying that are actually confident in their setup is multi-sig is the way to go you don't
278
00:22:26,831 --> 00:22:35,171
have one single point of failure um but when you just think of uh i don't know like the the dice
279
00:22:35,171 --> 00:22:42,671
method thing i don't think it's a we shouldn't go backwards um in a way where i think it's
280
00:22:42,671 --> 00:22:49,111
definitely the way to verify right but as far as us making bitcoin more accessible to more people
281
00:22:49,111 --> 00:22:53,611
i don't think you're gonna talk to somebody about bitcoin that has no idea what it is
282
00:22:53,611 --> 00:22:57,991
and actually orange pill them and they finally get it and then you're like now come rolls 200
283
00:22:57,991 --> 00:23:01,651
dice rolls with me and like pick it you know what i mean like i'll show you how to do this i don't
284
00:23:01,651 --> 00:23:07,111
think people will do that. They'll think you're insane. And it's not a good to say that's the
285
00:23:07,111 --> 00:23:13,151
only way to know for sure. Then how are you going to onboard other people? I'm with you. Yeah.
286
00:23:13,151 --> 00:23:21,111
100%. So the company that helped me recover my Bitcoin, Nova Sapien at NovaSapien.com,
287
00:23:21,851 --> 00:23:31,431
he sells self-custody kits and he bundles them with the seed signer. And in the kit is
288
00:23:31,431 --> 00:23:37,671
a set of dice and a little piece of paper with all of the words from the word list on it.
289
00:23:38,311 --> 00:23:45,071
And he's like very hardcore about this. He's like, you use those dice. And he's got it down
290
00:23:45,071 --> 00:23:50,351
because of the way he configured it. He's got it down to less than 30 rolls or something to
291
00:23:50,351 --> 00:23:57,091
be able to get decent entropy. Well, I mean, as good an entropy as you're going to get.
292
00:23:57,091 --> 00:24:07,111
they're not casino dice but they are decent I'd say high quality dice it's a decent kit
293
00:24:07,111 --> 00:24:15,951
and I have used that to generate private keys and when all this went down I was you know sleeping
294
00:24:15,951 --> 00:24:23,011
like a baby because I knew that I had generated my own entropy and you know unless Bitcoin itself
295
00:24:23,011 --> 00:24:31,011
was broken, which was not the case, then I was okay. And so I think at the end of the...
296
00:24:31,011 --> 00:24:31,851
Go ahead.
297
00:24:31,911 --> 00:24:40,351
Sorry. For people who don't do multi-sig and they want to make sure and sleep like a baby,
298
00:24:40,351 --> 00:24:52,211
you are able to create your own entropy. But people who don't do that, it's kind of
299
00:24:53,011 --> 00:25:20,571
I don't know. I think it's one of those things where you are just, you're trusting that the whole reason, just in case anyone doesn't understand, is because if this situation happened, like what happened with Cold Card, and they don't do the right method, it can not create enough of that and then be able to make it easier to guess, I guess, is what it was.
300
00:25:20,571 --> 00:25:47,551
That's absolutely right. Yes. And then, of course, so there's two points there, I would say. First, if you're not rolling dice yourself, you are relying on some sort of random number generator to generate your entropy. And that is a digital thing that may or may not be compromised. And I would venture to guess that probably most of them aren't.
301
00:25:47,551 --> 00:25:48,551
Okay?
302
00:25:48,551 --> 00:25:53,131
In fact, many people said about Conecard that the entropy was very strong.
303
00:25:53,131 --> 00:26:19,748
NVK was going out there and saying that entropy was great and there no issues and you know so but nonetheless you relying on a digital device and that is complexity OK And unless you a hardware guru who is willing to break out the soldering iron you know and understands all of you know microcircuitry and all this stuff then you really have no way of verifying that whether that is legitimate
304
00:26:19,748 --> 00:26:26,148
And not to mention the firmware side and the interaction with the hardware and and all this stuff.
305
00:26:26,148 --> 00:26:30,648
whereas if you have dice and a piece of paper with words on it anybody can understand that
306
00:26:30,648 --> 00:26:38,728
and i think that's part of it is you want to um you know don't abstract away the complexity
307
00:26:38,728 --> 00:26:43,948
from people make it as simple as possible it may seem a little silly to be sitting there rolling
308
00:26:43,948 --> 00:26:50,468
dice and have this you know piece of paper with words on it but it is it's just it's very
309
00:26:50,468 --> 00:26:53,908
straightforward it's something anybody can understand and the fact that anybody can understand
310
00:26:53,908 --> 00:26:59,048
it is the value. And it also happens to be the strongest form of entropy that you can get,
311
00:26:59,268 --> 00:27:04,228
I would argue, without complexity, without trusting anybody else. And that's why
312
00:27:04,228 --> 00:27:09,068
Nova Sapien built that in the way that they did, because they didn't want to trust anybody else.
313
00:27:09,068 --> 00:27:14,688
And I'm thankful that they did. And that's what he taught me when I did the workshop with them.
314
00:27:14,928 --> 00:27:22,628
And, you know, that's good. Now, the flip side of that, that you alluded to rightly,
315
00:27:22,628 --> 00:27:30,068
is that in terms of accessibility, in terms of onboarding hundreds of millions of people into self-custody,
316
00:27:30,608 --> 00:27:33,248
you know, that is not a very feasible thing to do.
317
00:27:33,368 --> 00:27:37,628
We can't, you know, the average American has the attention span of a goldfish.
318
00:27:38,268 --> 00:27:44,728
You know, they're not going to sit there and roll dice 30 times, much less 50 or 100 times to get random 12 words
319
00:27:44,728 --> 00:27:49,808
so they can be their own bank account and have, you know, all of these other security considerations
320
00:27:49,808 --> 00:27:55,768
like backups and, you know, putting it in a titanium and having multiple geographic locations
321
00:27:55,768 --> 00:28:00,408
and like time delays and all these other things that go into being your own bank.
322
00:28:00,448 --> 00:28:05,588
It's a lot of responsibility to take sovereignty into your own hands.
323
00:28:06,208 --> 00:28:12,268
And, you know, obviously I encourage people to do it, but, you know, and this is a principle
324
00:28:12,268 --> 00:28:12,928
in cybersecurity.
325
00:28:12,928 --> 00:28:22,548
The problem with operational security is you don't know that your operational security is bad generally until it's too late.
326
00:28:23,108 --> 00:28:29,548
And then when you're talking about your own money with your own savings, you know, there's only so many times where it can be too late.
327
00:28:29,688 --> 00:28:34,388
And then, you know, you're wiped out, you know, like you get one try to get this right.
328
00:28:34,608 --> 00:28:39,148
You don't get to, you know, by trial and error.
329
00:28:39,188 --> 00:28:41,128
You don't get a lot of trial and error with this, I guess.
330
00:28:41,128 --> 00:28:46,988
And so it's definitely a difficult thing to have to balance.
331
00:28:47,288 --> 00:29:04,348
And I think there is a much deeper issue there in terms of just people having short attention spans and people not really being educated in general or willing to take on the responsibility of being their own bank.
332
00:29:04,628 --> 00:29:10,128
And that's really kind of a separate discussion than self-custody in and of itself.
333
00:29:10,128 --> 00:29:10,948
Does that make sense?
334
00:29:11,128 --> 00:29:32,968
Yeah, 100%. Yeah, so I like that there are options like what Nova Sapien did with doing it less times and creating enough entropy. So it's nice that you could do it that way and not feel like, what if you screw up? Because I've heard other people even say like, if people that don't know what they're doing, they could screw up the dice roll method if you don't do it right.
335
00:29:32,968 --> 00:29:52,088
So for him to be able to come up with a method and make sure. And then on the other hand, is it, I mean, it's almost like you need to be able to do that when you're going to hold it yourself and have a single, you know, and not do a multi-sig, you need to be able to make sure and not trust.
336
00:29:52,088 --> 00:29:57,968
because obviously, I mean, I think that the more I've learned about this, the more it's
337
00:29:57,968 --> 00:30:04,988
a cold card issue. But then again, when the next phase of wallets comes out in the future
338
00:30:04,988 --> 00:30:09,568
and somebody else has an alliance with somebody else and they want to support their friend because
339
00:30:09,568 --> 00:30:14,668
they did a different thing and they have really good marketing teams, are you just going to trust
340
00:30:14,668 --> 00:30:20,248
and not verify that this thing is secure? Or, you know, how do we know? Like, how do we
341
00:30:20,248 --> 00:30:25,728
make sure that it's almost like there should be a rating of how many people
342
00:30:25,728 --> 00:30:33,608
were able to check this entropy and see that it was thoroughly checked or something some way for
343
00:30:33,608 --> 00:30:40,088
people to actually know that there's been a review and people can check on everybody's
344
00:30:40,088 --> 00:30:45,188
this is this is a really tough problem because on one hand you've got getting an audit of the
345
00:30:45,188 --> 00:30:49,208
source code and then auditing every update of the source code because you never know when it could
346
00:30:49,208 --> 00:30:57,708
change right so and any any open source security critical software has the same issue you know I
347
00:30:57,708 --> 00:31:04,648
think when you maybe have something like that happen though where it's a a fracture in the
348
00:31:04,648 --> 00:31:29,404
community and people see something like what was going on during the time with foundation and and then you see the transition from Trezor to cold card and they didn have have fit you know they didn cause an issue over that happening So then it was fine But then when you see somebody else being upset about this and then changing something of theirs in a moment of anger you know what I mean
349
00:31:29,404 --> 00:31:39,164
It's like you might be making mistakes or, you know, that should be a time where when you see fracture in the community is when people should start looking even more, you know, I don't know.
350
00:31:40,004 --> 00:31:45,224
I just can't get over the fact that like how it got to be promoted.
351
00:31:45,924 --> 00:31:47,484
Okay, well, so I'll go on.
352
00:31:47,664 --> 00:31:50,404
So in early 2022, James O'Byrne, by the way,
353
00:31:50,444 --> 00:31:52,604
he was a very well-respected Bitcoin Core contributor.
354
00:31:52,904 --> 00:31:53,564
I've met him.
355
00:31:53,664 --> 00:31:54,324
I like him.
356
00:31:54,564 --> 00:31:55,684
So don't get me wrong.
357
00:31:56,264 --> 00:32:00,424
But, you know, he's, he releases something called Cold Core,
358
00:32:00,724 --> 00:32:02,204
which is a wallet tool.
359
00:32:02,684 --> 00:32:06,684
And he says in the readme of the tool, he says, quote,
360
00:32:06,684 --> 00:32:17,604
The only wallet supported at the moment because it is open source and supports air gapped use via partially signed Bitcoin transactions referring to ColdCard.
361
00:32:18,404 --> 00:32:26,184
So for technically serious users, ColdCard is now the default signing device, not just through marketing, but through official tooling.
362
00:32:26,684 --> 00:32:33,844
OK, and I'll point out also, as I did in the article, that there's a difference between source available and open source.
363
00:32:33,844 --> 00:32:51,104
And the official guidance on MIT plus Commons clause is that it technically is not open source. But this line between source available and open source starts to get murky after 2021 when they move to this different firmware.
364
00:32:51,104 --> 00:33:00,344
um and then in 20 again in 2022 uh in june portland hodl again a well-respected bitcoin
365
00:33:00,344 --> 00:33:06,284
core contributor working at anchor watch um i don't know if he did this in his capacity at
366
00:33:06,284 --> 00:33:13,444
anchor watch but he works there he publicly reviews the mk4 stm 32 random number generator
367
00:33:13,444 --> 00:33:19,264
driver in being used by the cold card and comes to the conclusion that quote their driver
368
00:33:19,264 --> 00:33:28,504
implementation for the STM32RNG is correct, end quote. And he is right. The driver is correct
369
00:33:28,504 --> 00:33:33,764
because the bug is not in the driver. It is in the build guard in a different file,
370
00:33:34,244 --> 00:33:39,884
the guard that decides whether the driver is called at all. So basically, he accurately
371
00:33:39,884 --> 00:33:46,804
reviewed the driver implementation, said it was correct, but didn't realize that that correct
372
00:33:46,804 --> 00:33:57,124
driver wasn't actually running, which, you know, it's tough. I mean, some people were trying to
373
00:33:57,124 --> 00:34:02,504
look at this and it was hard to see exactly what was going on and what wasn't going on.
374
00:34:03,264 --> 00:34:03,424
Right.
375
00:34:05,924 --> 00:34:12,164
So the community is trusting credential review stops at the wrong files for years of false
376
00:34:12,164 --> 00:34:20,404
confidence uh again in we've got february 22nd 2023 adam back co-founder and ceo of blockstream
377
00:34:20,404 --> 00:34:27,764
uh and inventor of hash cash he's cited in the white paper many bitcoiners know this uh
378
00:34:27,764 --> 00:34:37,444
he calls cold card quote super high assurance in a public um post about bit 85 seed derivation
379
00:34:37,444 --> 00:34:46,364
so again another very prominent voice well-respected voice as you know assuming
380
00:34:46,364 --> 00:34:53,144
and promoting this as a high security software um and i'm by the way when i name all these people
381
00:34:53,144 --> 00:34:57,944
i'm not saying that they knew that it was broken and they were they were lying they
382
00:34:57,944 --> 00:35:07,104
genuinely believed what they were saying um and then in uh december of 2024 1031 which is a bitcoin
383
00:35:07,104 --> 00:35:14,124
focused venture fund leads coin kite series a 1031 managing partners are matt odell and marty
384
00:35:14,124 --> 00:35:21,124
bent uh cold cards two most prominent promoters uh you that's a interpretation but i would say
385
00:35:21,124 --> 00:35:27,624
they're probably some of the most prominent people in the in the space um and uh the investment
386
00:35:27,624 --> 00:35:33,924
formalizes what audience had been treating as independent advice and i i would say that you know
387
00:35:33,924 --> 00:35:41,664
I don't know exactly to what extent they were disclosing that while they were promoting cold
388
00:35:41,664 --> 00:35:48,824
card that they also had financial interest in this company that they're promoting and when that
389
00:35:48,824 --> 00:35:53,644
started and this is all kind of murky I'm not trying to throw stones at these people okay we
390
00:35:53,644 --> 00:36:01,404
were all in sort of an echo chamber believing that this was um you know the the vetted software when
391
00:36:01,404 --> 00:36:07,764
it wasn't. But, you know, this kind of gives a picture of how this was building up and how
392
00:36:07,764 --> 00:36:17,244
one layer after the other was basically, it was a self-referential credibility layering.
393
00:36:17,564 --> 00:36:33,280
You know the engineers said it was fine So the educators said it was fine And then the institution you know the educators listened to the engineers and then institutions listen to the educators and then each layer was listening to the layer below it
394
00:36:33,280 --> 00:36:39,200
nobody was actually checking um so they were checking they were checking the wrong things
395
00:36:39,200 --> 00:36:45,960
and they didn't realize what they were not looking at so i want you to continue but one question i
396
00:36:45,960 --> 00:36:53,040
want you to keep in mind for after this is, you know, the AI or if there was AI involved in
397
00:36:53,040 --> 00:37:00,260
discovering this weakness and also if there could be AI involved, which I think there was to discover
398
00:37:00,260 --> 00:37:06,200
how it was being done and what was happening and then how we can use AI in the future to, you know,
399
00:37:06,200 --> 00:37:09,560
because human error, right? So we might not check the correct files. We might not check.
400
00:37:10,120 --> 00:37:14,460
So maybe that, you know, even though AI can be used to discover the bug, it could be used to fix
401
00:37:14,460 --> 00:37:18,800
the bugs before they become a big problem possibly. So I just want to know your thoughts on that.
402
00:37:19,920 --> 00:37:30,020
So I'm seeing that claim a lot. I don't know. I don't think anybody knows whether or not AI was
403
00:37:30,020 --> 00:37:37,140
used. A lot of people keep citing Kimmy. They're saying, oh, well, this open weights model is out
404
00:37:37,140 --> 00:37:43,020
and it's got frontier quality and it doesn't have any guardrails. So obviously that's to blame.
405
00:37:43,020 --> 00:37:45,680
and that may or may not be true.
406
00:37:45,920 --> 00:37:47,140
I don't think there's any proof
407
00:37:47,140 --> 00:37:48,420
on one or the other on that.
408
00:37:49,780 --> 00:37:52,540
I can say with some certainty
409
00:37:52,540 --> 00:37:54,460
that Frontier models,
410
00:37:54,600 --> 00:37:56,060
even the ones with guardrails,
411
00:37:56,160 --> 00:37:58,020
can be relatively easily tricked
412
00:37:58,020 --> 00:37:59,420
to do just about anything
413
00:37:59,420 --> 00:38:00,500
you want them to do
414
00:38:00,500 --> 00:38:03,280
if you frame it in the right way.
415
00:38:03,840 --> 00:38:04,680
So you could say,
416
00:38:05,620 --> 00:38:07,900
even the ones with guardrails,
417
00:38:07,900 --> 00:38:08,700
you could say,
418
00:38:08,700 --> 00:38:14,080
I am concerned that I might lose my Bitcoin
419
00:38:14,080 --> 00:38:18,220
and I want you to help me find some way I might
420
00:38:18,220 --> 00:38:24,340
get into this software to protect my grandma or whatever.
421
00:38:24,900 --> 00:38:27,160
There's any number of ways you can put it.
422
00:38:28,740 --> 00:38:31,020
So what I will say to that end,
423
00:38:31,120 --> 00:38:33,900
so I don't think the AI angle is a really compelling angle.
424
00:38:33,900 --> 00:38:47,520
What I will say, though, is that I do have a message here, which I believe was in their Telegram, the cold card Telegram, or a well-known group chat.
425
00:38:47,520 --> 00:38:52,360
It's by a guy named Legost, and it's April 7th, 2021.
426
00:38:52,880 --> 00:38:57,380
So this is within months of the firmware change.
427
00:38:57,380 --> 00:39:15,760
He says this version 4 firmware appears to be a radical deviation to all prior firmware since 2018 with, quote, all crypto and BIP39 related code replaced by libngu.
428
00:39:15,760 --> 00:39:26,600
and then he questions the credibility of the author of lib ngu and basically says that this is
429
00:39:26,600 --> 00:39:33,020
very inappropriate for them to switch to this library and this was right afterwards so you know
430
00:39:33,020 --> 00:39:40,780
people were already raising the alarm bells as far as the switch at the time and then it goes on
431
00:39:40,780 --> 00:39:47,940
in 2024, we've got Block Unmasked, which is a blockchain forensics and fraud investigation firm.
432
00:39:47,940 --> 00:39:52,880
They are working with multiple victims where Bitcoin had disappeared from their cold card
433
00:39:52,880 --> 00:39:59,320
wallets with no evidence of phishing, malware, or seed exposure. Sounds familiar? They trace the
434
00:39:59,320 --> 00:40:05,840
losses to weak seed entropy, and according to Block Unmasked, file private reports with CoinKite
435
00:40:05,840 --> 00:40:13,280
and law enforcement. This is in 2024. No action, no visible action follows from that.
436
00:40:14,360 --> 00:40:16,980
Wait, so was CoinKite contacted?
437
00:40:17,560 --> 00:40:22,720
That's what they said. They said that they contacted them. And now that's what they said.
438
00:40:22,860 --> 00:40:27,600
I have not verified that personally. I haven't talked to them or anything, but they are claiming
439
00:40:27,600 --> 00:40:34,220
that in their public pronouncements. And that does sound actually exactly familiar. And it is
440
00:40:34,220 --> 00:40:36,500
insane that that was something that
441
00:40:36,500 --> 00:40:37,580
should have been
442
00:40:37,580 --> 00:40:39,800
urgent to
443
00:40:39,800 --> 00:40:42,460
work on people.
444
00:40:42,660 --> 00:40:44,320
Right. So even in
445
00:40:44,320 --> 00:40:45,640
2021, people were
446
00:40:45,640 --> 00:40:47,880
calling this out.
447
00:40:48,000 --> 00:40:49,040
Then in 2024,
448
00:40:50,040 --> 00:40:52,440
they didn't exactly
449
00:40:52,440 --> 00:40:54,380
know what was wrong, but they knew
450
00:40:54,380 --> 00:40:55,760
that there was an entropy problem.
451
00:40:56,560 --> 00:40:58,240
They should have known that there was an entropy
452
00:40:58,240 --> 00:41:00,160
problem. And then NVK
453
00:41:00,160 --> 00:41:01,840
in April 2025
454
00:41:01,840 --> 00:41:03,140
was
455
00:41:04,220 --> 00:41:13,080
says in a Bitcoin magazine interview one month before an audit, he says, quote, since version one,
456
00:41:13,080 --> 00:41:17,740
we always released schematics so people can go and build it themselves and prove the things.
457
00:41:17,860 --> 00:41:24,420
Because the whole point for us is provability. Every claim we make, we need to be able to
458
00:41:24,420 --> 00:41:29,740
substantiate in a way that the users can prove it themselves, end quote. So that's very recently.
459
00:41:29,740 --> 00:41:30,820
That's last year.
460
00:41:31,260 --> 00:41:32,620
They're still selling.
461
00:41:32,760 --> 00:41:36,240
This is verifiability as a product, okay?
462
00:41:36,580 --> 00:41:39,960
And then May 2025, O'Bear.
463
00:41:39,997 --> 00:41:41,417
Bearn audits cold card firmware.
464
00:41:41,797 --> 00:41:45,577
This is James O'Bearn, again, very well-known core contributor.
465
00:41:45,957 --> 00:41:47,257
I respect this man.
466
00:41:48,357 --> 00:41:51,217
I think he is a very smart guy.
467
00:41:51,217 --> 00:41:58,497
And he audits cold card firmware to determine where the RNG random number generator is actually sourced.
468
00:41:58,897 --> 00:42:01,497
He finds it routing through libngu.
469
00:42:02,337 --> 00:42:09,957
He points out at the time that that library, libngu, has six GitHub stars and one maintainer
470
00:42:09,957 --> 00:42:14,857
who may or may not be associated with CoinKite, though there is some evidence to point to that.
471
00:42:15,297 --> 00:42:18,437
He flags the hard-coded constants as sloppy.
472
00:42:18,437 --> 00:42:25,097
He questions whether the hardware RNG is actually running, and he advises CoinKite to remove
473
00:42:25,097 --> 00:42:33,677
libngu and link against libsep256k1 which is the standard crypto primitive library in bitcoin
474
00:42:33,677 --> 00:42:42,597
the same one that bitcoin core uses and is highly audited or reviewed when this was may 2025
475
00:42:42,597 --> 00:42:51,157
um coin kite's response was quote if something were wrong we'd already know about it by now
476
00:42:51,157 --> 00:42:55,597
everything is properly configured for the production boards end quote
477
00:42:55,597 --> 00:43:03,437
obeirne does not push further and in august 4th just uh today and this was was this specifically
478
00:43:03,437 --> 00:43:10,137
for the devices the mk4 you said for this was their firmware the so it would be all their devices
479
00:43:10,137 --> 00:43:15,137
yeah all their devices use this firmware yeah and then then today literally today august 4th he
480
00:43:15,137 --> 00:43:19,197
He calls not following up on this, quote, a horrible mistake.
481
00:43:19,497 --> 00:43:21,717
So he was like, he called them out on it.
482
00:43:21,957 --> 00:43:23,317
He told them they needed to check.
483
00:43:24,077 --> 00:43:27,477
And he just, he left it at that point.
484
00:43:27,557 --> 00:43:31,417
And I was like, okay, you know, that's it, I guess.
485
00:43:32,317 --> 00:43:38,077
And he, I guess, is like many other people looking back in regret for not doing more.
486
00:43:38,077 --> 00:43:40,897
And I'm, again, I'm not throwing stones at him or anything.
487
00:43:41,357 --> 00:43:43,417
Good on him for at least calling him out.
488
00:43:43,417 --> 00:43:54,837
And I'm sure that and the block unmasked incident and other instances will come out in court at some point.
489
00:43:55,837 --> 00:43:58,657
You know, what was known and when and who knew it.
490
00:43:58,657 --> 00:44:01,357
And, you know, that I'm not involved in any of that.
491
00:44:01,517 --> 00:44:12,797
But this is establishing a record that is going to be very important for people trying to claw back funds or whatever is going to happen.
492
00:44:12,857 --> 00:44:13,377
I don't know.
493
00:44:13,417 --> 00:44:16,837
maybe a class action at this point you're right because this point it's not
494
00:44:16,837 --> 00:44:23,177
it's not like oh i just lost my bitcoin and it was it was like this is like you said the biggest
495
00:44:23,177 --> 00:44:32,637
hack or problem or incident for self-custody yeah this is the worst uh disaster that's ever
496
00:44:32,637 --> 00:44:37,797
occurred in my in my view so that's not going to just disappear just like no one's gonna
497
00:44:37,797 --> 00:44:42,357
have to account for it or what happened you know there's going to be questions
498
00:44:42,357 --> 00:44:45,537
you know i think you're right about that there will be questions because
499
00:44:45,537 --> 00:44:51,697
again all the people that were affected were the plebs not all of them but you know what i mean like
500
00:44:51,697 --> 00:44:57,877
the people that trusted they did the best they could they did what they thought was the most
501
00:44:57,877 --> 00:45:03,837
secure and and um and it could have been their life savings you know that that it's just gone
502
00:45:03,837 --> 00:45:12,917
um i'll add uh it gets it gets worse uh so uh why am i like i just can't believe it that's why i'm
503
00:45:12,917 --> 00:45:18,637
like yeah i honestly i couldn't either i think everybody's pretty much in shock about this um
504
00:45:18,637 --> 00:45:23,617
so there's there was a outlet called wallet scrutiny uh project that verifies whether
505
00:45:23,617 --> 00:45:29,437
hardware wallet firmware builds are reproducible they reported that older card cold card uh
506
00:45:29,437 --> 00:45:36,457
releases could not be independently reproduced. And NVK responded publicly, quote, reminder that
507
00:45:36,457 --> 00:45:42,857
wallet scrutiny is either incompetent or malicious, end quote, followed by immediately by, quote,
508
00:45:42,957 --> 00:45:49,957
is it time to sue them? End quote. I was gonna say this before you said this specific
509
00:45:49,957 --> 00:45:57,917
quote or what this is. I was thinking about exactly that, about how people in Bitcoin should
510
00:45:57,917 --> 00:46:04,037
drop the ego because it almost feels like a defense mechanism somebody's trying to tell you
511
00:46:04,037 --> 00:46:09,797
hey I see something wrong and you're like are you just you know what I mean trying to attack me like
512
00:46:09,797 --> 00:46:16,717
you know what I mean it becomes like this thing where um you start to think everybody is trying to
513
00:46:16,717 --> 00:46:23,017
just hate on you or something and you're just like dismissive of all the of all the hate and
514
00:46:23,017 --> 00:46:28,397
are just trying to not care and not listen and move forward. But it's like, you should pay
515
00:46:28,397 --> 00:46:34,617
attention. Not everyone's just trying to be, you know, like these are legitimate concerns.
516
00:46:35,737 --> 00:46:41,057
And you need to let go not just him, but anybody listening in life. You know, it's like something
517
00:46:41,057 --> 00:46:45,697
where as Bitcoiners, we should have that. I think that most Bitcoiners should have this,
518
00:46:46,297 --> 00:46:56,693
you know ego death sort of a thing about them because you have to realize you were wrong about so much about how the world works It painful to be wrong I think that the hard part
519
00:46:57,633 --> 00:47:04,413
Right. Yeah. So, I mean, we've already, anyone who understands this has gone through that one time, two times, many times in our life about different things.
520
00:47:04,493 --> 00:47:10,893
But I think when it comes to creating a product or a company, people just get, it's like their baby and they get really defensive.
521
00:47:10,893 --> 00:47:37,113
And it's like this becomes your identity. And now you are tied to this thing as a reputation, you know, and you're trying to uphold that. And you have people that have counted on you. But that's even the more reason why you should take things seriously. You have people counting on you. You have trust. Like, look how many people supported this wallet, you know? I just am blown away how people just blindly support it.
522
00:47:37,113 --> 00:47:51,453
Well, yeah, then I will apply that same sort of lens to Bitcoin itself, Bitcoin's own source code, other wallets, other security software, you know, just the whole gamut.
523
00:47:51,453 --> 00:47:55,713
And I think really what it boils down to is people respond to incentives.
524
00:47:56,493 --> 00:48:00,233
And quite often, they're not responding to reason.
525
00:48:00,433 --> 00:48:01,813
They're responding to status.
526
00:48:02,273 --> 00:48:06,173
They're responding to, you know, purely incentives, who's paying them.
527
00:48:06,173 --> 00:48:23,733
Or they're responding to, and this is a little bit deeper, but maintaining their own trauma. They don't want to, you know, confront prior trauma. And so it's easier to dismiss others rather than confront that. But that's kind of a separate issue.
528
00:48:23,733 --> 00:48:28,353
What I'll say on this front is this has been a very traumatizing event for everybody.
529
00:48:28,613 --> 00:48:36,753
And if, you know, you are in this position where you've lost money or you know somebody who's lost money, please check on them.
530
00:48:37,553 --> 00:48:40,393
And don't leave them alone.
531
00:48:41,493 --> 00:48:51,273
And, you know, make sure that everybody's getting support in this very dark chapter in Bitcoin right now.
532
00:48:51,273 --> 00:48:54,893
So, you know, that's because they did everything right.
533
00:48:55,773 --> 00:48:56,613
They did everything right.
534
00:48:56,673 --> 00:48:58,273
They did everything they were supposed to.
535
00:48:59,013 --> 00:49:02,653
You know, it's not our job as people in Bitcoin.
536
00:49:02,813 --> 00:49:05,013
I mean, we do believe don't trust verify.
537
00:49:05,193 --> 00:49:14,973
But when enough people that you learned from, that you found out about Bitcoin from, that have educated you on Bitcoin are telling you this is the best thing to do.
538
00:49:15,273 --> 00:49:21,193
And if they're not technologists, they're not going to they're not going to be the ones to find find the buck, you know.
539
00:49:21,273 --> 00:49:30,853
And if you're one of those people, just rest assured you can rebuild your life. It's not over. You can start again.
540
00:49:30,853 --> 00:49:46,633
I saw a post that was so inspiring and I want to share it. I feel like he is an amazing person. He made a post on X that he lost, I don't know, one and a half million dollars.
541
00:49:46,913 --> 00:49:47,653
I saw that one, yeah.
542
00:49:47,653 --> 00:49:57,013
You did. And then at the end of this, his story, he said at the bottom, you know, but this is I'm going to keep building. I'm an entrepreneur.
543
00:49:57,433 --> 00:49:57,933
I'll make it back.
544
00:49:58,633 --> 00:50:16,393
Yeah, I'll make it back. I'm going to keep working. I'm like, that was beautiful. And what what an amazing person that can just realize, yeah, I'm going to keep working. I'm going to keep it's not like to not just beat yourself up about it and just sit there staring at the blockchain explorer.
545
00:50:16,393 --> 00:50:24,113
watching yeah i thought that was spend your money yeah that was don't do that right and i do think
546
00:50:24,113 --> 00:50:29,053
that the people that's been the other beautiful thing knowing that how many people like you have
547
00:50:29,053 --> 00:50:35,453
reached out and helped friends secure their um their their bitcoin and being able to make sure
548
00:50:35,453 --> 00:50:40,473
that they feel safe and they don't have to worry so people like you i'm so grateful for in this
549
00:50:40,473 --> 00:50:46,333
community that are out here speaking and letting people know um you know what happened also so
550
00:50:46,333 --> 00:50:53,573
that we can do things better you know um so there's been people even in that i've that i know
551
00:50:53,573 --> 00:50:59,713
have bitcoin in real life people have asked me i've asked them like how were you affected are
552
00:50:59,713 --> 00:51:04,233
you okay you know everyone's checking on each other and we're all asking everyone that we know
553
00:51:04,233 --> 00:51:08,913
or you know that we see in real life for like how they're doing and i haven't known anyone that's
554
00:51:08,913 --> 00:51:18,333
been affected um unless you know just online but i haven't um i i hope that they're gonna take that
555
00:51:18,333 --> 00:51:22,053
guy's which i do want to find his post but i hope that they're gonna take his lead and just keep
556
00:51:22,053 --> 00:51:29,333
positive and keep moving keep building um and i actually you know i had i had shared some news
557
00:51:29,333 --> 00:51:34,693
about you know being able to work in a space where we are building a better future and more open and
558
00:51:34,693 --> 00:51:41,853
decentralized future and being able to have Bitcoin as payment and being able to empower
559
00:51:41,853 --> 00:51:45,153
merchants. And that's what I thought of when I thought of him saying he's an entrepreneur.
560
00:51:45,793 --> 00:51:49,313
I was like, anyone who's created something or selling something, I'm like, there's a future
561
00:51:49,313 --> 00:51:55,973
out there where you can have and retain more of your sales and you can find a job in Bitcoin with
562
00:51:55,973 --> 00:51:59,973
whatever skills you have. I feel like we are still so early. You can do so much in this space. So I
563
00:51:59,973 --> 00:52:10,569
I think that for anyone who thinks well what am I even gonna do There so much opportunity and you can grow So that what I happy about And you know we were all we all here for
564
00:52:10,569 --> 00:52:17,169
each other and I'm so grateful to talk about this. But let me just, let me just say that just for
565
00:52:17,169 --> 00:52:21,729
anybody who isn't already in the know and they were wondering what the heck we're talking about.
566
00:52:22,289 --> 00:52:29,769
So I was just, this, all this leads up to the hack itself. And I'll just add real quickly,
567
00:52:29,769 --> 00:52:37,889
In May of 2026, Stefan Lavera covers the MK5 launch on his podcast with NDK.
568
00:52:38,549 --> 00:52:46,629
BTC Sessions, popular Bitcoin YouTube educator, publishes a MK5 walkthrough, calls it the gold standard.
569
00:52:47,489 --> 00:52:51,449
Framing continues right up until the drain, which occurs, starts.
570
00:52:53,089 --> 00:52:55,989
And again, I'm not calling any of these people out specifically.
571
00:52:56,149 --> 00:52:57,909
I'm not saying they're bad people or whatever.
572
00:52:57,909 --> 00:53:09,969
I'm just pointing out that there was a social cohesion around this wallet and prominent voices, ones that many Bitcoiners recognize were all saying the same thing, singing to the choir.
573
00:53:09,969 --> 00:53:20,989
And it comes up to July 30th, 594 Bitcoin was swept from 500 addresses in 41 minutes.
574
00:53:21,809 --> 00:53:23,589
Largest wallets were hit first.
575
00:53:23,869 --> 00:53:27,269
The attacker had already sorted by balance.
576
00:53:27,909 --> 00:53:31,409
So these wallets had obviously been scoped out.
577
00:53:33,269 --> 00:53:36,309
And just to kind of give people a little background on that,
578
00:53:36,689 --> 00:53:41,209
essentially when this device has poor entropy,
579
00:53:41,369 --> 00:53:42,929
when the cold card had poor entropy,
580
00:53:43,209 --> 00:53:45,609
when you generated your seed phrase from it,
581
00:53:46,249 --> 00:53:49,829
it's effectively like everybody got the same seed phrase.
582
00:53:49,969 --> 00:53:51,809
It's like if you go to a password generator
583
00:53:51,809 --> 00:53:54,309
and everybody generates the same password.
584
00:53:54,309 --> 00:54:22,349
And then, you know, they think that they've generated a normal seed phrase or a normal password. But if everyone has the same one, then it's not particularly secure. And the attackers had clearly scoped out, they'd figured out what that private key was, they'd scoped out all these addresses, they'd found the biggest ones. And then in a 41 minute time period, they sweep 500 addresses and get 594 Bitcoin.
585
00:54:22,349 --> 00:54:24,729
same day
586
00:54:24,729 --> 00:54:27,389
NVK
587
00:54:27,389 --> 00:54:28,329
posts publicly
588
00:54:28,329 --> 00:54:28,969
quote
589
00:54:28,969 --> 00:54:30,549
no need to panic
590
00:54:30,549 --> 00:54:31,389
end quote
591
00:54:31,389 --> 00:54:32,749
and blames a
592
00:54:32,749 --> 00:54:33,069
quote
593
00:54:33,069 --> 00:54:34,749
bad or leaked seed
594
00:54:34,749 --> 00:54:35,569
end quote
595
00:54:35,569 --> 00:54:37,249
links a victim's
596
00:54:37,249 --> 00:54:38,109
own transaction
597
00:54:38,109 --> 00:54:39,469
as supposed evidence
598
00:54:39,469 --> 00:54:40,489
this post
599
00:54:40,489 --> 00:54:41,629
is later deleted
600
00:54:41,629 --> 00:54:42,109
July
601
00:54:42,109 --> 00:54:42,949
and that's another thing
602
00:54:42,949 --> 00:54:43,689
it's like there's some
603
00:54:43,689 --> 00:54:44,189
kind of like
604
00:54:44,189 --> 00:54:44,929
hey there's something
605
00:54:44,929 --> 00:54:46,129
happening and it was like
606
00:54:46,129 --> 00:54:48,169
I don't want the optics
607
00:54:48,169 --> 00:54:48,909
to show
608
00:54:48,909 --> 00:54:50,509
that this looks bad
609
00:54:50,509 --> 00:54:50,969
on us
610
00:54:50,969 --> 00:54:51,869
instead of just
611
00:54:51,869 --> 00:54:52,289
checking
612
00:54:52,289 --> 00:55:03,249
Right, exactly. Well, it's the same pattern. From the very beginning up to the very moment, he just couldn't get over himself, basically.
613
00:55:03,249 --> 00:55:09,849
and that gets worse in that respect.
614
00:55:10,629 --> 00:55:14,829
But first in same day,
615
00:55:15,429 --> 00:55:18,989
Kevin Loet, I may be butchering your name,
616
00:55:19,089 --> 00:55:20,369
so I'm sorry if you listen,
617
00:55:20,369 --> 00:55:23,629
co-founder of Wizards Ardene,
618
00:55:24,129 --> 00:55:25,929
a Bitcoin wallet security firm,
619
00:55:26,109 --> 00:55:30,509
is the first to publicly hypothesize an RNG defect.
620
00:55:30,509 --> 00:55:37,449
Rob Hamilton, CEO of Anchor Watch, is the first to say entropy and publish on-chain analysis.
621
00:55:38,289 --> 00:55:47,709
And O'Bearn sends the widest reach warning, which had 423,000 views 64 minutes before the official advisory from CoinKite.
622
00:55:47,709 --> 00:56:00,649
Greg Sanders, known as Instagibs, a Bitcoin core contributor at Spiral, blocks Bitcoin division, independently reproduces the vulnerability 13 minutes before CoinKite's advisory.
623
00:56:01,869 --> 00:56:05,949
And then an official advisory happens.
624
00:56:06,389 --> 00:56:11,229
They point out, okay, this isn't a bad or leaked seed.
625
00:56:11,229 --> 00:56:14,809
This is actually a vulnerability.
626
00:56:14,809 --> 00:56:17,889
and then they advise everybody to start moving their funds.
627
00:56:18,509 --> 00:56:24,549
And if I remember correctly, they focused on the MK3 first,
628
00:56:24,649 --> 00:56:29,249
and it wasn't clear if the MK4 and the cold car queue were affected yet.
629
00:56:29,909 --> 00:56:31,489
It later came out that they were.
630
00:56:31,569 --> 00:56:34,889
The entropy was a little bit stronger on the later devices,
631
00:56:35,249 --> 00:56:38,689
but nowhere near what it should have been.
632
00:56:38,689 --> 00:56:49,889
and then on july 31st nvk says quote i am sorry i and i am devastated
633
00:56:49,889 --> 00:56:58,629
this is our core responsibility and we fell short end quote one sentence on accountability
634
00:56:58,629 --> 00:57:05,049
and then he shifts to attributing the discovery to ai framing it as an unprecedented attack
635
00:57:05,049 --> 00:57:08,169
rather than a findable bug in a public repo,
636
00:57:08,709 --> 00:57:23,486
which again gets back to this lack of accountability Even when he apologizing he still shifting blame onto AI or somebody else or something else which I think was a very bad look personally
637
00:57:24,886 --> 00:57:40,486
And that's probably why people are talking about that, you know, Kimmy AI, because maybe it's like, and who knows if it's true, but it's just, it's easier to blame some advanced, you know, advanced AI for figuring something out, you know.
638
00:57:40,486 --> 00:57:54,526
Well, and just go back to what I said earlier, like you've got somebody in the Telegram channel in 2021 calling this out, saying this was a mistake. This wasn't vetted. You're moving away from vetted software.
639
00:57:54,526 --> 00:58:14,646
And then you've got that, what was it, Block Unmasked, the forensics and fraud investigation firm saying explicitly that they believe that there was weak seed entropy involved in multiple cases that they were following.
640
00:58:14,646 --> 00:58:35,226
And then O'Bearn in 2025, he explicitly says, you know, that they should, he advises them as a well-known credentialed Bitcoin contributor to remove LibNGU and link the proper library.
641
00:58:36,026 --> 00:58:40,066
And they say, well, if there was something wrong, we'd already know about it.
642
00:58:40,066 --> 00:58:49,326
So it's rich for them to blame AI when they had multiple warnings clearly and just didn't seem to care.
643
00:58:49,986 --> 00:58:58,826
And that leads to this other theory, which I am not pushing, by the way, but other people are suggesting that this might have been an inside job.
644
00:58:59,046 --> 00:59:01,646
I don't know. I have no idea.
645
00:59:02,246 --> 00:59:07,506
I think that's honestly, it's as likely as anything else.
646
00:59:07,506 --> 00:59:19,206
You know, whether it's stupidity or malice, I'm, you know, it could be either way at this point. Neither is good. I think maybe the consequences will be pretty severe.
647
00:59:19,206 --> 00:59:27,146
That's why people start to go that direction, whether it's true or not, because of the way that it was dismissed and ignored.
648
00:59:27,646 --> 00:59:31,686
Like nobody that would actually care about this happening.
649
00:59:33,826 --> 00:59:40,226
But if it was your company, you would want like what you're doing, like you're reaching out, you were saved.
650
00:59:41,246 --> 00:59:42,826
You know, your Bitcoin was saved.
651
00:59:42,946 --> 00:59:44,386
Novosapiens helped you secure it.
652
00:59:44,466 --> 00:59:47,406
You know, you want to help other people secure theirs and save theirs.
653
00:59:47,406 --> 01:00:05,086
And then you have people who are building things who just have that attitude about them that makes you question. That's why I think people are, even if it's not true, that's, I think, why they are even asking these questions, because it's like, why would you ignore it? Why would you not cure?
654
01:00:05,086 --> 01:00:19,186
Well, I'll point out to their credit, at least from their public pronouncements, from what I've seen, they're destroying their inventory. That's what they say. I don't know if that's true or not. They're saying they're destroying their inventory, so they're not going to ship out any more devices with the faulty firmware.
655
01:00:19,186 --> 01:00:26,066
as far as I understand it, their team is actively supporting people who are coming to them with
656
01:00:26,066 --> 01:00:32,986
issues. But given the scale of this, it's really the burden of it has been put on the entire
657
01:00:32,986 --> 01:00:39,606
community. And as you pointed out earlier, the community, to their credit, is the network itself
658
01:00:39,606 --> 01:00:45,846
is standing up and trying to save as many people as they can, which is fantastic.
659
01:00:45,846 --> 01:00:55,546
but still that it's still ongoing august 1st which was just a few days ago another wave
660
01:00:55,546 --> 01:01:06,246
1100 bitcoin approximately 75 million from 2600 addresses stolen swept next day wave three 1300
661
01:01:06,246 --> 01:01:15,506
yeah 1300 bitcoin approximately uh 1889 million from 4500 addresses that was sunday that was the
662
01:01:15,506 --> 01:01:24,486
second yeah that was the second right um and then august 3rd uh yesterday wave 4 uh sweep rate 45x
663
01:01:24,486 --> 01:01:35,886
pre-incident baseline 1500 bitcoin getting us to 100 million total um and now 7300 addresses so um
664
01:01:35,886 --> 01:01:43,286
yeah and so today i think galaxy research has pointed out there are 15 independent attackers
665
01:01:43,286 --> 01:01:45,386
confirmed. Opportunists are piling
666
01:01:45,386 --> 01:01:47,426
in. They're
667
01:01:47,426 --> 01:01:49,386
estimating between 100 million.
668
01:01:49,466 --> 01:01:51,166
I've seen some estimates as high as
669
01:01:51,166 --> 01:01:51,626
130.
670
01:01:53,666 --> 01:01:55,186
CoinKite is,
671
01:01:55,606 --> 01:01:56,966
like I said, they're
672
01:01:56,966 --> 01:01:59,386
claiming they halted shipments, destroying
673
01:01:59,386 --> 01:02:00,286
remaining inventory.
674
01:02:01,346 --> 01:02:02,926
And, you know, this is
675
01:02:02,926 --> 01:02:04,246
just where we're at now.
676
01:02:05,746 --> 01:02:07,266
I think it's, well, how
677
01:02:07,266 --> 01:02:09,606
you even said opportunists
678
01:02:09,606 --> 01:02:10,866
are piling in too because
679
01:02:10,866 --> 01:02:13,126
now they know
680
01:02:13,126 --> 01:02:16,886
there's some kind of vulnerability so that more people are attempting.
681
01:02:17,246 --> 01:02:23,946
And then what I thought, I think here's what I was going to say earlier about what I thought
682
01:02:23,946 --> 01:02:29,306
was so wild to me was because I was seeing that the vulnerability was in certain devices.
683
01:02:29,982 --> 01:02:41,182
And then when I saw, you know, everybody, like big accounts, people that, you know, a lot of names that people follow, people from CoinKite, I guess.
684
01:02:41,182 --> 01:02:50,182
I think it was mainly like I remember Jack Maulers and like certain other people sharing, just get all of your Bitcoin off of cold card.
685
01:02:50,722 --> 01:02:54,902
Like all cold, they didn't say this device specifically, they just said move it away from cold card, period.
686
01:02:54,902 --> 01:03:15,862
And I think that was just seeing big voices like that say that, just move away, move away, get it out. I was shocked because I'm like, okay, if this is a vulnerability in certain devices, then why is it that there's suggestion to get away from the entire line of devices?
687
01:03:15,862 --> 01:03:19,722
and then I think that's why people started to question
688
01:03:19,722 --> 01:03:24,122
any other single signature if you didn't roll your own dice.
689
01:03:24,882 --> 01:03:29,642
They started to question the safety of any cold storage setup that they had
690
01:03:29,642 --> 01:03:30,782
and where did they move coins.
691
01:03:31,002 --> 01:03:34,282
So people reached out to me, what do you think about this?
692
01:03:34,322 --> 01:03:35,022
What should I do?
693
01:03:36,942 --> 01:03:41,682
And then I was just suggesting multi-sig setups
694
01:03:41,682 --> 01:03:45,902
or for now a trusted exchange if you're on a cold card.
695
01:03:46,382 --> 01:03:49,362
But I was just, I was questioning too,
696
01:03:49,422 --> 01:03:51,702
do I advise people just to stick with anybody else
697
01:03:51,702 --> 01:03:52,522
who's not cold card?
698
01:03:52,602 --> 01:03:54,722
Because then if people are already worrying about
699
01:03:54,722 --> 01:03:57,302
every device from their line,
700
01:03:57,422 --> 01:03:58,542
why is it that people are,
701
01:03:59,042 --> 01:04:00,922
it's making people concerned.
702
01:04:01,722 --> 01:04:03,342
As it should, completely understandable.
703
01:04:03,702 --> 01:04:04,942
Yeah, no, 100%.
704
01:04:04,942 --> 01:04:06,842
And then of course the question is like,
705
01:04:06,942 --> 01:04:08,762
well, if this went on for this long
706
01:04:08,762 --> 01:04:10,202
with this prominent wallet,
707
01:04:10,202 --> 01:04:13,322
But what's going on with the other wallets, right?
708
01:04:14,642 --> 01:04:15,162
And so...
709
01:04:15,162 --> 01:04:17,642
It's like you said, it wasn't just the initial code
710
01:04:17,642 --> 01:04:20,822
that everyone audited and verified and checked,
711
01:04:20,902 --> 01:04:23,442
but then it's every implementation that comes out after it.
712
01:04:23,722 --> 01:04:26,802
Every, you know, that's the thing.
713
01:04:27,002 --> 01:04:30,222
There was software updates and it's scary.
714
01:04:30,342 --> 01:04:31,682
I mean, that's the scary part.
715
01:04:31,682 --> 01:04:34,682
So that's why it's like the thing about...
716
01:04:35,462 --> 01:04:37,502
I mean, I like that what NovaSapiens is doing
717
01:04:37,502 --> 01:04:43,702
with making it easier for the people who are going to verify for themselves and make sure they
718
01:04:43,702 --> 01:04:49,762
are a full-on Bitcoin or they want to do the right thing. I think this is definitely a great way. Like,
719
01:04:49,782 --> 01:04:54,602
I would love to get that myself because I just thought, I think it's nice to have a simpler
720
01:04:54,602 --> 01:05:02,002
version. But then the people who aren't going to do that, I mean, you need, we need to just be,
721
01:05:02,162 --> 01:05:06,442
or the community needs to be better, just making better products for people so we don't have to be
722
01:05:06,442 --> 01:05:08,502
in fear and being able to check and know.
723
01:05:09,102 --> 01:05:10,522
I just don't know how,
724
01:05:11,442 --> 01:05:15,402
if this has something to do with Bitcoin Commons
725
01:05:15,402 --> 01:05:17,842
and with the governance of Bitcoin.
726
01:05:18,662 --> 01:05:19,842
Sure, yeah.
727
01:05:20,082 --> 01:05:22,462
Because, yeah, like how, can you explain?
728
01:05:22,562 --> 01:05:24,142
Well, I'll get into that.
729
01:05:24,962 --> 01:05:28,022
Just to kind of dovetail off of your last point, though.
730
01:05:28,082 --> 01:05:29,002
Yeah, you don't have to answer that.
731
01:05:29,002 --> 01:05:30,102
Well, it is an important point.
732
01:05:30,582 --> 01:05:34,482
And, you know, I guess I'm a Bitcoin OG.
733
01:05:34,482 --> 01:05:37,802
and I didn't really think of myself that way
734
01:05:37,802 --> 01:05:41,042
until very recently when I recovered Bitcoin from 2010.
735
01:05:41,862 --> 01:05:45,982
And I perhaps in an ill-advised fashion
736
01:05:45,982 --> 01:05:48,862
immediately went public with that fact,
737
01:05:49,202 --> 01:05:51,422
which I thought and still think
738
01:05:51,422 --> 01:05:54,502
was actually protecting me to be public.
739
01:05:55,502 --> 01:06:00,082
So I have a microphone where I can scream from the rooftops
740
01:06:00,082 --> 01:06:00,962
if something is happening,
741
01:06:00,962 --> 01:06:04,282
but it also, you know, there's a risk associated
742
01:06:04,282 --> 01:06:14,262
with that. And, you know, people make assumptions and that's fine. But because of that, I almost as
743
01:06:14,262 --> 01:06:20,402
soon as I got involved in this industry, I started asking myself the very same question. That was
744
01:06:20,402 --> 01:06:25,842
before any of this happened. I'm sitting there wondering, you know, like, I got to worry about
745
01:06:25,842 --> 01:06:34,722
wrench attacks. You know, I'm publicly at the Strategy Hub, which is a co-working space hosted
746
01:06:34,722 --> 01:06:39,782
by Strategy in Tyson's Corner, Virginia. And that's on my website. Okay. It's like, it's not
747
01:06:39,782 --> 01:06:45,942
hard to figure out. And I tell people that routinely on podcasts and whatnot. So these are
748
01:06:45,942 --> 01:06:54,482
considerations that I have. So, and I, you know, from that perspective, I really like what Nova
749
01:06:54,482 --> 01:06:59,522
Sapien is doing with their self-custody kits. And I'm not going to make recommendations,
750
01:06:59,722 --> 01:07:05,982
but I'll tell you what they recommend. And they recommend a single SIG plus passphrase as a bare
751
01:07:05,982 --> 01:07:13,862
minimum for any amount of Bitcoin over, say, $100. And single SIG plus passphrase is very similar to
752
01:07:13,862 --> 01:07:20,402
like a, it's not multi-SIG, but it's functionally equivalent to say a two of two multi-SIG.
753
01:07:20,862 --> 01:07:25,462
So you have to have your single SIG and you have to have your passphrase in order to access
754
01:07:25,462 --> 01:07:26,002
your Bitcoin.
755
01:07:26,402 --> 01:07:27,402
And that's what they recommend.
756
01:07:27,582 --> 01:07:32,862
And I think for most Bitcoiners who want to do self-custody, that is more than sufficient.
757
01:07:33,302 --> 01:07:49,085
If you store it in three separate places you store your single signature seed phrases in three separate geographically distinct locations in case you know there a terrible fire
758
01:07:49,545 --> 01:07:52,605
So can you explain that for anybody who doesn't understand what that means?
759
01:07:52,685 --> 01:07:56,725
So when you're separating your seed phrase into three locations or your device,
760
01:07:56,965 --> 01:07:58,725
or how do you separate that for people?
761
01:07:58,725 --> 01:07:59,405
Sure, yeah.
762
01:07:59,565 --> 01:08:02,965
So a seed phrase is 12 or 24 words.
763
01:08:03,285 --> 01:08:06,265
It's just words, literally, and it comes from a word list,
764
01:08:06,265 --> 01:08:13,565
a predefined word list. And you roll your dice and that gives you your words. And then you literally
765
01:08:13,565 --> 01:08:18,745
write it down on a little piece of paper. And you get your 12 words and that's a single sig.
766
01:08:19,005 --> 01:08:25,725
Or you can do 24 words and that is really just more words. It's not, I would argue it's really
767
01:08:25,725 --> 01:08:31,585
not that much more secure, but let's just say 12 words. And you've got your 12 words. Then you copy
768
01:08:31,585 --> 01:08:36,225
your 12 words onto two other pieces of paper. And then you've got three pieces of paper with
769
01:08:36,225 --> 01:08:40,885
12 words on them, and then you store one here, one over there, and one some third place.
770
01:08:41,405 --> 01:08:41,545
Okay?
771
01:08:41,725 --> 01:08:43,145
It's really that simple.
772
01:08:43,465 --> 01:08:48,325
And as long as they're not all in the same place, then you can be reasonably assured that
773
01:08:48,325 --> 01:08:56,225
if there's some major disaster or an EMP or a tidal wave or a fire or whatever happens,
774
01:08:56,225 --> 01:09:00,885
then at least one of your pass raises will survive.
775
01:09:01,525 --> 01:09:01,965
I love that.
776
01:09:01,965 --> 01:09:09,305
So instead of, because by separating them geographically, you're able to create a multi-sig.
777
01:09:09,305 --> 01:09:16,125
So when people say they get, like with cold card, people would suggest to create a multi-sig on a cold card.
778
01:09:17,085 --> 01:09:25,885
But so with other wallets, can you separate your seed phrase in order to make, like, do you have to have it separated through the device so that you know that there's a multi-sig setup?
779
01:09:25,885 --> 01:09:34,885
So at least with the SeatSigner, the SeatSigner device doesn't actually store your private keys except very temporarily when you're making transactions.
780
01:09:35,305 --> 01:09:36,925
And it doesn't save them permanently.
781
01:09:37,265 --> 01:09:41,185
And when you unplug the device, the device actually forgets your private keys.
782
01:09:41,185 --> 01:09:51,485
And I could say, load my private keys onto a SeedSigner, make a transaction, unplug it, and then give you the SeedSigner that I just used to make a transaction with my private keys.
783
01:09:51,805 --> 01:09:59,145
And there would be no risk that you could extract my private keys from that SeedSigner once you plugged it in again, which is very...
784
01:09:59,145 --> 01:10:02,085
Is that because of the passphrase?
785
01:10:03,005 --> 01:10:06,785
No, that's just because of the fact that SeedSigner uses volatile memory.
786
01:10:06,905 --> 01:10:09,405
And then when you unplug it, it clears the memory entirely.
787
01:10:10,065 --> 01:10:12,865
Now, with respect to the passphrase, it's just so we're clear.
788
01:10:13,265 --> 01:10:17,685
So let's say you do your seed phrase, you've got your 12 words,
789
01:10:17,685 --> 01:10:21,225
and then you distribute them in three separate geographic locations.
790
01:10:21,785 --> 01:10:22,865
Now, that's single signature.
791
01:10:23,205 --> 01:10:26,285
And Novosapien would say you don't want to put any more than, say,
792
01:10:26,305 --> 01:10:30,305
$100 worth of Bitcoin on a 12-word single signature wallet.
793
01:10:31,045 --> 01:10:34,485
What you then do is you do a passphrase,
794
01:10:34,665 --> 01:10:37,365
and that's called a single signature plus passphrase.
795
01:10:37,365 --> 01:10:44,245
A passphrase is kind of like a 13th word or a 25th word on your single signature.
796
01:10:45,145 --> 01:10:50,325
It doesn't need to be a word from the word list.
797
01:10:51,165 --> 01:10:54,285
It can be like any set of characters.
798
01:10:54,465 --> 01:11:01,005
It can be like a password, just an arbitrary set of numbers and characters, whatever you want.
799
01:11:01,005 --> 01:11:07,605
But Novosapine, I think, recommends four words, all lowercase, with no special characters or spaces.
800
01:11:08,545 --> 01:11:09,505
That's just their recommendation.
801
01:11:09,785 --> 01:11:11,385
But you can put whatever you want in your passphrase.
802
01:11:11,385 --> 01:11:18,165
And then you have to store your passphrase separately in three separate geographic locations.
803
01:11:18,485 --> 01:11:23,185
And then you can't store them in the same location as your single signature, okay?
804
01:11:23,285 --> 01:11:27,145
Because if someone were to compromise any one of those locations,
805
01:11:27,145 --> 01:11:31,785
then you don't want them to have both your signal signature and your passphrase,
806
01:11:31,965 --> 01:11:33,165
because then they have your Bitcoin.
807
01:11:33,605 --> 01:11:35,825
So essentially you need six separate locations,
808
01:11:36,005 --> 01:11:37,805
three locations for your signal signature,
809
01:11:38,225 --> 01:11:40,645
and then three locations for your passphrase,
810
01:11:41,005 --> 01:11:44,425
and then the twain shall not meet until you need to make a transaction,
811
01:11:44,785 --> 01:11:47,365
which if you're talking about your life savings,
812
01:11:47,785 --> 01:11:49,605
then that's probably pretty infrequently.
813
01:11:50,605 --> 01:11:53,525
So that's essentially his method, and that's a pretty complicated method.
814
01:11:53,525 --> 01:12:02,285
However, if you're talking about having, you know, a lot of Bitcoin or hundreds of Bitcoin, you know, you're going to go to the nth degree to protect that.
815
01:12:02,425 --> 01:12:04,345
And if that's what it takes, then that's what it takes.
816
01:12:04,425 --> 01:12:09,645
And you can put them in safe deposit box and you can put them in time delay vaults at banks.
817
01:12:09,745 --> 01:12:14,425
And there's also you can, you know, bury a hole, bury, you know, dig a hole and bury it underground.
818
01:12:14,545 --> 01:12:18,445
You can do whatever you want. You can, you know, etch it into the floorboard floorboards.
819
01:12:18,445 --> 01:12:23,785
There's like all kinds of ways that people do this, but I generally like his method.
820
01:12:24,805 --> 01:12:27,645
And that's distinct from multisig.
821
01:12:27,765 --> 01:12:30,165
So single sig plus passphrase is not multisig.
822
01:12:31,625 --> 01:12:33,265
Multisig is different.
823
01:12:33,585 --> 01:12:39,845
Multisig is like having three separate 12 word seed phrases.
824
01:12:40,245 --> 01:12:41,485
And then you need all three.
825
01:12:41,585 --> 01:12:43,125
They're all different seed phrases.
826
01:12:43,905 --> 01:12:45,845
And then you need all, well, not all three.
827
01:12:45,845 --> 01:13:01,608
You need a quorum like two of three or three of five in order to make transactions And you should also keep your public keys in your descriptors And that actually more complicated I think And so then it gets to this question about accessibility
828
01:13:03,488 --> 01:13:09,748
You know, is single SIG plus passphrase more accessible for average Bitcoiners or average
829
01:13:09,748 --> 01:13:15,148
people to use to secure their funds, which is like a two of two multi SIG without all
830
01:13:15,148 --> 01:13:17,068
of the complexity of multi SIG?
831
01:13:17,068 --> 01:13:19,208
I would say probably yes.
832
01:13:19,988 --> 01:13:32,228
However, if you're talking about a lot of Bitcoin, you probably should, again, do your own research.
833
01:13:32,228 --> 01:13:33,748
This is just me talking, okay?
834
01:13:34,008 --> 01:13:46,628
But yeah, if it's a lot of Bitcoin, you probably should do your due diligence and go through the extra effort of doing multi-sig and putting it into steel and doing all those things.
835
01:13:47,708 --> 01:13:49,568
Putting it into steel meaning like a casing?
836
01:13:49,988 --> 01:13:57,748
well actually etching the uh oh yeah the seed phrase into into steel yeah in case there's a
837
01:13:57,748 --> 01:14:03,568
fire or something it won't paper will burn obviously yes this is the new micro seed little
838
01:14:03,568 --> 01:14:10,828
cool you got one of those nice yeah um so that's the thing if i do like the dice rolling thing with
839
01:14:10,828 --> 01:14:15,648
nova sapien i don't have to do all that stuff and i could save it here so yeah the metal is
840
01:14:15,648 --> 01:14:16,748
definitely putting it to steal.
841
01:14:17,048 --> 01:14:19,308
Well, so I'll continue real quick,
842
01:14:19,548 --> 01:14:22,048
and then we'll move to the Bitcoin core point.
843
01:14:23,148 --> 01:14:25,148
So I'm also, and by the way,
844
01:14:25,188 --> 01:14:26,868
I am affiliated with Novus APN.
845
01:14:27,008 --> 01:14:28,228
They helped me recover my Bitcoin,
846
01:14:28,368 --> 01:14:31,248
and I support them forever for that.
847
01:14:31,248 --> 01:14:34,328
But I'm also affiliated with a company called BitVault,
848
01:14:35,068 --> 01:14:38,468
and they have got a website, bitvault.sv,
849
01:14:38,608 --> 01:14:40,868
and they just released their iPhone app,
850
01:14:41,348 --> 01:14:45,028
and it is a multi-sig-only, non-custodial...
851
01:14:45,648 --> 01:14:52,728
wallet with time delays built into it and arrest codes and various other ways of mitigating
852
01:14:52,728 --> 01:15:00,788
wrench attacks. And I think they've been doing great work. Like I said, they just released their
853
01:15:00,788 --> 01:15:07,388
iPhone version and they're hyper-focused on multi-sig. So with a lot of these wallets,
854
01:15:07,488 --> 01:15:14,108
they do everything. They're like a kitchen sink of different custody solutions. But BitVault is
855
01:15:14,108 --> 01:15:21,588
focused, hyper-focused on just multi-sig, just non-custodial, and with time delays built into it
856
01:15:21,588 --> 01:15:28,088
as a kind of a first-party solution. A time delay is essentially kind of like a bank vault or a vault
857
01:15:28,088 --> 01:15:34,508
at a grocery store where you can only access the Bitcoin once every week or two weeks or whatever
858
01:15:34,508 --> 01:15:42,348
you set your time delay to. And then the idea is that if someone is to, say, try to wrench you
859
01:15:42,348 --> 01:15:47,868
to get your Bitcoin. Well, you don't even have access to your Bitcoin only but so often. So now
860
01:15:47,868 --> 01:15:53,148
they have to upgrade from wrenching you to kidnapping you, and then they have to hold you
861
01:15:53,148 --> 01:15:59,188
for potentially weeks at a time to wait for that time delay to go up. And then presumably during
862
01:15:59,188 --> 01:16:04,508
that time, somebody else in your multi-sig setup will notice that you have in fact disappeared
863
01:16:04,508 --> 01:16:11,508
and then move your Bitcoin out of that particular vault so they can't even get it anymore.
864
01:16:11,508 --> 01:16:18,288
So there's a and then there are other security measures that they're implementing in this wallet as well.
865
01:16:18,808 --> 01:16:24,528
Duress codes and various other things that are specifically designed for that.
866
01:16:25,608 --> 01:16:26,768
That's a really great solution.
867
01:16:26,768 --> 01:16:32,488
I love that because even if that does happen, then you have people in your multi-sync setup that can realize.
868
01:16:33,908 --> 01:16:38,468
And then, I mean, be able to move it for you, you know, because you need the two or three keys.
869
01:16:38,628 --> 01:16:40,008
So that is incredible.
870
01:16:40,008 --> 01:16:45,868
But thank you so much for sharing that because people are looking for all sorts of what can they do?
871
01:16:46,028 --> 01:16:48,508
You know, everyone's looking for I don't trust anybody.
872
01:16:48,688 --> 01:16:50,888
I want to make sure I want to make sure that I'm safe.
873
01:16:51,368 --> 01:16:58,788
And coming from somebody like you, you know, yes, you have affiliations, but it's also you have a story.
874
01:16:58,988 --> 01:17:06,368
You have a reason to protect Bitcoin and protect other people's Bitcoin the way you were protected and the way that people helped you when you want to.
875
01:17:06,508 --> 01:17:07,548
You know, I just love that.
876
01:17:07,548 --> 01:17:11,308
I love that's why I trust your opinion.
877
01:17:11,468 --> 01:17:12,388
I love your deep research.
878
01:17:12,508 --> 01:17:13,188
I love the facts.
879
01:17:13,308 --> 01:17:20,548
I love looking into the history of core and even just not, you know, and all these things, the implementations.
880
01:17:21,428 --> 01:17:24,488
I love what you're doing with commons, with governance.
881
01:17:25,488 --> 01:17:33,308
And I'm always just keeping up with all the things you're doing because you're doing great work and you are kind of like you're detective, but you're building things.
882
01:17:33,388 --> 01:17:35,088
You know, it's just it's so great to see.
883
01:17:35,088 --> 01:17:36,888
And I love having you on.
884
01:17:36,888 --> 01:17:45,888
And so, you know, I feel like you were the first person that I'm like, I have to I have to talk to him about all this stuff because it gives people, you know, knowledge.
885
01:17:45,888 --> 01:17:53,228
You know, you're able to understand what happened, but also peace of mind that you don't feel like you have to be afraid of being in Bitcoin.
886
01:17:54,388 --> 01:18:00,788
Right. And yeah, it is. There's a lot of like I said earlier, there's a lot of responsibility being your own bank.
887
01:18:00,828 --> 01:18:14,931
And thank you for your kind words I do appreciate that I dedicated my life to Bitcoin and I really trying my best to at least expose the issues in Bitcoin
888
01:18:14,931 --> 01:18:18,371
but also not just sit there and complain about it,
889
01:18:18,751 --> 01:18:24,451
but try and prove to people through code, through documentation,
890
01:18:24,451 --> 01:18:28,751
that things could be built differently if we just had the will to do it.
891
01:18:29,391 --> 01:18:31,511
I'm a big security buff.
892
01:18:31,511 --> 01:18:40,571
cybersecurity history. And, you know, I want to see a more secure Bitcoin, not just the
893
01:18:40,571 --> 01:18:46,591
implementation, but the whole ecosystem. And I think there is work to be done. And I think the
894
01:18:46,591 --> 01:18:53,811
Bitcoin that we have today could be a lot better than it is. And I hope to at least have a small
895
01:18:53,811 --> 01:19:01,891
part in moving that needle forward. To that end, I'll say, you know, we go over, you know,
896
01:19:01,971 --> 01:19:08,431
these issues with cold card, there was never any audits of cold card. People just they said,
897
01:19:08,791 --> 01:19:13,351
essentially, they kind of banked on this idea that, well, the source code is available. So
898
01:19:13,351 --> 01:19:19,131
we'll just treat that as audited, or the contributors have reputation, who promote it.
899
01:19:19,211 --> 01:19:23,191
So we'll just treat that as verification, which is a violation of don't trust verified.
900
01:19:23,811 --> 01:19:32,611
The Gold Trash Verify was, of course, written on all of the Tampa Revenant bags and in all of the marketing copy for cold card.
901
01:19:32,611 --> 01:19:41,331
Yet they and the whole community is treating the reputation of the people promoting it as verification.
902
01:19:41,651 --> 01:19:42,591
That's not verification.
903
01:19:44,071 --> 01:19:50,891
And institutional consensus, you know, everybody's saying that all the companies are saying cold card is the best.
904
01:19:50,971 --> 01:19:52,931
All the documentation recommends the cold card.
905
01:19:52,931 --> 01:20:05,151
And this consensus is treated as correctness. So if they're saying it, therefore, it's correct. Therefore, it's verified. If the source is available, therefore, it's audited. None of that is true.
906
01:20:05,151 --> 01:20:11,931
and so you know we need to i i think you know to emphasize and underline this point we in this
907
01:20:11,931 --> 01:20:18,171
community have to get more serious about that and not just with with wallet software but the whole
908
01:20:18,171 --> 01:20:25,471
ecosystem the implementations themselves bitcoin core lib bitcoin in the future bitcoin commons
909
01:20:25,471 --> 01:20:33,831
and there are others working on implementations uh hornet node is uh built being built out and i
910
01:20:33,831 --> 01:20:38,931
I know Reardon Code, he's now working on an implementation for Bitcoin.
911
01:20:39,931 --> 01:20:42,871
And there's Bitcoin Echo.
912
01:20:44,011 --> 01:20:47,931
And there's just a lot of people, you know, trying their hand at this space.
913
01:20:47,951 --> 01:20:48,871
And that's what we need.
914
01:20:48,991 --> 01:20:53,231
We need more eyes on this, more people thinking about how can we make Bitcoin better,
915
01:20:53,331 --> 01:21:00,291
not just going with the status quo, not just treating the fact that it's open source as if it's complete,
916
01:21:00,531 --> 01:21:03,031
as if it's verified, as if it's correct.
917
01:21:03,831 --> 01:21:14,531
And, you know, as far as Bitcoin Core goes, you know, its first independent audit was last year, 2025, 16 years after launch.
918
01:21:14,751 --> 01:21:20,071
OK, so for the vast majority of Bitcoin Core's history, of Bitcoin's history, there had not been an audit.
919
01:21:20,331 --> 01:21:24,851
We were relying simply on people having said they looked at the code and reviewed the code.
920
01:21:24,851 --> 01:21:28,871
And then the audit that we did get was a partial audit.
921
01:21:28,871 --> 01:21:46,971
And it only looked at roughly 20 to 30 percent of consensus critical code. So the majority of Bitcoin's own code has not even been audited. And, you know, we've got three failures in Bitcoin Core's history.
922
01:21:46,971 --> 01:21:54,231
We've got an inflation bug, which basically allowed you to mint an infinite amount of Bitcoin.
923
01:21:54,551 --> 01:21:56,931
You know, Bitcoin is supposed to have this 21 million cap.
924
01:21:57,151 --> 01:22:01,911
We all love touting, you know, 21 million this, 21 million that, everything divided by 21 million.
925
01:22:02,131 --> 01:22:02,911
You see it everywhere.
926
01:22:03,511 --> 01:22:07,811
But for 18 months, you could print as much Bitcoin as you wanted.
927
01:22:08,291 --> 01:22:11,271
You could just do whatever you wanted with the number of Bitcoin.
928
01:22:11,891 --> 01:22:17,831
Fortunately, that bug was not discovered until a white hat found it and responsibly disclosed it.
929
01:22:17,831 --> 01:22:24,631
And we were able to, Bitcoin was able to fix it without it ever being exploited.
930
01:22:25,031 --> 01:22:27,091
So that was huge.
931
01:22:27,191 --> 01:22:36,931
But the fact that it was out there, the fact that it got through production, it was live, it was deployed for 18 months is really unconscionable.
932
01:22:36,931 --> 01:22:45,711
And I'll keep in mind that was 2017, 2018, roughly at the end of the block size war.
933
01:22:45,871 --> 01:22:50,311
So there had been a massive conflict in the community for a long time.
934
01:22:50,391 --> 01:22:57,151
It had been eating up all of the institutional effort, people debating the block size.
935
01:22:57,491 --> 01:23:04,211
And then again, going back to people's egos, everybody was being really rude to each other.
936
01:23:04,211 --> 01:23:07,791
and there was personal attacks going on
937
01:23:07,791 --> 01:23:09,851
and people were getting doxed and DDoSed
938
01:23:09,851 --> 01:23:13,651
and all sorts of ugliness was happening in the community.
939
01:23:14,191 --> 01:23:16,431
Meanwhile, there was this inflation bug
940
01:23:16,431 --> 01:23:19,531
that was literally, could have been a total catastrophe
941
01:23:19,993 --> 01:23:23,413
getting into the code base while we're all distracted with our egos and,
942
01:23:23,593 --> 01:23:23,873
you know,
943
01:23:23,973 --> 01:23:25,053
who's got the bigger,
944
01:23:25,333 --> 01:23:25,633
whatever,
945
01:23:25,873 --> 01:23:26,193
you know,
946
01:23:26,313 --> 01:23:26,913
you know what I'm saying?
947
01:23:27,093 --> 01:23:29,513
So that just shows you that even the governance,
948
01:23:29,793 --> 01:23:30,073
uh,
949
01:23:30,073 --> 01:23:31,473
ever since about a year ago,
950
01:23:32,073 --> 01:23:33,153
um,
951
01:23:34,053 --> 01:23:34,333
you know,
952
01:23:34,333 --> 01:23:35,073
when it came with,
953
01:23:35,073 --> 01:23:36,093
came up to the,
954
01:23:36,093 --> 01:23:37,993
the node debate and all this stuff,
955
01:23:38,033 --> 01:23:40,913
instead of people arguing and having better ways to,
956
01:23:41,013 --> 01:23:43,593
to have this governance move forward,
957
01:23:44,593 --> 01:23:47,813
things like what was going on with cold card were ignored,
958
01:23:48,013 --> 01:23:48,253
you know,
959
01:23:48,253 --> 01:24:07,253
And nobody was really paying attention to that because there was so much chaos going on with the community arguing and fighting over the attention, the time that people are spending on things that aren't as enough to say that they're not important, but there should be better ways to move governance forward.
960
01:24:07,253 --> 01:24:22,393
And I think that's why I was wondering when it comes to the way that you're doing it with commons is by isolating that monolithic code that is consensus in Bitcoin to make sure that that stays consensus.
961
01:24:22,393 --> 01:24:36,833
Well, then you can do other implementations. So I was wondering when it comes to wallet security, if using that same idea to make sure that what makes OK that I don't know if you said driver, but whatever.
962
01:24:37,253 --> 01:24:44,133
device was checked and okayed, but then it wasn't being used. Like there has to be certain
963
01:24:44,133 --> 01:24:48,693
checkpoints or certain things. Honestly, to that end, I would just recommend what
964
01:24:48,693 --> 01:24:53,073
James O'Bearn himself recommended is that while it should use the standard drivers,
965
01:24:53,073 --> 01:24:57,933
which is the LibSep 256K1, which is the same one that Bitcoin Core uses and has the most
966
01:24:57,933 --> 01:25:04,933
vetting and review of any of them, which is not to say it's perfect, but we should stick with the
967
01:25:04,933 --> 01:25:11,293
best solutions and not roll our own solution by some guy who nobody knows who's got a poor commit
968
01:25:11,293 --> 01:25:19,613
history and six github stars like and and i i say that with you know i don't know how to put this i
969
01:25:19,613 --> 01:25:25,353
it's it's really tough and i you know it's you're just speaking truth and people need to hear it
970
01:25:25,353 --> 01:25:42,162
sometimes it like tough love you know you kind of have to you just have to say it like there ever so many people are afraid to say what what important and what would help and um and yeah when it comes to your security and your bitcoin you don want to just
971
01:25:42,882 --> 01:25:48,162
trust anybody you know you have to make sure it's verified so yeah something like that if it works
972
01:25:49,362 --> 01:25:54,242
right and so i'll but just to continue i'll add we talked about this i talked about the inflation bug
973
01:25:54,242 --> 01:25:59,042
um and a lot of people might point out well that was a long time ago that was 2018 you know it's
974
01:25:59,042 --> 01:26:00,042
It's been almost a decade.
975
01:26:00,602 --> 01:26:01,382
You know, that's old news.
976
01:26:01,542 --> 01:26:08,802
Well, as it turns out, in late 2025, Bitcoin Core introduced a wallet deletion bug, another
977
01:26:08,802 --> 01:26:13,502
catastrophic bug, which in some cases would delete people's wallets off of their hard
978
01:26:13,502 --> 01:26:16,342
drives, deleting their Bitcoin with no recovery path.
979
01:26:16,902 --> 01:26:23,222
And a maintainer warned about the exact failure mode a year before it fired and nothing was
980
01:26:23,222 --> 01:26:23,902
done about it.
981
01:26:23,902 --> 01:26:24,542
Okay.
982
01:26:24,542 --> 01:26:50,622
Okay. And then a sort of, and unfortunately, you know, that didn't, that was, that did get into production for two versions, but it didn't stay in production to that, that long. They found it, they pulled the releases. So it, you know, it could have been a lot worse. I have seen reports of people having lost their Bitcoin due to that as well. I don't know if they're confirmed or not, but I have seen reports like that.
983
01:26:50,622 --> 01:26:54,942
and then this year
984
01:26:54,942 --> 01:26:59,022
Bitcoin Core releases a private broadcast feature
985
01:26:59,022 --> 01:27:03,342
which under certain circumstances could be bypassed
986
01:27:03,342 --> 01:27:05,722
it was relatively trivial for an attacker
987
01:27:05,722 --> 01:27:10,082
to unmask the people trying to use private broadcast
988
01:27:10,082 --> 01:27:11,662
and get their real IP address
989
01:27:11,662 --> 01:27:14,322
so it was a private broadcast feature that wasn't private
990
01:27:14,322 --> 01:27:18,042
so you have all these examples of poor review
991
01:27:18,042 --> 01:27:21,922
and I think many Bitcoin core maintainers,
992
01:27:21,962 --> 01:27:23,042
I don't want to rag on them too much,
993
01:27:23,102 --> 01:27:25,862
I think the contributors themselves would say,
994
01:27:26,042 --> 01:27:27,142
we're not getting enough review.
995
01:27:27,302 --> 01:27:30,022
We need more people paying attention to this code base.
996
01:27:30,382 --> 01:27:32,182
Otherwise, this kind of stuff is going to happen.
997
01:27:32,942 --> 01:27:37,782
And to that end, that gets into a whole governance realm
998
01:27:37,782 --> 01:27:41,022
where Bitcoin core,
999
01:27:41,102 --> 01:27:43,042
and I think we talked about this on the last podcast,
1000
01:27:43,042 --> 01:27:44,202
maybe the first podcast.
1001
01:27:44,362 --> 01:27:59,611
Yeah I could link it down below too for them to check it out The attrition rate is so high and you know people come to the repository and maybe they got some bright idea and they just rejected and kind of you know dispensed with
1002
01:27:59,891 --> 01:28:04,291
And then most people, like almost 90% of people leave and never come back.
1003
01:28:04,591 --> 01:28:09,231
And then, you know, they're just not very welcoming to outsiders or new ideas or anything.
1004
01:28:09,771 --> 01:28:14,591
Most people leave and then they complain about a poor reviewer pool in the same breath.
1005
01:28:14,591 --> 01:28:24,131
So and this kind of is the same issue with cold card where, you know, technically it was open source sort of, but it didn't have the open source ethos.
1006
01:28:24,231 --> 01:28:32,291
It was source available. And so, you know, there was no incentive for people to actually do anything to actually look into this code.
1007
01:28:32,691 --> 01:28:41,791
Same thing with Bitcoin Core. You know, technically it's source available, but there's not much incentive for people to look at this code because they can't really do anything with it.
1008
01:28:41,791 --> 01:28:46,811
Most of their changes are rejected and they're basically told one way or the other to go away.
1009
01:28:47,051 --> 01:28:48,391
And that is problematic.
1010
01:28:49,091 --> 01:28:59,731
So, you know, and I'll also add the same people who told you that cold card was the gold standard are the same people whose judgment you're supposed to trust on Bitcoin Core.
1011
01:29:00,091 --> 01:29:04,531
OK, so, again, this is this is an issue.
1012
01:29:05,371 --> 01:29:06,811
The mechanism failed.
1013
01:29:06,811 --> 01:29:17,871
In a way, framing it that way, it's scary because they're trusting somebody and it's like they're not verified.
1014
01:29:17,951 --> 01:29:21,431
It goes back to that thing like, are they verifying anything or are they just like backing?
1015
01:29:22,231 --> 01:29:26,351
Is it a political? Is it an alliance? Is it a friendship?
1016
01:29:27,171 --> 01:29:32,751
And they just have to show optics and support because they have a reputation.
1017
01:29:32,751 --> 01:29:36,371
Let me just point out, I'm completely allocated to Bitcoin.
1018
01:29:36,811 --> 01:29:38,331
I do not promote shit coins.
1019
01:29:38,471 --> 01:29:40,031
I know a lot of people who talk like this,
1020
01:29:40,211 --> 01:29:42,571
they've got some alt coin or something that they're promoting.
1021
01:29:42,691 --> 01:29:44,031
That's not what I'm promoting at all.
1022
01:29:44,031 --> 01:29:48,131
I'm saying we need alternative implementations to Bitcoin
1023
01:29:48,131 --> 01:29:51,071
that are consensus valid with different teams,
1024
01:29:51,431 --> 01:29:52,891
different people reviewing the code,
1025
01:29:53,011 --> 01:29:55,411
looking at consensus and all of that stuff.
1026
01:29:55,531 --> 01:29:59,671
So I don't want to say we need to switch away from Bitcoin
1027
01:29:59,671 --> 01:30:11,740
or Bitcoin is like fundamentally flawed in some way that it insurmountable You just you just need more people like it like it takes a village like you need you know family to help raise kids
1028
01:30:11,740 --> 01:30:14,380
Like it's hard to do it on your own. It's like this is an important thing.
1029
01:30:14,540 --> 01:30:25,120
We have to make sure that there's enough people there helping and taking care and maintaining and offering their eyes and seeing new mistakes if there are any because we all want the same thing.
1030
01:30:25,640 --> 01:30:27,100
A hundred percent. Absolutely.
1031
01:30:27,100 --> 01:30:33,060
Yeah. Is there any other parts of that with the, this was amazing?
1032
01:30:33,060 --> 01:30:58,260
I'm glad to thank you for having me, honestly. And this is honestly a really dark chapter. It's just terrible. I honestly, I don't know what to say. I go through all this stuff. I'm rattling it off. But, you know, my heart hurts for all the people who have been affected. And, you know, I don't know what to tell you. I don't know what you're going through. I don't understand. But, you know, it's just it's really tough.
1033
01:30:58,260 --> 01:30:59,960
and I'm
1034
01:30:59,960 --> 01:31:02,780
as much as I can be I'm with you I'm doing my
1035
01:31:02,780 --> 01:31:04,420
due diligence I'm
1036
01:31:04,420 --> 01:31:06,900
you know trying to get to the bottom of this
1037
01:31:06,900 --> 01:31:08,880
and figure it out and make
1038
01:31:08,880 --> 01:31:10,800
sure that this kind of thing doesn't happen
1039
01:31:10,800 --> 01:31:12,480
again to whatever extent I can
1040
01:31:12,480 --> 01:31:14,880
Thank you thank you for
1041
01:31:14,880 --> 01:31:16,820
all the work you're doing in this space and for the
1042
01:31:16,820 --> 01:31:18,800
research and knowledge and the building
1043
01:31:18,800 --> 01:31:20,680
and everything that you do for
1044
01:31:20,680 --> 01:31:22,340
Bitcoin is
1045
01:31:22,340 --> 01:31:24,940
much appreciated and
1046
01:31:24,940 --> 01:31:26,880
it's nice to see
1047
01:31:26,880 --> 01:31:29,240
that you're doing what you can.
1048
01:31:29,460 --> 01:31:30,920
And so many people are,
1049
01:31:31,060 --> 01:31:33,080
but it's like you need that transparency.
1050
01:31:33,360 --> 01:31:33,940
We need to have,
1051
01:31:34,000 --> 01:31:34,840
and that's why I'm so thankful
1052
01:31:34,840 --> 01:31:35,820
you joined me today too,
1053
01:31:35,900 --> 01:31:37,840
to be able to put it out there.
1054
01:31:37,940 --> 01:31:40,180
I know that you have also your links.
1055
01:31:40,280 --> 01:31:40,840
I'm going to share
1056
01:31:40,840 --> 01:31:43,000
where you have all your information
1057
01:31:43,000 --> 01:31:44,880
and everything that you have shared,
1058
01:31:45,380 --> 01:31:47,180
but it's also nice to get it out
1059
01:31:47,180 --> 01:31:47,840
into more places.
1060
01:31:47,840 --> 01:31:49,920
So I'm so grateful to have you on the show
1061
01:31:49,920 --> 01:31:51,300
and discuss the situation.
1062
01:31:52,260 --> 01:31:54,120
I'm glad that you offered solutions too.
1063
01:31:54,240 --> 01:31:55,160
I'm super excited.
1064
01:31:55,320 --> 01:31:56,380
I'm going to share all those links.
1065
01:31:56,880 --> 01:32:08,380
So the links are below in the video and thank you. Thank you for this. Stay safe, everybody. And, you know, actually verify. Don't just trust.
1066
01:32:08,380 --> 01:32:12,620
Yeah, actually verify. Okay. Thank you, Stacey.
1067
01:32:12,760 --> 01:32:14,500
All right. Thank you. Thank you so much.