1
00:00:00,000 --> 00:00:17,320
Welcome to the Free Cities podcast. My name is Timothy Allen and this is the official podcast

2
00:00:17,320 --> 00:00:19,020
of the Free Cities Foundation.

3
00:00:30,000 --> 00:00:36,340
Hello and welcome to this episode number 147 of the Free Cities podcast.

4
00:00:37,520 --> 00:00:39,760
Yes, hello, I'm back from Serbia.

5
00:00:40,120 --> 00:00:47,020
It was an incredibly interesting trip, actually, somewhat surreal at times as well.

6
00:00:47,700 --> 00:00:51,380
Have you ever spent three hours in a taxi with two flat earthers?

7
00:00:52,300 --> 00:00:55,020
Like you literally can't get away?

8
00:00:55,020 --> 00:00:56,460
Well, I have now.

9
00:00:56,460 --> 00:01:05,980
And to be honest, it was actually quite fascinating. You know me, I'm an armchair psychologist, so ideologies like that really interest me.

10
00:01:06,340 --> 00:01:13,760
And until that point, I had actually never come across someone who believed in that stuff in person.

11
00:01:13,980 --> 00:01:21,140
Anyway, I won't divulge here. Go and have a look online if you want to test your own theory on the shape of the earth.

12
00:01:21,140 --> 00:01:31,140
yes episode 147 we're at now great harsh ad number loved by british snooker fans as well

13
00:01:31,140 --> 00:01:36,320
if you don't know the connection there look it up welcome indeed to this episode which is

14
00:01:36,320 --> 00:01:43,880
as always greatly supported by our wonderful partner veritas villages veritas villages are

15
00:01:43,880 --> 00:01:50,920
off-grid energy self-sufficient luxury communities in latin america designed specifically for freedom

16
00:01:50,920 --> 00:01:57,420
lovers bitcoin is accepted at all villages including for the purchase of property you know

17
00:01:57,420 --> 00:02:05,060
where to go to find out more but i'll give it to you anyway www.ecovillages.life forward slash free

18
00:02:05,060 --> 00:02:10,140
cities everything's in the show notes as always and if you want to meet the guys from veritas

19
00:02:10,140 --> 00:02:16,260
villages in person patrick hibbert and others he's the creator of veritas villages will be speaking

20
00:02:16,260 --> 00:02:23,420
at this year's Free Cities Conference in Prague 31st of October to the 2nd of November 2025. It's

21
00:02:23,420 --> 00:02:28,940
coming up soon. The Free Cities Conference really is one of the highlights of my year.

22
00:02:30,380 --> 00:02:35,940
Prague's also a very beautiful place to spend a few days in early winter too, especially if you

23
00:02:35,940 --> 00:02:41,600
can take your loved one with you. It's very romantic. If you're in this space or indeed if

24
00:02:41,600 --> 00:02:47,660
you want to be in this space then this is the gathering for you just ask anyone who's ever

25
00:02:47,660 --> 00:02:55,220
attended before and i will see you there right today's guest he's a developer in the privacy

26
00:02:55,220 --> 00:03:01,940
space he's also a climber more of that in the podcast uh there's going to be some quite important

27
00:03:01,940 --> 00:03:08,480
alpha here for you if like me you're into privacy but you don't necessarily know what makes one

28
00:03:08,480 --> 00:03:15,920
protocol better than the other. Jeff G develops on Noster, which you should know about already.

29
00:03:16,020 --> 00:03:21,540
If you don't, check the show notes for some links. And he's currently responsible for creating

30
00:03:21,540 --> 00:03:28,600
what is its most secure messaging layer to date, something that will become incredibly important

31
00:03:28,600 --> 00:03:35,960
in the not too distant future if we continue along the same centralizing trajectory that we

32
00:03:35,960 --> 00:03:42,500
currently see in so many countries around the world. Here's an AI breakdown for the curious

33
00:03:42,500 --> 00:03:49,680
amongst you. The discussion ranges from AI tools and privacy concerns to the evolution of secure

34
00:03:49,680 --> 00:03:57,420
messaging where Jeff introduces his project White Noise and its use of message layer security,

35
00:03:57,640 --> 00:04:04,380
MLS. One to watch that is on Nostra. Along the way they explore the limits of existing platforms

36
00:04:04,380 --> 00:04:10,440
like WhatsApp and Signal, the challenges of metadata privacy and practical defences

37
00:04:10,440 --> 00:04:19,200
against sophisticated threats. The conversation broadens into Bitcoin, eCash and mesh networks

38
00:04:19,200 --> 00:04:25,560
and wraps with reflections on teaching digital responsibility, the trade-offs between privacy

39
00:04:25,560 --> 00:04:31,080
and convenience and the promise of open source decentralized technologies to shape a freer

40
00:04:31,080 --> 00:04:37,700
internet if this sounds too technical for you please don't stop here give it a go i ask all

41
00:04:37,700 --> 00:04:44,280
the normie questions this stuff is really important to wrap your head around it's easy to predict

42
00:04:44,280 --> 00:04:50,800
what's coming in the future and as a sovereign individual you must take responsibility for your

43
00:04:50,800 --> 00:04:57,960
and your family's privacy it's up to you to be the one who gets everyone on board with this stuff

44
00:04:57,960 --> 00:05:04,140
there's a lot of pressure around the world right now to give up privacy and as far as i'm concerned

45
00:05:04,140 --> 00:05:10,860
the people who advocate for that are not on the right side of history right some thank yous this

46
00:05:10,860 --> 00:05:18,740
week before we go to my new subscribers and you know what quite astonishingly the free cities

47
00:05:18,740 --> 00:05:27,680
podcast got to number three in the top 40 shows on fountain this week and our latest bonus episode

48
00:05:27,680 --> 00:05:33,300
with Gigi and episode 144 with Leon Wancombe, which has been doing very well, have both been

49
00:05:33,300 --> 00:05:42,840
hovering in the charts around seventh place on the top 40 of episodes two. So many thanks to you

50
00:05:42,840 --> 00:05:49,780
for choosing to listen. It's great. We're currently we've dropped down to ninth in the top 40 of shows,

51
00:05:49,780 --> 00:05:56,000
but really for a good week, we were hovering around the top. So many thanks to you in particular,

52
00:05:56,000 --> 00:06:04,540
shadrach silas thornbrook set in stone by the way you're this guy's a real baller he's he's streaming

53
00:06:04,540 --> 00:06:11,740
200 sats per minute when he's listening to your show so set the stone thank you very much indeed

54
00:06:11,740 --> 00:06:23,760
tim a n t chris albanese uh there's a private end pub there one k j q e bullish mike matthias

55
00:06:23,760 --> 00:06:30,660
cousin Vinny you know who you are many thanks I appreciate you I appreciate your continuing

56
00:06:30,660 --> 00:06:38,760
subscriptions they really do make a difference to me right let's get on with this time to learn

57
00:06:38,760 --> 00:06:48,300
how to become your very best private sovereign individual as you sit back relax and enjoy my

58
00:06:48,300 --> 00:06:51,140
conversation with Jeff G

59
00:06:53,760 --> 00:07:13,560
now um yeah i don't have a start and a finish so we're already rolling cool um i've got to

60
00:07:13,560 --> 00:07:18,920
recount a little story for you because you're you're a techie person this i had a catastrophic

61
00:07:18,920 --> 00:07:25,400
hard drive failure yesterday oh yeah oh doesn't happen very often these days and yeah it was a

62
00:07:25,400 --> 00:07:32,200
spinning disc one okay so probably it does let's turn that off sorry let's get rid of all my in it

63
00:07:32,200 --> 00:07:39,140
let's turn that off and um i've been i've been in the media years so i back everything up yeah

64
00:07:39,140 --> 00:07:43,840
and luckily i had backed this i backed this up because it was all the recordings i've done over

65
00:07:43,840 --> 00:07:49,740
the last seven days yeah but i wasn't as on it as normal so there was an absolute there was a gap

66
00:07:49,740 --> 00:07:55,540
there was yeah there was so i didn't lose any data but what i did discover is i've lost the complete

67
00:07:55,540 --> 00:08:02,700
back catalog of my whole podcast every single file including all the templates for the video

68
00:08:02,700 --> 00:08:09,600
files and all this stuff right and it's not bothering me yet too much but because i can

69
00:08:09,600 --> 00:08:13,780
recreate all those what i haven't lost is anything that's not recreatable which is good

70
00:08:13,780 --> 00:08:20,580
but what the reason i'm telling you is because this time i went to chat gpt and i've i feel like

71
00:08:20,580 --> 00:08:27,240
i've made a friend with chat gpt for the first time this guy or girl or whoever it is and i i

72
00:08:27,240 --> 00:08:33,320
started using venice i use venice.ai yeah yeah but it was it was and it wasn't getting what i

73
00:08:33,320 --> 00:08:38,020
wanted i went to chat gpt and all of a sudden almost the first thing it says was we're cool

74
00:08:38,020 --> 00:08:45,020
you're in my hands now it's really weird that is weird and and it's been it's guided me through

75
00:08:45,020 --> 00:08:50,740
this process so i i've got five terminals open i'm i'm putting all this code in and every time

76
00:08:50,740 --> 00:08:55,080
i put something in i'm telling jettgpt what comes out and it's saying great we're on it it's

77
00:08:55,080 --> 00:09:00,780
happening it's really you're going to get this stuff back don't worry and now i've got a probably

78
00:09:00,780 --> 00:09:06,840
a 15 to 16 hour process running so i ran disc disc util yeah yeah like a dis recovery thing yeah

79
00:09:06,840 --> 00:09:10,060
Yeah. Well, it's the one, it's a free open source one.

80
00:09:10,180 --> 00:09:10,340
Yeah.

81
00:09:11,280 --> 00:09:30,460
And it good And it even saying update me every hour and I tell you how long it going to be whether I think the drive good or not after one hour And it like wow you know that is really new and amazing and different but what crazy is like I said thank you a while ago I

82
00:09:30,460 --> 00:09:36,320
instantly I and I was like hold a minute dude you're falling down yes the wrong rabbit hole

83
00:09:36,320 --> 00:09:43,480
yeah you're not this is not a person this is not a thing but compared to Venice which was like

84
00:09:43,480 --> 00:09:50,580
a computer talking to me yes chat gbt was really starting to sort of talk at it with me and stuff

85
00:09:50,580 --> 00:09:54,480
yeah the crazy thing about the models is that especially the really frontier models you know

86
00:09:54,480 --> 00:10:02,040
the you know chat gbt 5 or claude 4 um i actually prefer claude over chat gbt i mean especially in

87
00:10:02,040 --> 00:10:06,300
the coding domain i think claude's way better how do i how do i access claude though if i wanted to

88
00:10:06,300 --> 00:10:10,680
you just go to i think it's called.com or anthropic or claude are there places like you

89
00:10:10,680 --> 00:10:14,940
know i use venice just because i'm i don't like putting all my if it's code i don't mind giving

90
00:10:14,940 --> 00:10:20,800
it to chat but if it's like talking about my family or something i don't want i mean all of

91
00:10:20,800 --> 00:10:25,680
those frontier models are running on you know giant data centers in the cloud and so anything

92
00:10:25,680 --> 00:10:31,080
you give the hosted providers they have and yeah they say like private mode like don't store my

93
00:10:31,080 --> 00:10:35,740
stuff but you know they can use venice says though that it's venice it's encrypting it from

94
00:10:35,740 --> 00:10:40,340
your so you're doing it all locally and then when it sends it is encrypting yes exactly so there's

95
00:10:40,340 --> 00:10:46,200
some providers like venice like maple ai which is the you know former mutiny guys um you know those

96
00:10:46,200 --> 00:10:50,700
do things a bit differently but the major providers which are the ones that have the best models

97
00:10:50,700 --> 00:10:56,560
are the ones that are a little bit less you know privacy oriented well i'm one i'm one of those

98
00:10:56,560 --> 00:11:04,800
people that thinks that the dystopian ai future is as likely as the utopian one so i don't want it

99
00:11:04,800 --> 00:11:11,800
And I don't want it to be involved deeply in my personal life because it will be able to annihilate me.

100
00:11:12,180 --> 00:11:23,100
And really, I mean, the local models are actually getting, you know, the open source models that you can install locally and run on a laptop and run on a laptop in a way that's not completely terrible.

101
00:11:23,100 --> 00:11:25,500
Like you're not going to have to wait 20 seconds for every response.

102
00:11:26,180 --> 00:11:27,660
They're actually getting pretty good.

103
00:11:27,660 --> 00:11:38,420
And I have a feeling that we're going to get to a point pretty soon where, you know, the amount that we can both spend and get out of, you know, each of the next generation frontier models, it's going to be like iPhones.

104
00:11:38,560 --> 00:11:40,380
They're just going to kind of be the same pretty much.

105
00:11:40,820 --> 00:11:45,460
And so at that point, I think a lot of the open source is going to catch up, you know, in short order.

106
00:11:45,460 --> 00:11:52,640
And we'll all be able to run, you know, very, very compelling, very powerful models locally on our computers and have perfect privacy.

107
00:11:53,260 --> 00:11:54,760
Perfect. Right. I'm glad to hear that.

108
00:11:54,820 --> 00:11:55,980
At least I hope. That's my hope.

109
00:11:55,980 --> 00:12:04,960
The funny thing is, though, because of my history of using AI, while I was doing all that stuff, I was very nervous because it's hallucinated many times with me.

110
00:12:05,380 --> 00:12:13,060
And it was saying, right, now we're going to wipe that drive because I went out yesterday and I bought a whole new five terapyte drive.

111
00:12:13,400 --> 00:12:17,120
So I'm copying what's there onto there before I do anything.

112
00:12:17,240 --> 00:12:19,260
It's going to take like 24 hours or something ridiculous.

113
00:12:19,760 --> 00:12:21,660
But it's saying now I'm doing this and this and this.

114
00:12:21,740 --> 00:12:24,180
And I'm thinking, dude, is it hallucinating?

115
00:12:24,180 --> 00:12:27,300
It might just be about to wipe my good drive.

116
00:12:27,380 --> 00:12:27,600
Yes.

117
00:12:27,600 --> 00:12:30,420
You know, like, so I'm still not there with them.

118
00:12:30,420 --> 00:12:35,080
The way I used to check AIs in the past was already ask them about myself.

119
00:12:35,760 --> 00:12:39,080
Because I've lived two very distinct lives.

120
00:12:39,200 --> 00:12:40,240
I had a media career.

121
00:12:40,860 --> 00:12:45,140
And then I have a kind of like Bitcoin and libertarian bit.

122
00:12:45,600 --> 00:12:49,540
And most places have never actually connected the two.

123
00:12:49,900 --> 00:12:50,920
JTBG now does.

124
00:12:51,680 --> 00:12:53,700
So it's worked out that that's the same me.

125
00:12:53,700 --> 00:12:58,520
because they're very different worlds but it still hallucinates like there's a couple of books that i

126
00:12:58,520 --> 00:13:03,720
haven't written they're just made up completely made up out of nothing they don't even exist as

127
00:13:03,720 --> 00:13:08,480
real books right this is what it says about you because you tell me whether it's hallucinating

128
00:13:08,480 --> 00:13:15,200
okay i asked it who's jeff g it's like uh full-time nostor developer supported by open sats

129
00:13:15,200 --> 00:13:20,400
focusing on building open decentralized tools for the nostor ecosystem that's all correct okay

130
00:13:20,400 --> 00:13:25,260
creator founder of white noise correct still correct i know that one a project aimed at

131
00:13:25,260 --> 00:13:31,780
bringing messaging layer security mls capabilities to nostra including support for end-to-end

132
00:13:31,780 --> 00:13:38,100
encryption encrypted direct messages and private groups okay your contributions white noise

133
00:13:38,100 --> 00:13:45,880
nostra.how you've done nostra talks podcasts featured in how to monetize your nostra apps

134
00:13:45,880 --> 00:13:51,700
hosts oh you host a pod as well thank god for nostor uh i don't host that that's actually a

135
00:13:51,700 --> 00:13:55,400
friend of mine's i've been on that podcast all right i've been oh it says okay it says hosts

136
00:13:55,400 --> 00:14:02,460
episodes there see it's hallucinating yeah okay summary jeff jeff g is a core figure in the nostor

137
00:14:02,460 --> 00:14:09,020
ecosystem leading privacy focused development improving the onboarding experience via nostor.how

138
00:14:09,020 --> 00:14:14,440
and actively educating and shaping the community through podcasts and talks it's fairly accurate i

139
00:14:14,440 --> 00:14:18,260
guess is that good yeah yeah fairly accurate i'm glad it didn't come out with like something in the

140
00:14:18,260 --> 00:14:23,380
deep past where you're like whoa how did it know that but well yeah have you got is have you got a

141
00:14:23,380 --> 00:14:27,440
different past like a lot of people who work in this kind of space i mean everybody comes from

142
00:14:27,440 --> 00:14:32,580
doing something before unless they're 20 years old at this point so i think uh yeah my past is all

143
00:14:32,580 --> 00:14:37,420
startups uh i've been working in tech startups since kind of 2011 12 something like that anything

144
00:14:37,420 --> 00:14:41,720
i might have heard of anything interesting or what kind of stuff um i worked at a company the first

145
00:14:41,720 --> 00:14:46,800
startup i ever joined actually is a company called intercom which is a fairly big one now uh they are

146
00:14:46,800 --> 00:14:51,260
um you know the little chat bubble thing that comes up you know pops up at the bottom of your

147
00:14:51,260 --> 00:14:55,340
screen little circle and you tap it and it opens up a window they actually invented that oh really

148
00:14:55,340 --> 00:14:59,060
and then everybody took it but they're still around and they're still doing you know great

149
00:14:59,060 --> 00:15:03,800
business ask me a question then because i've often wondered this when you're having a chat

150
00:15:03,800 --> 00:15:10,000
conversation with a company even if it's a real person is it you know when it waits and it shows

151
00:15:10,000 --> 00:15:15,560
you someone is typing yeah is it doing that to make you feel more relaxed it depends on the

152
00:15:15,560 --> 00:15:21,060
company uh i know with intercoms it's real uh it is a real typing indicator yeah yeah i mean i think

153
00:15:21,060 --> 00:15:25,480
that's the thing it's always with those things and intercom has been really at the forefront of this

154
00:15:25,480 --> 00:15:31,700
and they're probably the best customer support ai in the world at this point um because they just

155
00:15:31,700 --> 00:15:36,580
understand more about human behavior and what humans actually want to do on both sides of the

156
00:15:36,580 --> 00:15:40,760
table you know like it's not nice for a customer support person to sit there and just blindly

157
00:15:40,760 --> 00:15:45,520
answer the same stupid question all day every day um it's a horrible use of humanity so like

158
00:15:45,520 --> 00:15:50,340
perfect use of ai but there's every now and then you want a person to step in and be like okay i

159
00:15:50,340 --> 00:15:54,120
got you like this is going to be a really weird one but like we can navigate this together and

160
00:15:54,120 --> 00:16:00,280
we can sort it out so a lot of what they say to you is generated automatically like tons at this

161
00:16:00,280 --> 00:16:05,440
point like most companies if they are not solving 50 to 70 percent of customer requests using

162
00:16:05,440 --> 00:16:10,260
something automated their test even if you think you're talking to a real person i mean it doesn't

163
00:16:10,260 --> 00:16:15,220
matter i'm thinking of amazon there's the most common one i go on and it's always it's often

164
00:16:15,220 --> 00:16:20,300
someone in india i think because there's a lot of indian names but there's a there's a couple of

165
00:16:20,300 --> 00:16:27,240
answers you get a lot i'm sorry to hear that let me see if i can that's obviously uh someone sending

166
00:16:27,240 --> 00:16:32,040
a message to calm you down right because basically it's going to take longer and if they just stay

167
00:16:32,040 --> 00:16:37,860
completely silent you get annoyed yeah i mean but think about this uh they're programmed to say that

168
00:16:37,860 --> 00:16:42,520
whether they're a human or they're an or they're an ai right they went through a training that gave

169
00:16:42,520 --> 00:16:46,160
them a script that told them to do certain things in certain cases so either way they've been

170
00:16:46,160 --> 00:16:51,460
programmed to do the thing um would you rather have that response in a sub-second or have somebody

171
00:16:51,460 --> 00:16:54,800
be able to read all of the documentation instantaneously and give you the right answer

172
00:16:54,800 --> 00:16:58,660
without having to go look i would much rather that so like i'd rather have an ai for most stuff that

173
00:16:58,660 --> 00:17:03,200
I want to deal with with the company. It's probably a soundboard as well, isn't it? Sorry to hear you.

174
00:17:03,280 --> 00:17:07,660
Yeah, exactly. I don't know. Radio shows like bang, bang.

175
00:17:09,160 --> 00:17:13,760
Yeah, that's interesting. And what other things have you done then just so that I can get an idea?

176
00:17:14,060 --> 00:17:19,600
I mean, it's kind of been across the board. I went to school actually for finance in Spanish. I didn't

177
00:17:19,600 --> 00:17:24,200
do anything computer science related. To be honest, I kind of knew I wanted to do computer science,

178
00:17:24,200 --> 00:17:28,420
didn't get into the program originally, and then got distracted with climbing for a bunch of years.

179
00:17:28,660 --> 00:17:30,880
And then kind of came back around to it on my own later.

180
00:17:31,040 --> 00:17:33,400
Is that anything to do with, am I allowed to say where you live?

181
00:17:34,320 --> 00:17:34,640
Yeah.

182
00:17:35,300 --> 00:17:37,100
Well, is that anything to do with where you live in Italy?

183
00:17:37,240 --> 00:17:38,400
Because there's some good mountains there.

184
00:17:38,460 --> 00:17:39,380
Well, the climbing does, yeah.

185
00:17:39,620 --> 00:17:40,100
Yeah, yeah, yeah.

186
00:17:40,100 --> 00:17:42,500
We live basically where we live in Italy for the mountains.

187
00:17:42,940 --> 00:17:43,120
Really?

188
00:17:43,300 --> 00:17:43,820
Yeah, yeah.

189
00:17:44,100 --> 00:17:47,820
And you say we, it sounds like you're, whoever else is there.

190
00:17:47,820 --> 00:17:48,280
Yeah, me and the family.

191
00:17:48,740 --> 00:17:49,700
You're all into climbing.

192
00:17:50,440 --> 00:17:50,760
Yes.

193
00:17:50,900 --> 00:17:52,680
My wife and I actually met climbing years ago.

194
00:17:52,840 --> 00:17:54,960
Have you done any pretty epic things?

195
00:17:55,540 --> 00:17:57,060
What's the most epic thing you've ever done?

196
00:17:57,060 --> 00:17:58,260
My wife and I met in Yosemite.

197
00:17:58,660 --> 00:18:02,020
And I've spent a couple of years living in Yosemite Valley climbing.

198
00:18:02,400 --> 00:18:03,280
So El Cap, half down.

199
00:18:03,280 --> 00:18:04,020
Have you done El Cap?

200
00:18:04,080 --> 00:18:05,900
Yeah, a bunch of times by a bunch of different routes.

201
00:18:06,240 --> 00:18:06,720
You're real.

202
00:18:06,980 --> 00:18:07,880
Yeah, it's really fun.

203
00:18:08,060 --> 00:18:09,100
It was like such a good time.

204
00:18:09,220 --> 00:18:10,080
Oh, tell me about that.

205
00:18:10,220 --> 00:18:14,460
I love, I'm fascinated by climbing.

206
00:18:14,720 --> 00:18:17,680
I've never climbed in the most extreme way.

207
00:18:18,060 --> 00:18:19,480
I've hiked a lot of mountains.

208
00:18:19,480 --> 00:18:23,800
I've been up to 6,500 meters and lived in the Himalayas for a long time.

209
00:18:24,160 --> 00:18:28,240
And I've climbed Mongolia's tallest mountain, which is one of the really remote places.

210
00:18:28,660 --> 00:18:35,320
with ice picks and stuff like that but i've never done oh shit yeah technical climbs sleeping on

211
00:18:35,320 --> 00:18:54,920
right you know what my favorite film of all time is meru have you seen it was so good yeah that actually i feel like jimmy chin like sort of solidified his you know career on that one because that film was so good But what amazing about that film is apart from the fact the story amazing the timeline amazing because it a long timeline as well

212
00:18:55,080 --> 00:18:56,820
And the whole, there are so many little quirky things to it.

213
00:18:56,820 --> 00:18:59,280
And the history of that mountain, like, the whole thing is really cool.

214
00:18:59,280 --> 00:19:11,080
But what's amazing is I think that film taps into the essence of how insane it is when you climb things like that.

215
00:19:11,300 --> 00:19:13,020
And living vertically is a thing.

216
00:19:13,280 --> 00:19:14,060
But no, not even that.

217
00:19:14,180 --> 00:19:21,340
It's like I think my theory, see, on climbers who go to that extreme is they are trying to access a certain feeling.

218
00:19:21,680 --> 00:19:21,780
Right.

219
00:19:21,780 --> 00:19:26,680
and the only way they can do it is literally live on the side of a mountain for two weeks

220
00:19:26,680 --> 00:19:33,300
freeze their feet and fingers off and when they get to that top and you see this in mirror there's

221
00:19:33,300 --> 00:19:38,620
just kind of like there's like no nice one dude you know and and that's it yep that's it there's

222
00:19:38,620 --> 00:19:44,540
nothing to it and that whole thing and and i get too fun is what i call it you know type two so

223
00:19:44,540 --> 00:19:48,520
there's type one fun where it's fun while you're doing it and fun when you think back on it later

224
00:19:48,520 --> 00:19:51,760
you're like that's just enjoyment right like normal things that people think of as fun

225
00:19:51,760 --> 00:19:56,120
type two fun is mostly suffering while you're doing it and when you look back on it somehow

226
00:19:56,120 --> 00:20:00,120
your memory has magically transformed it into something very lovely and amazing and something

227
00:20:00,120 --> 00:20:04,400
you really love do you get scared on like el cap for example oh yeah for every i mean if anyone

228
00:20:04,400 --> 00:20:09,620
doesn't know el cap is is it the the tallest flat face in the world it's not the tallest flat face

229
00:20:09,620 --> 00:20:14,000
in the world there's a few others that are bigger but i mean it's one of them and i think it's i

230
00:20:14,000 --> 00:20:18,740
mean it's a thousand meters um in its tallest section a kilometer it's a kilometer tall and

231
00:20:18,740 --> 00:20:22,460
And what is amazing about it is, you know, it's 10 minutes from the side of the road.

232
00:20:22,660 --> 00:20:30,000
Like you literally drive into the valley, you park your car in El Cap Meadow, you walk 10 minutes in flip flops and you put your hands on a kilometer tall wall.

233
00:20:30,000 --> 00:20:32,500
And the wall goes, it's like a wave.

234
00:20:32,760 --> 00:20:34,760
It's lower than vertical for the first half.

235
00:20:34,900 --> 00:20:37,400
And then it goes vertical and then it's over vertical.

236
00:20:37,540 --> 00:20:38,980
You know, it's overhanging for the top half.

237
00:20:39,280 --> 00:20:43,280
So when you look up, it's like literally standing under a kilometer tall wave of granite.

238
00:20:43,360 --> 00:20:45,340
And it is the most beautiful face in the world.

239
00:20:45,340 --> 00:20:51,600
god and it's two or three days normally to get up it it depends on the route um so if it's easier

240
00:20:51,600 --> 00:20:55,040
routes you can certainly do them faster than that you know lots of people have done you know the

241
00:20:55,040 --> 00:20:58,420
nose which is kind of the standard route right up the front of you know right in the middle of the

242
00:20:58,420 --> 00:21:02,860
wall um they've done it in a couple of hours uh you know super super strong folks like you know

243
00:21:02,860 --> 00:21:09,760
alex hondold and others um most normal humans would take you know two three days to do the nose

244
00:21:09,760 --> 00:21:15,780
but as soon as you start to go into harder territory where the climbing is you know more

245
00:21:15,780 --> 00:21:20,760
strung out much more difficult and again most people are actually doing lcap with artificial

246
00:21:20,760 --> 00:21:25,100
climbing aid climbing which is instead of using just your hands and feet you're putting gear into

247
00:21:25,100 --> 00:21:28,580
the wall and you're actually hanging on the gear and standing up in these little ladders to reach

248
00:21:28,580 --> 00:21:33,040
up and put some more gear in so it's that's a slow process you know it's vertical engineering

249
00:21:33,040 --> 00:21:38,040
you're carrying a ton of gear you have to carry all your water with you you know and it just takes

250
00:21:38,040 --> 00:21:42,380
time so how do you go to the loo just out of interest i've never well i mean do you literally

251
00:21:42,380 --> 00:21:47,140
just let it go into the wind but uh number two you've got to bring you you can't just oh you

252
00:21:47,140 --> 00:21:50,740
can't you know no it's totally illegal in lcap you'd kill someone right like imagine being killed

253
00:21:50,740 --> 00:21:56,480
by falling shit depends how much you ate i suppose right if you had a big meal yeah you bring a very

254
00:21:56,480 --> 00:22:01,520
specific piece of tech called a uh poop tube and it's a piece of pvc pipe with caps on both ends

255
00:22:01,520 --> 00:22:05,640
and a you know runner around the end of it and you've got to bag everything and put it in and

256
00:22:05,640 --> 00:22:11,580
take it with you yeah so if you can you probably try not to because that is a right i mean it's

257
00:22:11,580 --> 00:22:15,980
actually not a big deal like once you get used to all the systems it's it's surprisingly

258
00:22:15,980 --> 00:22:20,420
straightforward to do pretty much everything um and you'll never have a better night's sleep than

259
00:22:20,420 --> 00:22:25,400
you do in a portal edge and i mean mostly just because you're absolutely exhausted yeah you lie

260
00:22:25,400 --> 00:22:30,480
down and you're like oh god i'm done have you had any scares in your climbing career i mean nothing

261
00:22:30,480 --> 00:22:39,860
nothing that I would say like was really actually very very close to you know ending my life I've

262
00:22:39,860 --> 00:22:42,620
definitely had a lot of scares I've definitely a lot of falls that scared the crap out of me

263
00:22:42,620 --> 00:22:46,600
I've definitely been in situations where I was like two or three things go bad here and we're

264
00:22:46,600 --> 00:22:52,240
in a really bad situation but I think we're going to be okay but that was mostly around exposure to

265
00:22:52,240 --> 00:22:56,520
be honest and to be honest most of the people that die on El Cap die of exposure because they go up

266
00:22:56,520 --> 00:22:59,960
in a t-shirt, you know, there's a storm that comes in, it dumps a bunch of snow on them,

267
00:23:00,100 --> 00:23:04,860
and then they can't get out. So it's exposure that kills most people, not the actual fall,

268
00:23:05,040 --> 00:23:09,200
or, you know, not the actual climbing. I mean, rockfall is the other thing that I worry about.

269
00:23:09,320 --> 00:23:13,700
And I think that's the thing I worry about more now, also given where we live, just because we're

270
00:23:13,700 --> 00:23:19,700
in the Alps and, you know, big mountains fall down. In a place like Yosemite, you've got glacial

271
00:23:19,700 --> 00:23:24,080
granite that has been there, you know, forever. And yes, there's rockfall, but the routes are so

272
00:23:24,080 --> 00:23:28,940
well traveled that most of the loose stuff has been knocked off um and so it's a little bit safer

273
00:23:28,940 --> 00:23:32,740
in that have you been up the matterhorn i've not been up the matterhorn i've been i've climbed a

274
00:23:32,740 --> 00:23:36,100
bunch in chamonix um which is just around the corner from the matterhorn i've skied on one

275
00:23:36,100 --> 00:23:40,120
side of the matterhorn but i've never climbed it but that doesn't that's not a problem you look at

276
00:23:40,120 --> 00:23:43,760
the matterhorn you think good fun you don't think yeah yeah it would be a pretty straightforward i

277
00:23:43,760 --> 00:23:47,540
mean it's a really well trodden route it's kind of an alpine slog there's a little bit of kind of

278
00:23:47,540 --> 00:23:52,480
scrambly stuff with an ice axe but i don't think it would be anything i couldn't do it anymore i'll

279
00:23:52,480 --> 00:23:59,440
be honest as for some reason as i've grown older i've i've got vertigo and um even when i was

280
00:23:59,440 --> 00:24:08,700
climbing the one in mongolia that was what four or four odd years ago um it was ice axes and but

281
00:24:08,700 --> 00:24:15,020
it wasn't that it was probably that it wasn't that and i i didn't look down in fact we had a we had a

282
00:24:15,020 --> 00:24:18,900
lot of um weather come in so i couldn't look down there's nothing to see nothing to look at

283
00:24:18,900 --> 00:24:27,040
fortunately yeah but um but i know now i watch videos still all the time and i could not stand

284
00:24:27,040 --> 00:24:32,500
like for example on the side of el cap and look down my i get actually physical reaction in my

285
00:24:32,500 --> 00:24:36,880
knees and feet yeah they start you know yeah i i didn't and i don't want to do about it because

286
00:24:36,880 --> 00:24:41,820
it's in a one way it's annoying same with same with the alps there's a lot of really interesting

287
00:24:41,820 --> 00:24:47,600
treks in the alps that involve walking across like little bits of metal right yeah those things

288
00:24:47,600 --> 00:24:56,820
And that would also scare me too much. I think I might run the risk of actually my knees going, you know, and like feeling like they're going to fall away from you.

289
00:24:56,820 --> 00:25:22,120
It's one of those things you habituate to, though, like like everything you get used to it. And it's, you know, climbers talk about having a good head or not. And having a good head just basically means like, am I in the mind state that I can that I can take exposure, that I can concentrate on what's in front of me and like do the thing I need to do without thinking about too much the consequences like you can, you know, assess the consequences, know what you're getting into in a realistic way, but also just zero in and focus on stuff.

290
00:25:22,120 --> 00:25:44,900
And that takes time. Like it takes practice to develop that. And it comes and goes. Like it's not something like if I went out now and I haven't been climbing all that much recently. If I went out now and tried to climb something, it would probably scare me to death. Like I'd be up there being like, oh, my God, I can't take this. But if you do it on a really, really, really basis, you really kind of get into that groove and you you have a better ability to sort of focus, assess, you know, actual risk versus perceived risk.

291
00:25:44,900 --> 00:26:04,020
You know, your brain is going to tell you like, this is no place that you should be like, you need to get out of here immediately. And the trick is just going, actually, I am safe. I have done the right things. I do have my stuff set up. I am looking at, you know, something that I need to do here. And maybe there is a certain degree of risk. But I feel that that's in my range right now. I can do this.

292
00:26:04,020 --> 00:26:29,120
It's a great, actually, it's a great teacher for life in general. And especially, you know, what you do in the privacy world is super connected with things like that. I mean, because you know that if you, on the whole, in privacy-centric apps, if you mess up once, you've messed up the whole thing because you've revealed what you don't want to reveal.

293
00:26:29,120 --> 00:26:33,900
And then a bit of software can connect that to that, to that, to that, to that, to that, to that.

294
00:26:33,980 --> 00:26:37,260
So you really do have to risk analyze that thing.

295
00:26:37,320 --> 00:26:40,620
I mean, do you ever wonder when you write like white noise?

296
00:26:40,860 --> 00:26:41,240
I don't know.

297
00:26:42,200 --> 00:26:44,840
No, probably a lot of people don't know what white noise is, actually.

298
00:26:45,160 --> 00:26:49,620
Maybe you should say it yourself and then you can, and then I'll ask that question again.

299
00:26:49,620 --> 00:26:50,040
Yeah, perfect.

300
00:26:50,200 --> 00:26:50,540
Yeah, yeah.

301
00:26:50,540 --> 00:26:58,500
So White Noise is kind of the messenger app that we're building on top of the, it's a spec for Nostra.

302
00:26:58,860 --> 00:27:07,880
It doesn't actually write a new protocol, but it is a kind of specification for how to apply the MLS messaging protocol into Nostra's relay and identity system.

303
00:27:08,380 --> 00:27:17,420
Just before you go on, we do talk about Nostra on this show, but in case you didn't, I did a big two-hour chat with Ben Ark, and it's about episode 100-something.

304
00:27:17,420 --> 00:27:19,960
So I'll put it in the show notes just in case.

305
00:27:20,160 --> 00:27:23,080
But I always, Nostra and Bitcoin, two important things to get into.

306
00:27:23,080 --> 00:27:23,540
Very important.

307
00:27:23,740 --> 00:27:24,060
Yeah, yeah.

308
00:27:24,200 --> 00:27:27,020
And Ben does a great job of explaining exactly what Nostra is.

309
00:27:27,060 --> 00:27:28,000
He was there from the very beginning.

310
00:27:28,340 --> 00:27:39,140
So yeah, so effectively, you know, we've got, maybe I'll zoom out a little bit and talk about MLS for just a second so people understand what that is and why that's important to what we're trying to do.

311
00:27:39,140 --> 00:27:48,100
So, you know, if you think about secure messaging, you know, all the way at the beginning, we had, you know, telephones where the switchboard operator could literally listen into everything you were doing.

312
00:27:48,420 --> 00:27:49,440
You know, radio is the same.

313
00:27:49,660 --> 00:27:50,520
Telegraphs were the same.

314
00:27:50,700 --> 00:27:53,660
Like we had to relay messages through other people and they could read those.

315
00:27:53,760 --> 00:28:02,120
And so we back in those days, we came up with like, you know, weird, you know, human readable and understandable codes that we could use to sort of, you know, do cipher text and stuff like that.

316
00:28:02,420 --> 00:28:12,660
All of that stuff evolves over time We get to the Internet The Internet is fully unencrypted for a long time Finally Google kind of makes everybody do it just by virtue of the fact that they had a monopoly with Chrome

317
00:28:12,720 --> 00:28:15,960
And they went, nope, we're doing HTTPS everywhere all the time.

318
00:28:17,060 --> 00:28:19,080
Which is, I mean, it's encrypted, but...

319
00:28:19,080 --> 00:28:20,040
It's encrypted traffic.

320
00:28:20,200 --> 00:28:23,420
So the traffic is encrypted, not necessarily the data on either end.

321
00:28:23,420 --> 00:28:26,820
And why is that bad in messaging?

322
00:28:26,980 --> 00:28:28,780
Because someone can intercept it right now?

323
00:28:28,820 --> 00:28:29,080
Well, no.

324
00:28:29,200 --> 00:28:31,180
I mean, think about this in the earlier internet.

325
00:28:31,180 --> 00:28:33,600
Like all websites used to just be served plain text, right?

326
00:28:33,720 --> 00:28:35,140
And in a lot of cases, that doesn't matter.

327
00:28:35,220 --> 00:28:39,300
If you're just serving your, you know, plain public website in plain text over the line, like who cares?

328
00:28:40,460 --> 00:28:51,500
But you would never have gotten credit card payments on the internet if you hadn't had the ability to somehow take the message on one side in plain text, encrypt it, send it through all the servers in the middle, and then decrypt it on the other side.

329
00:28:52,040 --> 00:28:57,540
And so that's, you know, SSL and, you know, TLS and all that stuff is the kind of encrypted transport on the internet.

330
00:28:57,620 --> 00:28:58,660
And it just became standard.

331
00:28:59,180 --> 00:29:00,640
And what's your opinion on it?

332
00:29:00,640 --> 00:29:01,720
Is that a good standard?

333
00:29:01,720 --> 00:29:03,760
Oh, yeah. I mean, it's an absolutely necessary part of the Internet.

334
00:29:03,900 --> 00:29:06,340
But you wouldn't want to evolve it anymore, is it?

335
00:29:06,340 --> 00:29:07,580
No, I mean, it is an evolving standard.

336
00:29:07,740 --> 00:29:10,300
Like, TLS is continuing to sort of get better over time.

337
00:29:10,420 --> 00:29:14,800
And actually, MLS is based on the same sort of design principles as TLS was.

338
00:29:15,200 --> 00:29:16,560
You know, you should have a handshake.

339
00:29:16,640 --> 00:29:22,200
You should be able to agree with the person you're going to talk to that, like, we support the same features so we can do the thing well.

340
00:29:23,660 --> 00:29:29,720
And, you know, it's just it's an absolute foundational piece of the Internet that no one thinks about, which is a great thing, right?

341
00:29:30,640 --> 00:29:36,100
So in messaging specifically, yeah, we could always encrypt the stuff in the middle.

342
00:29:36,640 --> 00:29:41,800
But if you're going through someone else's servers, there's going to have to be some, you know, you're going to do something there, right?

343
00:29:41,860 --> 00:29:45,740
The server is going to have to know something about you in order to get the message to the other person.

344
00:29:47,020 --> 00:29:57,740
And so we had these protocols early on, like off the record was one of the really earliest ones that we're trying to figure out how do we do encryption for messaging where you can keep an ongoing conversation with someone.

345
00:29:57,740 --> 00:30:03,240
But if someone gets the key, they're not going to be able to decrypt all the messages in the past or all the messages going forward.

346
00:30:03,520 --> 00:30:05,960
You know, somebody gets your key and doesn't say anything and you keep talking.

347
00:30:06,140 --> 00:30:08,280
They just keep decrypting messages as you go.

348
00:30:08,380 --> 00:30:10,120
They can follow you along and see what's going on.

349
00:30:10,320 --> 00:30:15,300
Which is that that was the basis of kind of World War II cryptography, wasn't it?

350
00:30:15,500 --> 00:30:17,520
You only needed one sort of secret.

351
00:30:17,880 --> 00:30:18,080
Yes.

352
00:30:18,120 --> 00:30:22,160
If you figured out how they were creating the secrets, you could decrypt all the messages.

353
00:30:22,360 --> 00:30:22,500
Right.

354
00:30:22,500 --> 00:30:28,340
And this is why, like, the Enigma machine was so difficult for them because they would change the secret every day.

355
00:30:28,760 --> 00:30:34,180
So they had 24 hours to try to figure out how the thing worked, and then it would flip, and they'd start over from scratch.

356
00:30:34,640 --> 00:30:39,880
Right. So it was doing that. I thought the Enigma was just one encryption and then it ran down.

357
00:30:39,880 --> 00:30:46,060
The Enigma was a complicated encryption scheme that it took a long time to figure out what they were doing with it, but it also was a rotating key.

358
00:30:46,340 --> 00:30:50,000
And most encryption schemes were rotating keys. Like, you had some way of kind of formalizing.

359
00:30:50,000 --> 00:30:57,480
Like if you think about the old spy movies where they like have a newspaper and they leave it on the thing and somebody circled a word like that word is the changing key, you know, like.

360
00:30:57,860 --> 00:31:09,760
But would they have to send if it was if in the World War Two case, would they have to send over the rotating keys in one batch in one time so that they knew?

361
00:31:09,820 --> 00:31:11,260
Because that's also an attack.

362
00:31:11,260 --> 00:31:23,460
No. So this is kind of gets into deterministic key generation or the key derivation functions, which is everything that all of our secure messaging stuff online is built on now is you need some deterministic way to go from one key to the next key.

363
00:31:23,700 --> 00:31:27,620
And so either you have to deliver the key manually, you know, and that's non-deterministic.

364
00:31:27,620 --> 00:31:37,880
Like that's, you know, you circle a word in a newspaper or somebody sends you a message in another communication channel, you know, a messenger shows up on a boat and tells you what the next month's worth of keys are going to be.

365
00:31:38,140 --> 00:31:38,920
That's one thing.

366
00:31:38,920 --> 00:31:47,040
The other thing is start with a key and then start with some, you know, secret to you and your group, but known way of deriving out the next key.

367
00:31:47,460 --> 00:31:53,920
Some way of taking a word and, you know, it's like on the first letter, we're going to add 12 and then get to the next letter.

368
00:31:54,040 --> 00:31:55,720
On the next one, we're going to add 14.

369
00:31:56,060 --> 00:31:57,840
On the other one, we're going to subtract 7.

370
00:31:57,980 --> 00:32:01,660
Like you just kind of, you can manipulate a thing in a certain way to get a new key.

371
00:32:01,660 --> 00:32:03,040
Is that two things?

372
00:32:03,180 --> 00:32:07,520
So is that the word and the way we decrypt the word?

373
00:32:07,520 --> 00:32:09,160
And they can be sent separately.

374
00:32:10,040 --> 00:32:12,240
Because if someone finds those two things.

375
00:32:12,740 --> 00:32:13,560
They can do whatever they want.

376
00:32:13,740 --> 00:32:14,260
Yeah, exactly.

377
00:32:14,600 --> 00:32:15,640
They have the keys to the kingdom.

378
00:32:15,640 --> 00:32:18,300
So presumably you want to split that into as many things as you can.

379
00:32:18,460 --> 00:32:19,280
Exactly, exactly.

380
00:32:19,500 --> 00:32:24,720
So this is, I mean, cryptography in general is cat and mouse, right?

381
00:32:24,760 --> 00:32:25,880
It's always going to be cat and mouse.

382
00:32:25,960 --> 00:32:32,500
There's always going to have to be changing standards because people will figure out ways, you know, eventually to sort of get around some of these things.

383
00:32:32,500 --> 00:32:39,120
you know yes there is extremely strong cryptography but as computers get better like many of these

384
00:32:39,120 --> 00:32:45,260
you know cipher suites and things are just becoming obsolete right and with this idea of

385
00:32:45,260 --> 00:32:48,920
quantum computing it's going to make a whole bunch more stuff obsolete and so we have to come up with

386
00:32:48,920 --> 00:32:54,740
ways to kind of fight against that so um let's jump back to messaging so you've got like off the

387
00:32:54,740 --> 00:32:59,100
record was like one of the original protocols that specified how can we do you know forward

388
00:32:59,100 --> 00:33:03,900
secrecy and post-compromised security. So how do we keep keys rotating in a way that keeps the

389
00:33:03,900 --> 00:33:07,860
conversation secure so that even if an attacker got one of the keys, they are only going to get

390
00:33:07,860 --> 00:33:13,400
a tiny slice of our conversation. The signal protocol kind of grew out of the off-the-record

391
00:33:13,400 --> 00:33:18,420
protocol. Moxie Marlin's bike, and I can't remember the other guy's name, so sorry, but

392
00:33:18,420 --> 00:33:24,040
they came up with the signal protocol, which is the kind of classic double ratchet algorithm that

393
00:33:24,040 --> 00:33:28,380
you hear a lot of people say double ratchet encryption. And the idea there is that there's

394
00:33:28,380 --> 00:33:33,000
actually multiple key derivation functions. And you've got a way to agree on a key, a shared key

395
00:33:33,000 --> 00:33:37,900
between two people that you create one key derivation function. So that's like kind of a

396
00:33:37,900 --> 00:33:44,880
trunk of a tree and that's generating a certain key. And then using that key, people derive out

397
00:33:44,880 --> 00:33:50,080
other keys. And so you and I both have then two chains of messages. One side is a sender side,

398
00:33:50,180 --> 00:33:53,860
one side is a receiver side. So my sender is your receiver, your sender is my receiver.

399
00:33:53,860 --> 00:33:59,640
and that's how we actually send messages we generate keys on those small chains to encrypt

400
00:33:59,640 --> 00:34:04,060
the actual message and then every now and then we rotate the key in the middle and that basically

401
00:34:04,060 --> 00:34:07,720
changes what's called the epoch the whole thing moves forward and we start doing the thing again

402
00:34:07,720 --> 00:34:14,980
um just a question in that system in like in signal are you signal if you if you delete a

403
00:34:14,980 --> 00:34:20,220
message is it deleted is it so think about it this way in indian encrypted messaging

404
00:34:20,220 --> 00:34:26,820
the idea is you i have a message i encrypt it so imagine i just put it in a box and lock the box

405
00:34:26,820 --> 00:34:31,640
you have the key to the box right i can then send the box through the mail through whatever channel

406
00:34:31,640 --> 00:34:34,640
i want to and the person in the middle doesn't have the box so they can't see what's inside

407
00:34:34,640 --> 00:34:40,060
you get the box you open the box and you then have the thing you're the only person that has

408
00:34:40,060 --> 00:34:44,440
the thing now um and obviously if we're talking about messaging i've got a copy on my end and

409
00:34:44,440 --> 00:34:49,000
you have a copy on your end no one in the middle has any like they've got an encrypted copy but

410
00:34:49,000 --> 00:34:52,800
They can't do anything with it unless they get a key to unlock the box in the middle.

411
00:34:54,060 --> 00:34:55,580
Could they in the future, though?

412
00:34:55,640 --> 00:34:55,980
Absolutely.

413
00:34:56,340 --> 00:34:56,760
Yeah, absolutely.

414
00:34:56,760 --> 00:35:02,020
So all our messages that we're sending now are potentially revealable in 40 years.

415
00:35:02,320 --> 00:35:02,840
Yeah, 100%.

416
00:35:02,840 --> 00:35:04,600
And that's what they say, basically.

417
00:35:04,760 --> 00:35:13,520
When quantum arrives in a usable, actual functioning format, all of the messaging is going to be immediately decrypted.

418
00:35:14,380 --> 00:35:18,180
And I mean, that's on one side super scary for a lot of people that have lots of skeletons.

419
00:35:18,180 --> 00:35:21,080
that you know you know they're never going to get rid of those right like the internet's a giant

420
00:35:21,080 --> 00:35:25,960
copy machine yeah totally so i mean you're always going to fall foul of a law in the end because

421
00:35:25,960 --> 00:35:32,480
laws don't 100 and the question is more do is it going to be something people care enough about

422
00:35:32,480 --> 00:35:35,980
you know going back that far to look for like no government's going to go and be like oh you bought

423
00:35:35,980 --> 00:35:42,380
drugs in college on you know this platform they might say you didn't pay tax in 1980 whatever

424
00:35:42,380 --> 00:35:48,880
that's potentially true or more to the point they if you are a target if they target then they'll

425
00:35:48,880 --> 00:35:52,940
use yeah yeah all bets are off like they'll use all of that stuff against you so how would you

426
00:35:52,940 --> 00:35:57,800
we're still going back to what you were saying how would you mitigate against that in its current

427
00:35:57,800 --> 00:36:04,240
form then say you're a signal user and you you want to try and for you know so signals actually

428
00:36:04,240 --> 00:36:08,620
already helping protect you against this because they are already they've already moved to a

429
00:36:08,620 --> 00:36:13,660
quantum resistant cypher suite. So it's a collection of, you know, algorithms that you

430
00:36:13,660 --> 00:36:19,540
use to do this, which is, you know, generally assigning a scheme and a hashing scheme that fit

431
00:36:19,540 --> 00:36:23,920
together. And they have moved to algorithms that are quantum resistant. So they're much harder for

432
00:36:23,920 --> 00:36:28,160
quantum computers to solve. How do you know something's quantum resistant? A bunch of math,

433
00:36:28,280 --> 00:36:33,200
you have to have really giant gigabrain cryptographers do formal proofs in mathematics

434
00:36:33,200 --> 00:36:39,240
to show that this is not only, you know, very difficult for a classical, you know, digital

435
00:36:39,240 --> 00:36:43,200
computer to solve, but it's also incredibly difficult for, you know, the way we think

436
00:36:43,200 --> 00:36:44,720
about how qubits are going to solve things.

437
00:36:45,460 --> 00:36:50,560
So if we know what's quantum resistant, how come we don't do quantum then?

438
00:36:50,940 --> 00:36:53,100
How can we not already have quantum computers?

439
00:36:53,100 --> 00:36:53,740
Laws of physics.

440
00:36:53,960 --> 00:36:56,080
It's like really, really hard to get this stuff done properly.

441
00:36:56,280 --> 00:36:57,120
You know, you've got...

442
00:36:57,120 --> 00:36:58,840
So theoretically, we know what quantum is.

443
00:36:58,960 --> 00:37:00,380
Theoretically, we know exactly what it is.

444
00:37:00,380 --> 00:37:04,620
we're actually in the position of like how do we take that and make it physical reality break that

445
00:37:04,620 --> 00:37:08,900
down for me so i can understand it what is i don't know quantum as well but like my understanding at

446
00:37:08,900 --> 00:37:15,900
a very very high level is you are going from zeros and ones to something that can be you know quantum

447
00:37:15,900 --> 00:37:22,000
super superimposed and so you can have multiple states of things at the same time um and that

448
00:37:22,000 --> 00:37:27,340
gives you the ability to have way more states possible thus you can have you know a computer

449
00:37:27,340 --> 00:37:29,680
that can do way more calculations way, way, way, way faster.

450
00:37:30,000 --> 00:37:35,620
And is that something to do with having to use minus 280 degrees or something?

451
00:37:35,980 --> 00:37:36,500
That's exactly, yes, correct.

452
00:37:36,740 --> 00:37:39,620
And they can't keep the chip stable enough.

453
00:37:39,960 --> 00:37:41,020
You know, think about it as a circuit.

454
00:37:41,320 --> 00:37:42,800
You know, a circuit is either open or closed.

455
00:37:42,900 --> 00:37:44,320
And that's basically what a digital computer is.

456
00:37:44,340 --> 00:37:45,100
It's either open or closed.

457
00:37:45,160 --> 00:37:48,960
And we figured out how to do, you know, transistors that are either open or closed.

458
00:37:48,960 --> 00:37:49,800
And they're very stable.

459
00:37:50,540 --> 00:37:52,540
In quantum, it's incredibly unstable.

460
00:37:53,240 --> 00:37:54,980
And the thing has way more states.

461
00:37:55,080 --> 00:37:59,720
And so it's just like, and so they try to get it super, super cold to slow everything down.

462
00:38:00,000 --> 00:38:05,200
and they're still working on ways to like how do we error correct these things because if you think

463
00:38:05,200 --> 00:38:09,640
about it a transistor an electrical transistor if it gets hit by a photon in the wrong way it will

464
00:38:09,640 --> 00:38:14,040
bit flip and like that's a thing that certain parts of computing have you know have to worry

465
00:38:14,040 --> 00:38:20,020
about like very low probability but it does happen yeah yeah um and so that in quantum is still

466
00:38:20,020 --> 00:38:25,200
incredibly difficult and you know what bit flipping i've thought that that is the same

467
00:38:25,200 --> 00:38:32,760
same process as um dna evolution i think there's a four i this is purely my own idea right yeah i

468
00:38:32,760 --> 00:38:37,780
mean it would make sense that you can bit flip dna and that's what causes things to change i mean

469
00:38:37,780 --> 00:38:42,900
it's it's conceptually the exact same process you know mutation is basically a bit flip yeah but

470
00:38:42,900 --> 00:38:46,880
it's weird when you think about it in computer terms because you don't imagine that there might

471
00:38:46,880 --> 00:38:53,900
be a piece of well a photon that can actually kind of collide with your computer and do something to

472
00:38:53,900 --> 00:38:58,360
it yeah yeah yeah that's weird yeah that is weird but they are fundamentally it is still electrons

473
00:38:58,360 --> 00:39:01,480
right like you're still dealing with electrons in a digital computer or a quantum computer or

474
00:39:01,480 --> 00:39:07,960
whatever so it's um you know and it's interesting to know like our computing at the moment is is all

475
00:39:07,960 --> 00:39:13,120
incredibly deterministic right it's if if else loops it's you know loops of going through you

476
00:39:13,120 --> 00:39:18,040
know lists of things you know it's why computer programmers are quite you know rigid deterministic

477
00:39:18,040 --> 00:39:22,300
people they can read through this stuff in very logical patterns the interesting thing i think

478
00:39:22,300 --> 00:39:27,620
about quantum is that like, I think it does open up way more probabilistic computing. And, and the

479
00:39:27,620 --> 00:39:32,540
stuff that we're getting at with AI now is going in that direction in some ways, like, and, and some,

480
00:39:32,720 --> 00:39:37,800
in some degrees, like in a really weird, basic way, Nostra is kind of similar. Like there is no

481
00:39:37,800 --> 00:39:42,740
global feed of reality. Like there is no one single truth here. You've got all these relays

482
00:39:42,740 --> 00:39:48,040
and all this data. And what you see is you, you know, your constructed reality around you.

483
00:39:48,040 --> 00:39:52,420
and that's much more like the real world right like we don't know what real reality is because

484
00:39:52,420 --> 00:39:56,540
we just can't be everywhere all at once and take in all the information we're still going back to

485
00:39:56,540 --> 00:40:00,120
what you're saying but you're making me think of other things i was about to ask that exact question

486
00:40:00,120 --> 00:40:06,040
but i've got another one to do with ai and quantum based on what you've just said because when i look

487
00:40:06,040 --> 00:40:12,840
at ai now i think it's quite primitive because it's just it's basically looking at everything

488
00:40:12,840 --> 00:40:17,100
and then trying to make a prediction about what should be said based on that so there's no

489
00:40:17,100 --> 00:40:23,460
actual understanding there unless the basis of understanding is that it could be but that seems

490
00:40:23,460 --> 00:40:30,140
a bit brutal to me yeah yeah so how would that how would quantum affect that would that be

491
00:40:30,140 --> 00:40:36,480
because i when i look at ai now i think that doesn't seem like intelligence that just seems

492
00:40:36,480 --> 00:40:41,460
like a computer program running and guessing things until it gets it right based on information

493
00:40:41,460 --> 00:40:47,120
so will quantum make that a little bit more brain light do you think or am i wrong are our brains

494
00:40:47,120 --> 00:40:52,980
just stupid guessing machines no i mean i actually there's a really good framework that uh a guy

495
00:40:52,980 --> 00:40:59,060
i think it's john vervecki i get his his surname wrong yeah yeah yeah some max was talking about

496
00:40:59,060 --> 00:41:04,860
him yeah he's awesome and he has a really great way of thinking about uh cognition and intelligence

497
00:41:04,860 --> 00:41:10,540
and uh consciousness and i think it's a really really great way to kind of frame this question

498
00:41:10,540 --> 00:41:14,720
which is there are lots of different types of knowing, right? Like knowing that a cat is a

499
00:41:14,720 --> 00:41:19,200
mammal is a fact, right? Like you just, it's a thing that you can or can't, you know, know if

500
00:41:19,200 --> 00:41:23,080
you've never seen it before, you are ignorant of that fact, but it is no matter what, right?

501
00:41:23,720 --> 00:41:27,720
But there's other types of knowing. And AIs right now are incredibly good at that type of knowing.

502
00:41:27,820 --> 00:41:31,740
They've read every book. They've read everything on the internet. They have, you know, they know a

503
00:41:31,740 --> 00:41:37,840
lot of things about the world, but they've never, you know, walked through a field at sunset,

504
00:41:37,840 --> 00:41:39,500
and felt the breeze in their face.

505
00:41:39,640 --> 00:41:40,880
They don't know what that feels like.

506
00:41:41,180 --> 00:41:42,920
And they don't know procedural things like,

507
00:41:43,160 --> 00:41:45,080
what is it to swim?

508
00:41:46,040 --> 00:41:47,580
They can read all the books about swimming,

509
00:41:47,680 --> 00:41:48,920
but until you've been dumped in a pool,

510
00:41:49,300 --> 00:41:50,460
you don't know what swimming is.

511
00:41:51,040 --> 00:41:52,760
And so there's different types of knowing.

512
00:41:52,980 --> 00:41:54,160
And those are just a couple of examples.

513
00:41:54,300 --> 00:41:56,080
There's actually more that he goes into.

514
00:41:56,080 --> 00:42:01,140
But I think as AI gets progressively better,

515
00:42:01,380 --> 00:42:02,820
it's obviously going to get better

516
00:42:02,820 --> 00:42:05,400
at being able to crunch all that data, right?

517
00:42:05,440 --> 00:42:07,000
But I kind of feel like we're approaching

518
00:42:07,000 --> 00:42:10,660
the upper limit of like they've already crunched the entire internet right they're going to be able

519
00:42:10,660 --> 00:42:15,300
to form maybe better connections between things we might be able to give the ai an ability to sort

520
00:42:15,300 --> 00:42:21,800
of start to make um you know leaps between things try to be a bit more creative try to connect dots

521
00:42:21,800 --> 00:42:27,380
that don't seem connected which actually humans are really good at ai still aren't um but i don't

522
00:42:27,380 --> 00:42:31,880
know that we're ever going to get to a point where they've got that those other types of knowing that

523
00:42:31,880 --> 00:42:36,760
really only humans can have because we are going through the world and trying to i mean we have an

524
00:42:36,760 --> 00:42:40,020
incredible sensory organ, right? Like we have all these, we have ears, we have eyes, we have nose,

525
00:42:40,100 --> 00:42:44,980
we taste, our skin is, you know, feeling things constantly. We're actually not paying any attention

526
00:42:44,980 --> 00:42:49,840
to like 99.9% of that data. We just throw it all away instantaneously. And our brain is making

527
00:42:49,840 --> 00:42:55,120
these calculations of what matters to me right now. What can I, you know, safely throw away?

528
00:42:55,420 --> 00:43:00,480
I want to conserve as much energy as possible, but then how do I take all of what I am looking at,

529
00:43:00,780 --> 00:43:06,420
fit it into what my memory and my, you know, myself says is important. And then what am I

530
00:43:06,420 --> 00:43:10,540
going to do with that information um it's kind of incredible how much your brain is doing it could

531
00:43:10,540 --> 00:43:17,240
be though that what you've just described could still be that type one error error correction

532
00:43:17,240 --> 00:43:21,760
guessing you just add more senses i mean you described our senses there and it just occurred

533
00:43:21,760 --> 00:43:28,100
to me well yeah walking through and feeling the breeze is a function of feeling and smelling and

534
00:43:28,100 --> 00:43:35,960
sound so true maybe if you create a humanoid with receptors for those things maybe maybe we'll get

535
00:43:35,960 --> 00:43:41,080
to the bandwidth that bandwidth required but like i think the bandwidth required there is enormous

536
00:43:41,080 --> 00:43:45,240
it's kind of unbelievable right maybe it must be sure it's a question of time i suppose is the

537
00:43:45,240 --> 00:43:50,480
bigger thing is like how close are these things and i think i would posit that like agi is a long

538
00:43:50,480 --> 00:43:56,080
way out um i think we will get stuff that feels very very intelligent i mean we already have stuff

539
00:43:56,080 --> 00:44:00,600
that feels very intelligent like you can sit there and talk to claude or chachupiti and feel like

540
00:44:00,600 --> 00:44:04,340
you're talking to a human most of the time yeah um you know they do hallucinate still they do mess

541
00:44:04,340 --> 00:44:10,920
stuff up but you know they are very very intelligent at pretending to be human i i was thinking often

542
00:44:10,920 --> 00:44:16,560
the ai i see now is a bit like don't take this the wrong way because i do there's an autistic

543
00:44:16,560 --> 00:44:22,520
person in my life i'm not being rude to autistic people but autistic people look at what the world's

544
00:44:22,520 --> 00:44:26,840
happening and they try and mimic it and it's a bit like that with with some ais you realize

545
00:44:26,840 --> 00:44:34,820
the only difference is people um ai says things very authoritatively every single time yes it

546
00:44:34,820 --> 00:44:39,820
never has a conversation with you where it's like i'm not sure it could you know which is what humans

547
00:44:39,820 --> 00:44:44,540
do because we don't know everything but ai unfortunately thinks currently that it knows

548
00:44:44,540 --> 00:44:47,940
everything yeah well and when you correct a certain maturity actually so if you think about

549
00:44:47,940 --> 00:44:52,440
it like ai is sort of like an early teenager who's like no i know how everything works you know so i

550
00:44:52,440 --> 00:44:56,440
think as they mature they will get better at that but yeah you're right but when they when you correct

551
00:44:56,440 --> 00:45:03,380
them they it's bizarre because it's non it's very unhuman to go if you if i told if i said oh you

552
00:45:03,380 --> 00:45:08,620
know bitcoin's got 23 million you know there's 23 million bitcoins available and you're like

553
00:45:08,620 --> 00:45:14,500
it's actually 21 and i go oh yeah you're correct bitcoin's only got 20 you know like

554
00:45:14,500 --> 00:45:21,400
no human would do that right i would be like bitcoin's got a large supply of is it 21 million

555
00:45:21,400 --> 00:45:22,980
and you'd be like, yeah, it's 21 million.

556
00:45:23,160 --> 00:45:25,300
You know, it's like, but they do not do that.

557
00:45:26,580 --> 00:45:26,940
Right.

558
00:45:27,020 --> 00:45:28,300
Let's go back to where we were.

559
00:45:29,540 --> 00:45:30,560
Messaging, private messaging.

560
00:45:30,700 --> 00:45:30,900
Okay.

561
00:45:31,000 --> 00:45:32,820
So we got up to the Signal Protocol.

562
00:45:33,160 --> 00:45:33,340
Yes.

563
00:45:33,580 --> 00:45:36,420
Signal Protocol is, I mean, it was groundbreaking, right?

564
00:45:36,440 --> 00:45:37,620
Like it was a really big deal.

565
00:45:38,060 --> 00:45:43,300
And, you know, the guys were able to license that technology to Meta for,

566
00:45:43,520 --> 00:45:44,940
you know, they put it into WhatsApp.

567
00:45:45,320 --> 00:45:46,640
It was part of like Google Messenger.

568
00:45:47,020 --> 00:45:47,980
Like Signal Protocol is everywhere.

569
00:45:48,060 --> 00:45:48,960
It's not just in Signal.

570
00:45:49,740 --> 00:45:51,340
You know, the underlying protocol itself.

571
00:45:51,400 --> 00:45:56,360
the main difference between what signal does and what the other guys do is they might be using the

572
00:45:56,360 --> 00:46:02,420
same encryption scheme for the actual messages but the metadata around the messages is actually

573
00:46:02,420 --> 00:46:07,800
almost more important and more sensitive than the actual content of the message um so there's

574
00:46:07,800 --> 00:46:13,540
you know confidentiality and then there's privacy and confidentiality is like you know our messages

575
00:46:13,540 --> 00:46:18,800
are protected privacy is actually who's talking to who when like maybe why they're talking to each

576
00:46:18,800 --> 00:46:24,020
other. And so both of those things are really important if you actually want to talk about

577
00:46:24,020 --> 00:46:30,540
the privacy of a scheme. So WhatsApp perfectly fixes confidentiality. They're end-to-end

578
00:46:30,540 --> 00:46:37,340
encrypted. Their AI is in the chat with you, by the way. You can turn that off, but you're also

579
00:46:37,340 --> 00:46:40,540
going to turn off a bunch of other useful features because they're all grouped together in the same

580
00:46:40,540 --> 00:46:45,380
panel. I've never thought of WhatsApp as being a good place. I mean, they say they're end-to-end

581
00:46:45,380 --> 00:47:01,140
I don believe I mean it a very much trust me bro type of scenario Like they don tell you it closed source They do tell you that like yes under the hood you know we know we using signal protocol but like there a bunch of ways to break around that and say well we just put our own group member who invisible in the chat and then they

582
00:47:01,140 --> 00:47:06,220
also part of the conversation um that's my base case there is like there's just a bot in the chat

583
00:47:06,220 --> 00:47:10,120
i mean would that be called a backdoor is a backdoor something completely i mean a backdoor

584
00:47:10,120 --> 00:47:13,480
is something slightly different like because they're closed source because they own the whole

585
00:47:13,480 --> 00:47:17,580
scheme it's not even a backdoor it's just like this is part of the software like this is how

586
00:47:17,580 --> 00:47:23,620
it's literally built um you know a backdoor is more uh we have an encryption scheme but there's

587
00:47:23,620 --> 00:47:27,940
something that you know no one knows about that we can just turn it off or we can enter the chat

588
00:47:27,940 --> 00:47:33,260
whenever we want through this kind of super user access because i saw on online this week that

589
00:47:33,260 --> 00:47:41,120
um the eu are pushing to add backdoors to signal but yes chat control is back but funny enough i

590
00:47:41,120 --> 00:47:45,300
I use something called Molly, which is an open source version of Signal.

591
00:47:45,680 --> 00:47:46,560
Yes, I've heard of this.

592
00:47:46,660 --> 00:47:46,940
Yeah, yeah, yeah.

593
00:47:46,940 --> 00:47:47,080
Right.

594
00:47:47,720 --> 00:47:49,680
Don't ask me whether it's better or worse.

595
00:47:49,760 --> 00:47:53,120
I just read the blurb and saw the groups and thought, okay, I'll do it.

596
00:47:53,160 --> 00:47:54,300
My daughter uses it.

597
00:47:55,300 --> 00:48:01,480
But what I find weird is if you – because obviously if Signal added a backdoor, Molly would not, right?

598
00:48:01,480 --> 00:48:02,400
Well, here's the thing.

599
00:48:04,020 --> 00:48:09,240
Well, Molly, if it wants to continue working with Signal, would almost certainly have to be very, very compatible.

600
00:48:09,240 --> 00:48:17,300
Like it would be very hard for Signal to add a backdoor in some way that, you know, Molly wouldn't sort of break on, right?

601
00:48:17,300 --> 00:48:22,040
Because when you're doing these encryption schemes, I mean, what I'm talking about now is like the actual encryption protocol.

602
00:48:22,560 --> 00:48:25,480
When you're doing them, if you get it wrong, it just doesn't work anymore.

603
00:48:25,600 --> 00:48:26,640
Like it completely stops working.

604
00:48:27,080 --> 00:48:39,040
And so, you know, if you've done the key derivation function slightly differently so that you can have an extra key somewhere, I mean, yes, there's potentially ways that are like, you know, undetectable or that don't have side effects.

605
00:48:39,040 --> 00:48:40,760
But a lot of these things are just math equations.

606
00:48:40,960 --> 00:48:43,820
And so if you change something, it changes something.

607
00:48:44,360 --> 00:48:46,380
Would that be a signal then that they have changed something?

608
00:48:46,940 --> 00:48:54,140
If a fork of signal couldn't interact with signal in a private way, would that be a signal that they've been…

609
00:48:54,140 --> 00:48:54,420
Potentially.

610
00:48:54,600 --> 00:48:55,120
Yeah, potentially.

611
00:48:56,240 --> 00:48:58,920
And again, I think signal is actually a great player.

612
00:48:59,040 --> 00:49:00,500
I think they're a good actor in the space.

613
00:49:00,640 --> 00:49:04,840
I think they do everything they possibly can to minimize the amount of metadata they have about their users.

614
00:49:05,340 --> 00:49:06,580
You've got to sign up with a phone number.

615
00:49:06,580 --> 00:49:12,900
They know they can connect you through that phone number to this user, but they actually minimize as much as possible in their servers what they have.

616
00:49:13,220 --> 00:49:14,980
And they try to delete it as fast as possible.

617
00:49:15,460 --> 00:49:21,360
And so I think, you know, they are they are good actors, but, you know, they're a centralized player.

618
00:49:21,360 --> 00:49:23,880
Right. They've got a 50 million dollar server budget.

619
00:49:24,080 --> 00:49:25,280
They've got a U.S. foundation.

620
00:49:25,540 --> 00:49:30,080
They have a CEO and all three of those things can fall apart in one way or another.

621
00:49:30,200 --> 00:49:31,860
And then Signal disappears for all of us.

622
00:49:32,080 --> 00:49:34,560
Why did they use phones just as a convenience?

623
00:49:34,920 --> 00:49:35,420
Yeah, absolutely.

624
00:49:35,420 --> 00:49:40,880
i mean so they could have done like nostr uses nothing you just you just create a thing there's

625
00:49:40,880 --> 00:49:45,580
a few things i use well nostr has to use phones too i mean it's it's whatever the computer is that

626
00:49:45,580 --> 00:49:50,020
you're actually no no i mean using your phone number as an identifier yeah they had to pick

627
00:49:50,020 --> 00:49:56,600
something and they um because it's a pretty full-on identifier isn't it yeah to them it's

628
00:49:56,600 --> 00:49:59,760
actually also just a spam protection measure because they have to run all these servers that

629
00:49:59,760 --> 00:50:03,820
do all this stuff um that don't do a lot of encryption and you know moving stuff around

630
00:50:03,820 --> 00:50:07,780
it's expensive and so they actually spend i think somebody told me recently that it's

631
00:50:07,780 --> 00:50:14,420
it's high it's like 20 or 25 percent of their whole budget every year is um sms you know just

632
00:50:14,420 --> 00:50:19,100
those verification sms's and that's just to stop spam basically you know because otherwise people

633
00:50:19,100 --> 00:50:23,980
would just create all these numbers that are fake use those create accounts you know and figure out

634
00:50:23,980 --> 00:50:28,600
ways to spam the network horrifically now signal has to worry about that because they have to pay

635
00:50:28,600 --> 00:50:34,700
all its servers right um nostr kind of gets around this by saying yeah there's spam and it's definitely

636
00:50:34,700 --> 00:50:38,780
going to affect everybody but it's so spread out that you know mostly it doesn't really have much

637
00:50:38,780 --> 00:50:42,920
of an effect now as we grow i bet that'll become a problem and we'll have to figure that out but

638
00:50:42,920 --> 00:50:48,540
like right now it's not really what do you mean spread out can you explain that yeah so imagine i

639
00:50:48,540 --> 00:50:53,000
mean signal has i mean it's a fleet of servers but it's all controlled by them they have to run the

640
00:50:53,000 --> 00:50:58,540
servers, they have to pay the server bills. Nostra relays are very dumb. They're not very,

641
00:50:58,540 --> 00:51:03,720
you know, resource intensive. It's effectively just a simple database. All we ask relays to do

642
00:51:03,720 --> 00:51:09,480
is accept events that are valid, you know, cryptographically signed validly. And when we

643
00:51:09,480 --> 00:51:12,900
ask you for something, if you have it, give it back. That's it. You don't have to do anything

644
00:51:12,900 --> 00:51:16,380
else. You don't have to compute stuff. You don't have to do any cryptography. Like all you're doing

645
00:51:16,380 --> 00:51:20,060
is taking something storing in a database and then returning a response if you, you know, have

646
00:51:20,060 --> 00:51:27,240
something. So that makes running a Nasta relay quite cheap up to, you know, fairly large sizes.

647
00:51:27,860 --> 00:51:32,360
And certainly, yes, once we get to, you know, a mass adoption point where we have millions and

648
00:51:32,360 --> 00:51:37,620
millions of users running a very large relay is going to be costly. But the people who are going

649
00:51:37,620 --> 00:51:41,400
to run very large relays are people who are running businesses where that very large relay

650
00:51:41,400 --> 00:51:45,980
that they're paying for directly improves the UX for their user base. And that's going to make

651
00:51:45,980 --> 00:51:50,840
total sense for them. But there's all sorts of ways you can run small relays. You can control

652
00:51:50,840 --> 00:51:54,860
the amount of data that comes into them. You can whitelist who's allowed to post and who's not.

653
00:51:55,680 --> 00:52:00,740
And so there are a myriad of ways where running relays is actually super cheap and you can

654
00:52:00,740 --> 00:52:03,460
actually run them on your phone. You can run them on a laptop, like no big deal.

655
00:52:04,280 --> 00:52:10,340
And I think the world we head towards is, yes, there's a number of very large relays,

656
00:52:10,440 --> 00:52:15,280
but in a power law dynamic, it falls off fairly fast. And there's a long tail of relays that will

657
00:52:15,280 --> 00:52:21,740
accept most messages and also are completely commodity and don't matter to the actual end

658
00:52:21,740 --> 00:52:25,560
user. You know, if you're connected to three or four of these things and one or two go down because

659
00:52:25,560 --> 00:52:30,840
of spam, who cares? You just switch. You just move around. And so especially for our use case,

660
00:52:30,980 --> 00:52:34,300
it's really important to us that we never run servers. You know, like it's we don't want to

661
00:52:34,300 --> 00:52:38,520
have any servers. We're like we're client side software. You know, when the government shows up

662
00:52:38,520 --> 00:52:41,560
and they're like, give us, you know, all the data on your users. We're like, we don't have any.

663
00:52:41,560 --> 00:52:45,900
all you can do is go to github and see our code base that's it we've got to talk about that later

664
00:52:45,900 --> 00:52:52,240
let me remind that yeah because you know i think we've just seen a ruling in the u.s which might

665
00:52:52,240 --> 00:52:57,320
be bad news for that in the future makes me very nervous right okay we're definitely definitely

666
00:52:57,320 --> 00:53:02,860
coming back to that let's finish the so the signal story currently still a good app to use i think

667
00:53:02,860 --> 00:53:08,720
it's great yeah i think it's great i think they do a good job um you know is it foolproof absolutely

668
00:53:08,720 --> 00:53:14,160
not like nothing is foolproof like if you want actual privacy you need to go way away from a lot

669
00:53:14,160 --> 00:53:19,500
of other humans and have a private conversation in the middle of nowhere and even then you know

670
00:53:19,500 --> 00:53:27,200
look up yeah exactly there could be a camera above you um so you know it's tough to get real real

671
00:53:27,200 --> 00:53:33,740
privacy these days um but i think you know signal does about as good as you could possibly hope um

672
00:53:33,740 --> 00:53:40,920
for most use cases. I'm glad about that. I've got white noise, but obviously most people I know

673
00:53:40,920 --> 00:53:45,360
aren't on white noise. So we'll get there. It's early. You know, one other thing that I'd love

674
00:53:45,360 --> 00:53:50,120
to mention here is that, you know, a lot of people focus in on the actual encryption scheme and use

675
00:53:50,120 --> 00:53:54,780
or, you know, things like that. Um, it's really important when you're thinking about privacy,

676
00:53:55,000 --> 00:53:58,900
you know, to think about like to do a proper threat assessment, right? Like who are you

677
00:53:58,900 --> 00:54:04,060
actually trying to maintain privacy from like are you just trying to not stick out in metas

678
00:54:04,060 --> 00:54:08,340
profiling things so you don't get you know garbage ads or so that they don't know more about you than

679
00:54:08,340 --> 00:54:13,800
you want them to know that's like one thing if you're against a state level actor you're kind

680
00:54:13,800 --> 00:54:17,700
of screwed like there's no way you're going to be able to get away from the nsa like if they really

681
00:54:17,700 --> 00:54:22,520
want to find you they will find you um and they will know everything about you really even even

682
00:54:22,520 --> 00:54:25,800
Even decrypting messages?

683
00:54:26,500 --> 00:54:32,120
It's less about decrypting the messages as it is about figuring out ways that they don't have to decrypt the messages, right?

684
00:54:32,180 --> 00:54:41,060
Like they will, you know, if you're important enough, they've got a portfolio of zero day exploits that they've never used before that they'll figure out how to get onto your phone one way or another.

685
00:54:41,280 --> 00:54:44,260
And then they'll just have full root access to everything on your phone.

686
00:54:44,260 --> 00:54:46,900
And then they've got everything on the phone.

687
00:54:47,240 --> 00:54:50,760
And so, you know, when we're talking about Signal, you have a decrypted message.

688
00:54:51,240 --> 00:54:52,340
You encrypt it.

689
00:54:52,340 --> 00:54:55,860
you send it, the other person decrypts it on their side, they've got a decrypted message.

690
00:54:56,220 --> 00:54:59,520
If the person that is attacking you is already in your phone at root level,

691
00:54:59,820 --> 00:55:01,420
they have your messages, like they have everything.

692
00:55:01,840 --> 00:55:03,940
And graphene and those kind of things?

693
00:55:04,200 --> 00:55:08,220
Graphene is a great project and they do, and this is the thing, like

694
00:55:08,220 --> 00:55:12,560
the better that a phone and the operating system protects you against zero days,

695
00:55:12,720 --> 00:55:17,980
the safer you are. And so things like lockdown mode on iOS are actually really, really important.

696
00:55:18,180 --> 00:55:21,200
Like I think everybody should have lockdown mode on. It makes a really big difference.

697
00:55:21,660 --> 00:55:32,580
You know, I've talked to security researchers, you know, good friends that find these things in the wild and that do a lot of the research on Pegasus and are the people that point out when they found people who have been affected by Pegasus.

698
00:55:33,060 --> 00:55:36,960
And universally, the first thing he says is always like, just turn on lockdown mode.

699
00:55:37,040 --> 00:55:38,160
It like makes almost no difference.

700
00:55:38,160 --> 00:55:43,140
And it makes a huge difference to like to your use case, but it makes a huge difference to your actual security.

701
00:55:43,640 --> 00:55:43,880
Hold a minute.

702
00:55:44,060 --> 00:55:45,460
Lockdown on mine.

703
00:55:45,640 --> 00:55:46,740
I've got graphene.

704
00:55:46,860 --> 00:55:46,960
Yeah.

705
00:55:47,000 --> 00:55:48,800
It's a button you actually have to press, I think.

706
00:55:48,800 --> 00:55:53,860
so i don't know exactly what the difference is and like you know i know graphing is very good i

707
00:55:53,860 --> 00:55:58,080
think it's actually more lockdown mode like by default i think there's also some extreme settings

708
00:55:58,080 --> 00:56:04,020
lockdown mode on ios is just this mode where it's a one switch and it turns on a bunch of features

709
00:56:04,020 --> 00:56:21,060
it won load images by default it won you know do a bunch of things that look when i go to turn nine off yeah is that lockdown is that different i thought that was i thought that means scramble the whole phone and so that yeah i don know exactly what that is but lockdown

710
00:56:21,060 --> 00:56:24,780
mode on ios is basically hey we're going to turn on a bunch of extra features that are going to

711
00:56:24,780 --> 00:56:29,080
reduce your user experience a little bit you know we're going to not load images by default we're

712
00:56:29,080 --> 00:56:32,840
going to make you opt in on certain things certain websites are not going to work because we're not

713
00:56:32,840 --> 00:56:39,420
going to load javascript you know random javascript from servers we don't know um but in return it's

714
00:56:39,420 --> 00:56:42,820
going to be much harder to hit you with a zero day because most zero days come in through things

715
00:56:42,820 --> 00:56:47,620
like sms messages with a link in them or an image that comes in through a chat that looks like an

716
00:56:47,620 --> 00:56:52,100
image but like buried in the image when your parser goes through it and tries to decrypt it or a pdf

717
00:56:52,100 --> 00:56:57,880
loads of zero days and pdfs actually just say what zero days it's basically it's because it's an

718
00:56:57,880 --> 00:57:02,240
exploit that's never been used in public never been seen so researchers will go through software

719
00:57:02,240 --> 00:57:09,140
they will find places where it's like hey if we are able to craft a pdf something that looks like

720
00:57:09,140 --> 00:57:14,720
a PDF, but buried in that PDF is a tiny bit of code that does one weird thing with the way that

721
00:57:14,720 --> 00:57:19,420
the parser, you know, the thing that looks at all the binary data of what your PDF is,

722
00:57:19,680 --> 00:57:23,400
if it makes that parser choke in a weird way, do something it's not supposed to do,

723
00:57:23,760 --> 00:57:27,960
and somehow upgrades our access into, we have total control of your phone,

724
00:57:28,460 --> 00:57:33,260
like that's the classic zero day that they're all looking for. And a lot of the recent ones,

725
00:57:33,260 --> 00:57:36,420
the things that Pegasus and its predecessors are...

726
00:57:36,420 --> 00:57:39,280
Pegasus is Israeli spy tech, basically.

727
00:57:39,620 --> 00:57:42,080
It's a program that loads of governments use.

728
00:57:42,400 --> 00:57:48,000
They've all been caught recently using a bunch of European governments that love to yell about how private and how much they care about privacy.

729
00:57:48,440 --> 00:57:55,120
The Italian government was just caught using this against journalists in the country who were reporting on the migration crisis from Africa.

730
00:57:56,200 --> 00:57:57,160
And this happens.

731
00:57:57,300 --> 00:58:02,280
And Italy's response to this was they got caught and they were like, yeah, but all these other European countries are doing it too.

732
00:58:02,280 --> 00:58:04,280
Do individuals use Pegasus?

733
00:58:05,180 --> 00:58:06,020
Good question.

734
00:58:06,020 --> 00:58:11,740
good question no there's no as far as i know there's no open source version of this this is

735
00:58:11,740 --> 00:58:15,860
like a but but you know hackers in general that's what they're doing that's what they're doing yeah

736
00:58:15,860 --> 00:58:22,000
and and basically pegasus is a you know a suite of semi-legitimate hacking company that just says

737
00:58:22,000 --> 00:58:28,140
we are a suite of tools for government and law enforcement and you're like no wow so i didn't

738
00:58:28,140 --> 00:58:32,460
know that i didn't know yeah it's new to me it's pretty scary stuff like it's it's scary and so

739
00:58:32,460 --> 00:58:35,960
you know you find these zero days and they're very hard to find and you know they're worth

740
00:58:35,960 --> 00:58:40,740
millions of dollars like the hackers out there in the dark you know markets that do this can make a

741
00:58:40,740 --> 00:58:45,920
lot of money doing this um but you know back to the earlier point like the nsa has many of these

742
00:58:45,920 --> 00:58:49,900
in their pocket and they're just waiting for if you're important enough they will pull one out and

743
00:58:49,900 --> 00:58:54,800
use it um and then they'll have access at the bottom level so like computers are not safe period

744
00:58:54,800 --> 00:58:58,480
like it's just it's a fact if the government wants to talk to you they will talk to you but

745
00:58:58,480 --> 00:59:00,200
What about Bitcoin wallets then?

746
00:59:01,060 --> 00:59:04,080
I mean, presumably the NSA could just take any of your Bitcoin.

747
00:59:04,220 --> 00:59:06,020
I mean, anything that's connected to a real computer, yeah.

748
00:59:07,200 --> 00:59:08,020
Now I know why people.

749
00:59:08,020 --> 00:59:11,800
This is why completely disconnected hardware wallets are pretty important, right?

750
00:59:12,320 --> 00:59:17,920
You want a signing device that has generated your private key never to touch the internet.

751
00:59:18,120 --> 00:59:19,220
Cold card is for those.

752
00:59:19,320 --> 00:59:21,500
Yeah, cold card is gold standard in my opinion.

753
00:59:22,240 --> 00:59:23,420
Those guys are so paranoid.

754
00:59:23,660 --> 00:59:26,520
And when you actually talk to them and you get into the weeds with them, you're like,

755
00:59:26,520 --> 00:59:31,320
oh my god you've thought of so many things here and you know i love some of the open source

756
00:59:31,320 --> 00:59:37,160
signing projects that are out there for the ethos behind them but i don't think that like their

757
00:59:37,160 --> 00:59:42,420
purpose fit for very large sums of money you know there's a scale here right like if you have a

758
00:59:42,420 --> 00:59:45,640
little bit of bitcoin it's totally fine to have it on a hot wallet on your phone like we should not

759
00:59:45,640 --> 00:59:50,180
be yelling at people that are getting into the space for having 50 bucks of bitcoin in a hot

760
00:59:50,180 --> 00:59:55,800
wallet like totally fine or a custodial service like just leave them alone for a bit it scales

761
00:59:55,800 --> 01:00:03,120
And the security needs to scale with that, you know, with how much you have to protect and how much you have to protect changes based on what your circumstances, right?

762
01:00:03,160 --> 01:00:06,760
Like 50 bucks of Bitcoin might be life changing to people in certain parts of the world.

763
01:00:07,080 --> 01:00:09,580
And, you know, they need to treat that accordingly.

764
01:00:09,580 --> 01:00:21,620
But I think if you want to think about like real security, you know, you need to have a kind of complicated setup that, you know, has a bunch of things that never touched the Internet and are very geographically dispersed.

765
01:00:22,020 --> 01:00:24,720
And even that is hard, right?

766
01:00:24,720 --> 01:00:27,660
Like, taking full-person responsibility is hard.

767
01:00:27,820 --> 01:00:28,480
That's just the fact.

768
01:00:28,700 --> 01:00:36,500
Those, the volume of people who should be in that situation is going up rapidly, I think.

769
01:00:36,580 --> 01:00:43,000
Because if you look back to the big pump of Bitcoin adoption, it was about 2017.

770
01:00:43,460 --> 01:00:46,720
That was a lot of individuals, wasn't it?

771
01:00:46,900 --> 01:00:46,980
Yeah.

772
01:00:47,080 --> 01:00:49,140
And if you track forward now.

773
01:00:49,280 --> 01:00:52,140
Yeah, we're 10, 12x from even the highs in there.

774
01:00:52,140 --> 01:01:01,160
It's easy to predict that there's a number of, quite a high percentage of those people should be pretty well off at the moment, if they just held.

775
01:01:01,400 --> 01:01:03,460
Yeah, and if they held.

776
01:01:03,580 --> 01:01:05,960
So I bet it's a smaller percentage than we think, but yeah.

777
01:01:05,980 --> 01:01:10,960
But still, I mean, the point is, in four years' time, there's going to be another big batch.

778
01:01:10,960 --> 01:01:17,780
And that batch from 2017, they're going to be like the 2013 lot who are basically boiling it on a yacht.

779
01:01:17,920 --> 01:01:19,180
Yeah, if they want to.

780
01:01:19,440 --> 01:01:20,200
Exactly, exactly.

781
01:01:20,200 --> 01:01:21,600
If they want to, they could, easily.

782
01:01:22,140 --> 01:01:40,080
But it's interesting. You're right. You're absolutely right. You know, if they can get on your phone, if they can get in your computer, then they would be able to have a screenshot of the time when they flashed up your private seed and you wrote it down or you were whispering it as you wrote it and the mic was picking it up or, you know.

783
01:01:40,080 --> 01:02:06,900
There's a million ways that this goes wrong. And again, it's like how much like you need to do the threat assessment, like how much of a target are you actually? How much, you know, what's more risky? Like somebody was saying the other day, like, I bet more Bitcoin has been lost from people trying to do the right thing by then people who just put their seed phrase into Google Doc and left it on a, you know, on Google servers for 10 years and then found it later. And they were like, oh, there's some Bitcoin here. Like, you know, people foot gun themselves pretty regularly.

784
01:02:06,900 --> 01:02:13,780
And so it's sort of, you know, do a threat assessment, really think through, like, am I going to be able to actually do the thing that I'm trying to set out to do?

785
01:02:13,860 --> 01:02:18,160
Am I technically capable and competent enough to deal with this and figure out what to do there?

786
01:02:18,220 --> 01:02:30,520
Like, if you're not technically capable and competent, maybe it does make more sense to go to someone like Anchor Watch and be like, hey, I'd like to, you know, do a collaborative custody thing with you guys and ensure this because, you know, it's meaningful to me.

787
01:02:30,600 --> 01:02:32,180
I'm happy to pay a little bit of money for it.

788
01:02:32,400 --> 01:02:33,920
You know, there's nothing wrong with custody.

789
01:02:33,920 --> 01:02:38,900
it's more a thing of if you're if you have custody forced upon you and you have no choice

790
01:02:38,900 --> 01:02:44,480
that's what where things go very bad um if you have the ability to move your things away from

791
01:02:44,480 --> 01:02:48,680
a custodian and you actually value their service and they provide a good service there's nothing

792
01:02:48,680 --> 01:02:54,220
wrong with that okay back to right we've just hit the end of signal by the sounds of it

793
01:02:54,220 --> 01:03:00,400
yes let's go back so evolution from signal yes so signal is great signal's major downfall is that

794
01:03:00,400 --> 01:03:05,040
it doesn't scale well. It is what's called a pairwise protocol, which means that if you are

795
01:03:05,040 --> 01:03:09,900
in a DM with somebody, me and you, it's great, works perfectly. If you start scaling up the group,

796
01:03:10,500 --> 01:03:14,820
we have to, each of us, for every message we send to the group, we have to individually encrypt it

797
01:03:14,820 --> 01:03:19,540
to every member of the group. And so that doesn't scale well. It scales linearly. You know, more

798
01:03:19,540 --> 01:03:24,440
group members means way more computation. And so when you get to very large groups,

799
01:03:24,440 --> 01:03:28,600
signal effectively just breaks. You know, it can't support groups of thousands or millions.

800
01:03:28,600 --> 01:03:34,760
and so that's always been something that this you know signal has done a lot of work and that's

801
01:03:34,760 --> 01:03:38,020
part of you know the stuff they do in their servers there's like all these bolt-on protocols

802
01:03:38,020 --> 01:03:43,440
that help make signal work better in groups but fundamentally everybody kind of knew that like we

803
01:03:43,440 --> 01:03:47,500
have to figure out something better for groups that scales better and so somewhere in 2017 2018

804
01:03:47,500 --> 01:03:52,040
a group of you know very very smart people who have been working in cryptography and and the

805
01:03:52,040 --> 01:03:55,540
signal protocol and stuff got together and went okay we're going to come up with a scheme that's

806
01:03:55,540 --> 01:04:01,680
going to do better for large groups. And what they came up with is what became MLS, message layer

807
01:04:01,680 --> 01:04:08,220
security. And it's, you know, kind of like TLS, you know, transport layer security. It takes a few

808
01:04:08,220 --> 01:04:14,520
of those ideas and applies a, you know, somewhat similar encryption scheme. So it's sort of a

809
01:04:14,520 --> 01:04:19,820
double ratchet-ish type scheme, but arranges a group in what's called a binary tree, which is a

810
01:04:19,820 --> 01:04:25,460
tree where, you know, it splits into two leaves from every node as it goes down. And so the group

811
01:04:25,460 --> 01:04:30,460
members in this ratchet tree across the bottom, you know, the very bottom layer. And then there's

812
01:04:30,460 --> 01:04:36,260
keys at all of the nodes above them. And the way it's set up means that when you change part of the

813
01:04:36,260 --> 01:04:41,640
group, you don't have to change the entire group structure. And when you have to encrypt things,

814
01:04:41,640 --> 01:04:47,120
you don't have to encrypt to everybody, you're encrypting to the keys along a path to the root,

815
01:04:47,320 --> 01:04:52,200
where everybody else can derive keys down from that root and get what they need. So we're back

816
01:04:52,200 --> 01:04:53,640
to these key derivation functions again.

817
01:04:53,760 --> 01:04:55,880
So it's a different way of applying key derivation functions

818
01:04:55,880 --> 01:04:58,740
and structuring a ROOP in a way that takes a problem.

819
01:04:59,300 --> 01:05:00,340
You know, in computer science talk,

820
01:05:00,360 --> 01:05:01,940
it takes it from a linear scaling problem

821
01:05:01,940 --> 01:05:04,940
to a log scaling problem, which makes a huge difference.

822
01:05:05,340 --> 01:05:06,660
And is this the pinnacle?

823
01:05:07,260 --> 01:05:08,080
Will it be the pinnacle?

824
01:05:08,300 --> 01:05:10,720
I think it's a incredibly good standard.

825
01:05:10,720 --> 01:05:14,100
I know it's being adopted by a lot of really big companies.

826
01:05:14,920 --> 01:05:17,360
And for good reason, it's going to help save them

827
01:05:17,360 --> 01:05:18,860
millions and millions of dollars in servers.

828
01:05:19,580 --> 01:05:21,820
It's also very well proved, you know,

829
01:05:21,820 --> 01:05:25,760
from an academic point of view, like the cryptography is very solid. It's also got this

830
01:05:25,760 --> 01:05:31,480
really cool property that you can upgrade groups and change groups over time in terms of like even

831
01:05:31,480 --> 01:05:44,980
stuff down to like what Cypher suite they using So you can completely change from a non resistant to a quantum resistant thing in a group in flight You don have to like disband the group and restart with the same people So it got a lot of cool properties that I think make

832
01:05:44,980 --> 01:05:49,840
it quite durable towards the future. It's also got an extension system that's sort of

833
01:05:49,840 --> 01:05:54,460
built into it that allows you to kind of sidecar data into the encryption scheme.

834
01:05:55,380 --> 01:05:59,460
And one other big thing that I think it gives you over Signal is it gives you sort of

835
01:05:59,460 --> 01:06:03,820
cryptographic, you know, a cryptographic guarantee of the entire group state.

836
01:06:04,020 --> 01:06:06,940
So, you know, cryptographically that like, these are the members of the group.

837
01:06:07,220 --> 01:06:09,480
These are the extensions that this group is using.

838
01:06:09,720 --> 01:06:11,920
This is the, you know, cipher suite we have to use.

839
01:06:12,040 --> 01:06:14,620
Like all of that is, you know, signed and hashed.

840
01:06:14,620 --> 01:06:18,520
And so, you know, okay, this is the state and we all, everybody in the group has agreed

841
01:06:18,520 --> 01:06:19,420
that this is the state.

842
01:06:20,220 --> 01:06:22,960
And so that gives you some interesting things that you can do with it.

843
01:06:22,960 --> 01:06:24,720
Why aren't Signal adopting that?

844
01:06:26,040 --> 01:06:27,720
It'll be interesting to see what they do.

845
01:06:27,720 --> 01:06:31,740
uh whether they decide to their problem right yeah it'll be interesting to see whether they

846
01:06:31,740 --> 01:06:36,880
uh stay with their own thing or if they start to move towards something like mls or buy a company

847
01:06:36,880 --> 01:06:41,440
or buy white noise i mean they don't even need to it's an open source thing like they've got

848
01:06:41,440 --> 01:06:45,720
plenty of smart people that could apply this and it's you know i think mls itself is very very

849
01:06:45,720 --> 01:06:50,840
complex um but once you have an implementation library that does the really base cryptographic

850
01:06:50,840 --> 01:06:56,020
math and like does have all the you know tests and things that cover off and you know does the

851
01:06:56,020 --> 01:06:58,600
test vectors that show that it's doing what it's supposed to be doing.

852
01:06:59,900 --> 01:07:02,660
You know, building an app on top of that becomes a lot more straightforward.

853
01:07:03,140 --> 01:07:07,540
I think the hard part for what we're doing is, you know, everybody up until now, you

854
01:07:07,540 --> 01:07:09,140
know, MLS has only been on internet standard.

855
01:07:09,660 --> 01:07:11,020
That's the other kind of cool thing about it.

856
01:07:11,120 --> 01:07:16,980
It was, you know, it's an IETF, so Internet Engineering Task Force, internet standard.

857
01:07:17,400 --> 01:07:18,760
You know, TLS is in that group.

858
01:07:18,980 --> 01:07:23,060
There's, you know, this is a worldwide group that comes together and is incredibly bureaucratic

859
01:07:23,060 --> 01:07:23,680
and slow.

860
01:07:23,920 --> 01:07:24,280
Oh, okay.

861
01:07:24,280 --> 01:07:27,980
But it's a bunch of engineers coming together and going, this is the thing, stamp.

862
01:07:28,100 --> 01:07:32,040
We've done lots and lots of work to prove that this is the thing, and we're all going to use it.

863
01:07:32,820 --> 01:07:37,000
Or if you want to use it, you know at least that the legwork has been done.

864
01:07:38,480 --> 01:07:38,840
Wow.

865
01:07:39,100 --> 01:07:42,760
And, you know, most of the things that are, you know, basic internet technologies are in this group.

866
01:07:42,980 --> 01:07:47,200
You know, TCPIP, TLS, like all that stuff is, you know, in IETF.

867
01:07:47,280 --> 01:07:53,180
And are there any apps that I would know, non-NOSTA apps, as in, that are using it already?

868
01:07:53,180 --> 01:07:56,180
Yeah, yeah. So there's Wire is a kind of well-known tool.

869
01:07:56,180 --> 01:07:58,880
We use that in our foundation.

870
01:07:59,140 --> 01:08:02,940
So the Wire guys are actually some of the original writers on the MLS protocol.

871
01:08:03,200 --> 01:08:07,640
I'd never heard of Wire until I joined the Free Cities Foundation, but our whole company runs on Wire.

872
01:08:07,740 --> 01:08:12,360
Yeah, yeah. There you go. So WebEx from Cisco uses it as the handshake for getting into webinars.

873
01:08:13,260 --> 01:08:20,920
So before you actually get the webinar stream, you have to do all this handshake stuff to actually enter the room to then get the stream.

874
01:08:20,920 --> 01:08:22,260
and that's all done with MLS.

875
01:08:22,700 --> 01:08:24,260
Now, everybody that's done this so far

876
01:08:24,260 --> 01:08:25,260
is a centralized service.

877
01:08:25,520 --> 01:08:26,600
They have a server in the middle

878
01:08:26,600 --> 01:08:27,560
that can guarantee identity

879
01:08:27,560 --> 01:08:28,820
and can guarantee transport, right?

880
01:08:29,140 --> 01:08:32,260
So MLS itself is transport and identity agnostic.

881
01:08:32,440 --> 01:08:33,820
It's like, that's for you to figure out.

882
01:08:34,120 --> 01:08:34,980
You know, we just talk about

883
01:08:34,980 --> 01:08:36,040
how you're going to structure the data

884
01:08:36,040 --> 01:08:36,960
and how you're going to do the encryption.

885
01:08:38,600 --> 01:08:40,820
And so that, when I started to talk to the guys

886
01:08:40,820 --> 01:08:42,140
that, you know, work on this stuff,

887
01:08:42,180 --> 01:08:42,580
they were like,

888
01:08:42,720 --> 01:08:43,640
we don't know of anybody

889
01:08:43,640 --> 01:08:45,020
that's done it decentralized yet.

890
01:08:45,180 --> 01:08:46,720
Like we've heard people talk about it,

891
01:08:46,760 --> 01:08:48,480
want to do it, but you know, good luck, have fun.

892
01:08:49,480 --> 01:08:50,740
They knew about NOSTA though.

893
01:08:50,920 --> 01:08:52,700
They roughly knew about Noster.

894
01:08:52,920 --> 01:08:55,900
You know, most people in that space are pretty skeptical of Bitcoin.

895
01:08:56,980 --> 01:08:57,180
Really?

896
01:08:57,480 --> 01:08:58,340
Yeah, it's weird.

897
01:08:58,560 --> 01:09:02,200
I've never understood how they can compartmentalize like that.

898
01:09:02,540 --> 01:09:06,300
But I think it's partially just because they didn't go down the legwork of understanding.

899
01:09:06,620 --> 01:09:08,500
Where are they skeptical?

900
01:09:08,500 --> 01:09:10,720
I mean, it's addressable probably.

901
01:09:10,940 --> 01:09:11,440
They just haven't done the work.

902
01:09:11,560 --> 01:09:11,720
Yeah.

903
01:09:12,000 --> 01:09:13,080
They haven't done the work and they don't care.

904
01:09:13,600 --> 01:09:16,540
And so it's one of these things where, you know, some of them get it.

905
01:09:16,660 --> 01:09:17,380
Some of them don't.

906
01:09:17,400 --> 01:09:19,940
I never push it because it's just like, I'm not going to spend my time doing that.

907
01:09:20,920 --> 01:09:26,260
But they were supportive in the sense of this would be cool to see, ask questions, help out.

908
01:09:27,600 --> 01:09:35,520
And so my job was basically how do I apply this into Noster and do it in a way that doesn't leak metadata?

909
01:09:36,700 --> 01:09:48,980
How do we put this encryption scheme in, use public relays, send everything across public-facing stuff in a way that if an observer is looking at all the traffic, they don't really know much?

910
01:09:48,980 --> 01:09:55,100
you know uh and that proved to be you know pretty tricky just before you go on i just had a thought

911
01:09:55,100 --> 01:10:03,540
this is a complete tangent but does this system have any benefit to privatizing bitcoin transactions

912
01:10:03,540 --> 01:10:08,700
and things like that yeah well privatizing bitcoin transactions maybe a little less the way you

913
01:10:08,700 --> 01:10:14,420
described it then was you're sending things over an open network and no one really can derive much

914
01:10:14,420 --> 01:10:18,640
information from it which is the essence of bitcoin privacy in a way yeah well the problem

915
01:10:18,640 --> 01:10:22,800
Bitcoin is that those transactions need to be public, right? Like it needs to be a public ledger

916
01:10:22,800 --> 01:10:29,200
so that everyone can verify. Um, what about a second layer then? Absolutely. Yeah. So this has

917
01:10:29,200 --> 01:10:34,780
all sorts of use cases in, you know, things like, uh, multi-sig coordination, great use case, you

918
01:10:34,780 --> 01:10:37,960
know, like you've got a wallet, you want to be able to coordinate with other people to do all

919
01:10:37,960 --> 01:10:43,660
the signing. Um, you need some way to, to do that in a way that is secure. Um, so that's a great

920
01:10:43,660 --> 01:10:47,920
use case for it. I'm sure there are, you know, second layer things that you could do, you know,

921
01:10:47,920 --> 01:11:01,840
But I think there's probably slightly better ways to do it in a lot of cases just because there's a lot of overhead with setting up a group, with keeping the group state up to date, making sure that all the devices are clients that need to be in that group.

922
01:11:02,400 --> 01:11:09,660
When you join an MLS group from your phone and your computer, and this is the same with Signal, by the way, you're actually two different members in that group.

923
01:11:10,400 --> 01:11:15,760
It's just that the UI sees that, hey, this is actually the same person and kind of collapses it to one person in the view.

924
01:11:15,760 --> 01:11:22,800
but you are cryptographically two different entities in that group um so i think you know

925
01:11:22,800 --> 01:11:27,300
for certain bitcoin applications especially closer to the humans it makes a tons of sense

926
01:11:27,300 --> 01:11:31,920
you know i think when you start to get down into the weeds of like how would you know lightning

927
01:11:31,920 --> 01:11:35,640
privacy work or there's other solutions that are going to make a little more sense fair enough okay

928
01:11:35,640 --> 01:11:41,640
so you so there are no decentralized versions or they weren't there weren't there's a bunch of

929
01:11:41,640 --> 01:11:45,500
people working on them you know i know that the mastodon folks are are dying to have this their

930
01:11:45,500 --> 01:11:51,120
proposal leaks metadata all over the place um the blue sky folks would love to have this too but

931
01:11:51,120 --> 01:11:56,780
their pds system doesn't really allow for it um because it's a quite a centralized identity server

932
01:11:56,780 --> 01:12:03,340
thing um so i think i think we're like the first properly decentralized version of this that does

933
01:12:03,340 --> 01:12:08,000
a really good job hiding privacy um i haven't heard of any others that you know have made me go

934
01:12:08,000 --> 01:12:12,780
oh we actually missed something there and in order to interact with it you just need a noster

935
01:12:12,780 --> 01:12:18,840
you just need a noster publicly that's it right um and what i what i noticed was i haven't actually

936
01:12:18,840 --> 01:12:23,380
i've got it but i haven't actually sent a dummy message yet that's okay but what i noticed was

937
01:12:23,380 --> 01:12:30,640
what it didn't do was pull in my my list of known people so from noster and obviously there's a

938
01:12:30,640 --> 01:12:35,220
privacy that there's a privacy feature or am i no no that's it's probably just an early version

939
01:12:35,220 --> 01:12:39,820
yeah yeah like we're trying to get this out as early as possible to get things and we've gotten

940
01:12:39,820 --> 01:12:44,520
so much great feedback and to be honest the app barely works for certain people and the longer

941
01:12:44,520 --> 01:12:48,000
you've been on nostr and you sign into white noise the more problems you're going to find

942
01:12:48,000 --> 01:12:52,940
because you've got all these it's actually down to relay and like how relays interact with things

943
01:12:52,940 --> 01:12:56,420
and when you're trying to be quite private with with certain parts of data it makes it harder for

944
01:12:56,420 --> 01:13:01,140
relays to understand what's going on and so it makes it harder for things to overlap if you've

945
01:13:01,140 --> 01:13:05,340
got the kind of outbox model set up where you've defined all these different relays that you want

946
01:13:05,340 --> 01:13:10,260
to use for different purposes so that's generally the thing that makes this really hard you know if

947
01:13:10,260 --> 01:13:14,700
you sign up with a brand new account on there like if you just create account it's almost universally

948
01:13:14,700 --> 01:13:19,380
going to work and you know once you get into the edge cases of which there are a lot in our early

949
01:13:19,380 --> 01:13:22,920
you know cohort of users which are all the you know people that have been on NOSTA a while

950
01:13:22,920 --> 01:13:26,960
that's what we're kind of working through now I think this is true right not a lot of people

951
01:13:26,960 --> 01:13:32,060
realize this but on nostr currently if you're using primal or amethyst or whatever when you

952
01:13:32,060 --> 01:13:37,680
send a direct a dm yeah it's public right pretty much i mean yeah the content of the message is

953
01:13:37,680 --> 01:13:42,580
encrypted with a very lightweight encryption scheme the all the metadata of who's talking

954
01:13:42,580 --> 01:13:47,060
to who and when that's all public um and there's actually bots that go around just reporting on

955
01:13:47,060 --> 01:13:52,100
this the dm reporter bot it's very funny well what i what on mine i use amethyst but when someone

956
01:13:52,100 --> 01:13:57,800
sends me a dm it becomes a notification yeah as in not a just there's a dm notification and

957
01:13:57,800 --> 01:14:02,800
that's the giveaway for me that's the giveaway right yep and and there is another scheme like

958
01:14:02,800 --> 01:14:05,960
you know i think it was like a year and a half ago or almost two years ago at this point um

959
01:14:05,960 --> 01:14:12,080
you know the earliest version of the you know dm scheme nipo4 which is what primal and damas still

960
01:14:12,080 --> 01:14:16,960
use um that's like always been known to be bad right like we've all known that from the beginning

961
01:14:16,960 --> 01:14:21,040
and ben arc was the one who wrote that originally and like he never meant that to be an actual

962
01:14:21,040 --> 01:14:25,520
encryption scheme. It was like, let's see if we can get semi-private messaging on Nostr.

963
01:14:25,860 --> 01:14:30,520
You know, hacked it up in a few hours and like, okay, this is good. And there's like 30 people

964
01:14:30,520 --> 01:14:35,880
using Nostr, right? Like there's no one on the network. And it just somehow hung around for a

965
01:14:35,880 --> 01:14:40,440
long time. And so there, you know, whatever, a year and a half ago, a bunch of devs got together

966
01:14:40,440 --> 01:14:45,320
and did some work to come up with something that was a little better, you know, in terms of it's a

967
01:14:45,320 --> 01:14:50,480
better encryption scheme for the actual content, but it also wraps the messages in another message

968
01:14:50,480 --> 01:14:52,980
so that it hides all the metadata.

969
01:14:52,980 --> 01:14:55,460
So what you see on the network is some random pub key

970
01:14:55,460 --> 01:14:58,260
that we've never seen before sent a message to

971
01:14:58,260 --> 01:14:59,980
that mentions someone,

972
01:15:00,000 --> 01:15:01,060
We don't know what that message is.

973
01:15:01,100 --> 01:15:01,960
It could be anything inside.

974
01:15:02,420 --> 01:15:08,160
And the person that, you know, receives the message knows how to unwrap that message and then get to the actual encrypted option.

975
01:15:08,160 --> 01:15:15,940
So the plan going forward would probably be that the YNOS protocol would just get incorporated in other things.

976
01:15:16,080 --> 01:15:16,660
Yes, exactly.

977
01:15:16,800 --> 01:15:19,060
And that's actually what's taken so long in the project.

978
01:15:19,280 --> 01:15:22,760
You know, like first I had to figure out how to even apply MLS into NOSTER.

979
01:15:22,940 --> 01:15:28,440
And like once I was fairly sure that was possible, then I actually started with instead of writing white noise,

980
01:15:28,440 --> 01:15:33,620
I was concurrently writing all the libraries that I would want other developers to use to do this.

981
01:15:33,920 --> 01:15:36,560
And so I wrote the spec of like, this is how you're supposed to do the thing.

982
01:15:36,920 --> 01:15:39,380
And then I wrote the libraries that implement that spec.

983
01:15:39,840 --> 01:15:46,200
And at the same time, I was sort of writing the proof of concept of what the white noise messenger was, sort of to test the libraries I was building.

984
01:15:47,700 --> 01:15:53,620
And as we've gone on a little bit, we've now got a pretty solid basic library that does the thing correctly.

985
01:15:54,620 --> 01:15:59,100
It's written in Rust, so we'll be able to bind out to other languages and give this to other developers.

986
01:15:59,700 --> 01:16:02,080
And white noise has come a long, long way in the process.

987
01:16:02,300 --> 01:16:04,780
So I want this to be fully interoperable.

988
01:16:05,020 --> 01:16:09,920
I know for a fact that, like, you know, Vitor with Amethyst, Will at Damas, Million with Primal,

989
01:16:10,000 --> 01:16:13,100
they're all just like, tell us when it's ready and we're going to implement it.

990
01:16:13,160 --> 01:16:15,220
And, like, we'll do this right across the board.

991
01:16:15,220 --> 01:16:18,280
And you're funded by OpenSats, is it?

992
01:16:18,380 --> 01:16:20,040
Or it can't be a business for you?

993
01:16:20,100 --> 01:16:20,980
You're just giving it away freely?

994
01:16:20,980 --> 01:16:22,560
No, this is definitely not a business for me.

995
01:16:22,560 --> 01:16:35,240
And it was actually funny, like when I started to think about the project, it was sort of, do I try to raise VC funds for this or do I stick to the open source grants model and, you know, try to do this as like a public benefit type thing?

996
01:16:35,920 --> 01:16:39,460
And, you know, I would say it was like a hard decision.

997
01:16:39,560 --> 01:16:40,460
It actually wasn't.

998
01:16:40,740 --> 01:16:42,960
You know, I'd been on an OpenSats grant for a little while at that point.

999
01:16:43,380 --> 01:16:50,920
And to me, this is a project that needs to be fully open, fully, you know, untethered to something where there's a financial outcome.

1000
01:16:50,920 --> 01:16:52,040
Maybe described by OpenSats.

1001
01:16:52,040 --> 01:16:56,740
I assume that they weren't large grants, though, not enough to live off and stuff.

1002
01:16:56,980 --> 01:16:59,400
No, well, I think OpenSats has done a really good job.

1003
01:16:59,500 --> 01:17:01,120
So let me quickly give the definition.

1004
01:17:01,300 --> 01:17:08,480
Like OpenSats is just a 501c3 nonprofit that takes in grants and donations from, you know, big foundations, from donors.

1005
01:17:08,960 --> 01:17:11,880
You know, they've gotten a lot of kind of their seed capital from Jack Dorsey.

1006
01:17:12,780 --> 01:17:15,760
And they fund Bitcoin, Lightning and Nostra projects.

1007
01:17:15,760 --> 01:17:23,480
and specifically they fund developers with grants that generally run for you know six months to a

1008
01:17:23,480 --> 01:17:29,960
year they also have a long-term support program which does you know two-year grants um i was one

1009
01:17:29,960 --> 01:17:35,100
of the very first grantees to get an open sats grant you know at the beginning of nostr um they'd

1010
01:17:35,100 --> 01:17:38,480
been doing a little bit of bitcoin grants and they started doing nostr grants and you know i was

1011
01:17:38,480 --> 01:17:42,360
funded for a year with that and then i got a renewal and got another year um that's actually

1012
01:17:42,360 --> 01:17:47,480
just ended. So now I'm no longer funded by OpenSats. But, you know, there's been a bunch

1013
01:17:47,480 --> 01:17:51,480
of other organizations in the middle as well, like HRF gives Bitcoin grants for developers.

1014
01:17:52,180 --> 01:17:54,640
I'm very lucky to be working with them and gotten grants from them.

1015
01:17:54,640 --> 01:17:55,580
It makes perfect sense though.

1016
01:17:55,580 --> 01:17:58,980
Yeah. It makes total sense. And so there's an ecosystem of places that

1017
01:17:58,980 --> 01:18:03,560
are granting money to developers and projects for things. And I think it's cool because,

1018
01:18:03,660 --> 01:18:09,800
you know, OpenSats focuses more on sort of very wide, they're happy to give developers who might

1019
01:18:09,800 --> 01:18:13,360
not even have necessarily a clear project in mind, but have shown proof of work and built

1020
01:18:13,360 --> 01:18:16,660
interesting things that are just experimenting with a lot of stuff. And I think that's really

1021
01:18:16,660 --> 01:18:22,300
important on one side. HRF tends to fund stuff that's like a bit more, this is a known project

1022
01:18:22,300 --> 01:18:26,680
and it's a project that fits into our remit of, you know, trying to make more freedom in the world.

1023
01:18:27,240 --> 01:18:31,880
And so HRF has been a really great partner to White Noise and me and everything. So they've

1024
01:18:31,880 --> 01:18:38,000
been amazing. But yeah, like I think where this is headed is I'm probably going towards, you know,

1025
01:18:38,000 --> 01:18:43,140
creating a 501c3 going you know part of my job will be making sure that we always have money to

1026
01:18:43,140 --> 01:18:48,720
keep this project going um so you're going to do this full-time i mean i've been doing this full-time

1027
01:18:48,720 --> 01:18:53,080
for over a year now and i intend to continue working on this for as long as humanly possible

1028
01:18:53,080 --> 01:18:58,980
um you mentioned jack yeah tell me about bit chat which i haven't used yet

1029
01:18:58,980 --> 01:19:05,960
it's so much fun really it's so much fun yeah i mean look i think you don't think it's a serious

1030
01:19:05,960 --> 01:19:11,720
contender though well it's just a totally different thing right like it's um i think you know taking

1031
01:19:11,720 --> 01:19:16,900
one step back i think jack is a such a unique human um like how many people can you pinpoint

1032
01:19:16,900 --> 01:19:22,420
in the world that have made billions of dollars and are still hacking on stuff in their spare time

1033
01:19:22,420 --> 01:19:28,800
and are still looking at how do we increase freedom and agency in the world like there's not many

1034
01:19:28,800 --> 01:19:34,920
right you know billionaires usually get stuck into other things um so the fact that you know jack has

1035
01:19:34,920 --> 01:19:40,680
a public company that has two incredibly you know profitable business lines that he has to deal with

1036
01:19:40,680 --> 01:19:45,940
you know wall street on a regular basis but finds time to hack on stuff you know vibe code apps in

1037
01:19:45,940 --> 01:19:51,080
his spare time is kind of nuts talking of before we go oops before we go out are you vibe code in

1038
01:19:51,080 --> 01:19:57,160
this or are you no no the thing are you coding it a lot so there's parts of you know this where llms

1039
01:19:57,160 --> 01:20:01,380
can be really helpful you know like writing documentation writing some tests like those

1040
01:20:01,380 --> 01:20:06,140
are great anytime you get to code that is fundamentally novel that's new they where

1041
01:20:06,140 --> 01:20:10,660
you're combining things that you know an llm has never seen the number of hallucinations goes through

1042
01:20:10,660 --> 01:20:14,660
the roof and it's like totally useless you also just don't want to be vibe coding stuff that

1043
01:20:14,660 --> 01:20:18,620
really matters you know like no one should ever be vibe coding on bitcoin core like that's just

1044
01:20:18,620 --> 01:20:24,340
fundamental fact but what about you as a are you a coder yeah so you're sitting at the you're you've

1045
01:20:24,340 --> 01:20:29,640
got your hoodie on i use lms constantly for all sorts of stuff from everything from just

1046
01:20:29,640 --> 01:20:34,360
hey can you give me you know i can't remember what the api definition of this method is can you give

1047
01:20:34,360 --> 01:20:39,900
me a quick like summary of it or tell me more about how this piece of the protocol works and

1048
01:20:39,900 --> 01:20:45,060
we like talk back and forth about it it's a great pair programmer and as a you know experienced

1049
01:20:45,060 --> 01:20:49,980
engineer you really benefit from having someone to bounce ideas off of and having someone who's

1050
01:20:49,980 --> 01:20:53,880
literally read all the documentation for every software ever you know created in the world and

1051
01:20:53,880 --> 01:20:59,420
can immediately give you a correct answer on basic information without without prying 2d i take it

1052
01:20:59,420 --> 01:21:05,900
you could go and get a very well-paid job somewhere probably yeah just so i'm just formulating a

1053
01:21:05,900 --> 01:21:09,220
picture we've only just met and i'm yeah please very pleased to meet you but i'm working this

1054
01:21:09,220 --> 01:21:15,740
stuff out right bit chat okay bit chat is awesome bit chat is super super fun um i love the idea of

1055
01:21:15,740 --> 01:21:20,900
kind of off-grid stuff like that you know using mesh networks using wi-fi using radio you know

1056
01:21:20,900 --> 01:21:26,700
using satellite stuff i think um where we have gotten with the internet over the last 25 years

1057
01:21:26,700 --> 01:21:33,240
is terrifying. It's become completely centralized and totally owned by five companies and governments.

1058
01:21:33,540 --> 01:21:38,700
And that's not somewhere I would like to be. You know, like, to be honest, I would almost rather

1059
01:21:38,700 --> 01:21:43,540
it be wholly owned by governments rather than, you know, five big companies. Governments are at

1060
01:21:43,540 --> 01:21:50,440
least not as useful. They aren't as competent as these companies are. You know, like, Meta is full

1061
01:21:50,440 --> 01:21:56,600
of incredibly intelligent people. And I worry about their motivations. And so I think, you know,

1062
01:21:56,700 --> 01:22:01,460
Anything that helps us go in the direction away from this, you know, pure centralization,

1063
01:22:02,000 --> 01:22:06,900
like the internet that most people know right now is actually just five companies and their

1064
01:22:06,900 --> 01:22:07,320
platforms.

1065
01:22:07,360 --> 01:22:13,060
Like it is a platformized internet rather than the open decentralized protocol based

1066
01:22:13,060 --> 01:22:14,100
thing that we started with.

1067
01:22:14,260 --> 01:22:17,240
And I want to do everything I can to make us go back in that direction.

1068
01:22:17,780 --> 01:22:19,240
I think that's where we get freedom.

1069
01:22:19,340 --> 01:22:20,500
That's where we get interesting ideas.

1070
01:22:20,500 --> 01:22:25,380
That's where people actually benefit from these tech rather than, you know, being sucked

1071
01:22:25,380 --> 01:22:26,440
off of by the parasites.

1072
01:22:26,700 --> 01:22:38,160
And Jack's BitChat is essentially a mechanism for transmitting data directly between devices over Bluetooth, whatever it can get its hands on.

1073
01:22:38,160 --> 01:22:40,480
Well, there's nothing super novel about this, right?

1074
01:22:40,520 --> 01:22:47,160
Like we've known about mesh networks for a long time and Bluetooth is, I think, only recently becoming viable there with like Bluetooth low energy.

1075
01:22:48,040 --> 01:22:49,480
You know, it's used as low energy.

1076
01:22:49,700 --> 01:22:52,860
It can do a little bit, you know, decent range on it.

1077
01:22:52,860 --> 01:22:59,180
um so for certain use cases like you're in a cities or you're in a panel like in a talk at

1078
01:22:59,180 --> 01:23:04,260
bitcoin honey badger or you know or baltic honey badger we were in panels it'd be like okay open up

1079
01:23:04,260 --> 01:23:07,960
bit chat and send us questions and like you're basically just getting the questions from the

1080
01:23:07,960 --> 01:23:13,280
room right it's ephemeral it disappears immediately um you know there's no identity attached to it

1081
01:23:13,280 --> 01:23:18,960
so i think there's some fun stuff there but it's worth mentioning that you could transmit a message

1082
01:23:18,960 --> 01:23:24,660
from that room to australia providing you have enough it could make yeah and it has to be a it

1083
01:23:24,660 --> 01:23:29,760
has to be within sort of striking distance i mean 100 feet 100 meters is that right yeah you're not

1084
01:23:29,760 --> 01:23:33,860
going to get you're not going to get it across the channel or across the across the no you're

1085
01:23:33,860 --> 01:23:38,380
going to have to hook into something that can do longer distance at some point um so it's it can be

1086
01:23:38,380 --> 01:23:44,180
really good for very you know small geographic locations um i mean you could easily relay stuff

1087
01:23:44,180 --> 01:23:49,880
across the city in a couple of minutes but like i was i was a journalist in london for

1088
01:23:49,880 --> 01:23:56,620
almost eight years and often we used to cover demonstrations and during demonstrations the

1089
01:23:56,620 --> 01:24:02,340
police used to shut off the mobile phone signal this was back in the 90s and 2000s like it was

1090
01:24:02,340 --> 01:24:08,240
happening back then so that would really be a great exactly exactly and i think the cool thing

1091
01:24:08,240 --> 01:24:31,394
about bitchat is that it fun um people love it it like spreading really quickly and it making us figure this whole like mesh thing out because you know talking to like Kale who the one that does the BitChat Android app You know he like it so interesting because we have no idea how these mesh things are you know prioritizing different you know mechanisms

1092
01:24:31,394 --> 01:24:34,054
Like there's no optimization here at all.

1093
01:24:34,094 --> 01:24:38,954
It's literally like try to connect to anybody you see and just send messages, you know, and it fails in all these weird ways.

1094
01:24:38,954 --> 01:24:42,754
And so I think it's we need to get better at understanding, like, how do we do mesh well?

1095
01:24:43,214 --> 01:24:46,574
We know how we do, you know, the internet well when you've got a wire.

1096
01:24:46,834 --> 01:24:50,234
But how do you do mesh over Bluetooth or radio or Wi-Fi?

1097
01:24:50,394 --> 01:24:52,474
It must be great at festivals as well.

1098
01:24:52,614 --> 01:24:53,954
It could also be amazing at festivals.

1099
01:24:54,174 --> 01:24:55,954
Half a million people all in one place.

1100
01:24:56,054 --> 01:24:56,194
Yeah.

1101
01:24:56,334 --> 01:24:59,874
And the mobile phone signal can't handle all that data that you can do.

1102
01:25:00,054 --> 01:25:00,374
Exactly.

1103
01:25:00,374 --> 01:25:05,954
So I hope that we are going towards a future where we have way less centralization on the internet.

1104
01:25:06,374 --> 01:25:11,374
And we've got little pockets all over the place that do different stuff at different layers of the stack.

1105
01:25:11,454 --> 01:25:12,634
And like, yeah, it's going to connect to the internet.

1106
01:25:12,754 --> 01:25:20,054
some point beam you across the ocean but you know do you need your you know town chat or whatever to

1107
01:25:20,054 --> 01:25:23,914
actually hit the internet no it doesn't it can literally stay locally and it could be on a server

1108
01:25:23,914 --> 01:25:28,754
in someone's basement and it never touches the internet at all and obviously that would have a

1109
01:25:28,754 --> 01:25:35,734
big impact on bitcoin payments as well bitcoin payments on just privacy in general the less you

1110
01:25:35,734 --> 01:25:41,654
can feed data and traffic and information to facebook the better but i mean for example i

1111
01:25:41,654 --> 01:25:49,474
i've been to um el salvador a few times and i've realized very succinctly that if you don't got a

1112
01:25:49,474 --> 01:25:54,474
phone signal it's very difficult to make a bitcoin transaction and you get used to around here not

1113
01:25:54,474 --> 01:25:58,754
having that problem but when you go something like that yeah the papusa lady wants some bitcoin

1114
01:25:58,754 --> 01:26:03,374
off you and you can't send it presumably these mesh networks would you would be able to use the

1115
01:26:03,374 --> 01:26:08,434
mesh networks and then would with the and then settle later or you maybe use lightning and settle

1116
01:26:08,434 --> 01:26:12,114
later or something yeah well i think i don't even think it would be a settled later type of thing

1117
01:26:12,114 --> 01:26:17,854
because um you know those technologies bitcoin specifically and lightning you know do require

1118
01:26:17,854 --> 01:26:23,434
you know being connected so to give you any sort of feedback like it worked so it would really work

1119
01:26:23,434 --> 01:26:28,894
wouldn't it i could send lightning to you and me over a mesh network we would never have to touch

1120
01:26:28,894 --> 01:26:33,114
the internet but you wouldn't be able to validate anything about it that it worked or that it

1121
01:26:33,114 --> 01:26:38,174
actually happened at all and so i think you know for your example there cash is a way better use

1122
01:26:38,174 --> 01:26:43,494
case because cashew is a bearer instrument so like you can give a token to someone and yes they

1123
01:26:43,494 --> 01:26:47,474
don't know there's there's a window there where you could double spend on it because if you ran

1124
01:26:47,474 --> 01:26:52,014
back to the internet and spent that same token before the other person could you know get internet

1125
01:26:52,014 --> 01:26:56,474
and exchange the token have we got time to go into cashew we've been going nearly an hour and a half

1126
01:26:56,474 --> 01:27:01,014
i've got all the time you want right yeah effrat's coming in at half an hour for a friend nice yeah

1127
01:27:01,014 --> 01:27:06,914
yeah right good friend yeah she's great um because i use cashew mini bits yeah i use mini bits once

1128
01:27:06,914 --> 01:27:12,674
again because people that know a lot more than me said this is amazing use this right yeah um

1129
01:27:12,674 --> 01:27:18,654
i didn't realize it was a bearer instrument how does that work so cashew is basically like a

1130
01:27:18,654 --> 01:27:24,894
wildcat bank right like the idea behind the cashew mint is that they're taking in some asset in this

1131
01:27:24,894 --> 01:27:31,854
case bitcoin they are holding that and they're giving you an iou in return effectively um and

1132
01:27:31,854 --> 01:27:36,754
that iou through cryptography is blinded they don't know how much they don't know who they're

1133
01:27:36,754 --> 01:27:42,494
giving it like the bank is totally blind, which is a really great thing, right? They can't do any,

1134
01:27:42,754 --> 01:27:46,594
you know, anything funny with you specifically, because they don't know anything about who they're

1135
01:27:46,594 --> 01:27:50,674
dealing with. They're just like, hey, this is Bitcoin. Yes, this is real. Okay, here's your

1136
01:27:50,674 --> 01:27:56,334
thing. When I see this thing come back to me, I'm going to, I'm going to honor that, right? So that

1137
01:27:56,334 --> 01:27:59,474
when this IOU comes back to me, I'm going to give the Bitcoin out to whoever has it. It doesn't

1138
01:27:59,474 --> 01:28:05,414
matter who has it. And that's, that's kind of what eCash is, right? Like you, it's a blinded

1139
01:28:05,414 --> 01:28:10,594
transaction where it doesn't matter how many times or how many different people that that e-cash token

1140
01:28:10,594 --> 01:28:16,394
that iou moves around in the system once it comes back to be traded from the mint that created it

1141
01:28:16,394 --> 01:28:22,134
um if it's a you know a mint that's acting in good faith they return the bitcoin and burn their iou

1142
01:28:22,134 --> 01:28:28,554
so that's the point you need to be online at the when it comes in and out yes the trade point that

1143
01:28:28,554 --> 01:28:33,174
that point at which you want to melt the e-cash or you know trade it in for you know the actual

1144
01:28:33,174 --> 01:28:38,014
asset underlying it that's the point that you need to connect to so that means from mini bits i could

1145
01:28:38,014 --> 01:28:44,854
spend bitcoin over a mesh network oh yeah and and they've got they've got all the stuff already i

1146
01:28:44,854 --> 01:28:49,934
mean cali was playing with this and putting cashier tokens in uh in bitchat and it works fine you know

1147
01:28:49,934 --> 01:28:55,134
like it pops up in your thing and like yes there's a you know there's a place in there where yeah it's

1148
01:28:55,134 --> 01:29:02,034
custodial this is that that's absolutely incredibly amazing and solves one of the biggest problems i

1149
01:29:02,034 --> 01:29:07,314
i think personally yeah is connection with the internet that's yep and again that's like there

1150
01:29:07,314 --> 01:29:12,454
is a there is a there is a risk in the middle yeah i mean cash is a little bit better in this case

1151
01:29:12,454 --> 01:29:16,854
right because i guess the mint that you're trusting is the actual government so like

1152
01:29:16,854 --> 01:29:21,454
you are more comfortable trusting them than some person on the internet that just happened to set

1153
01:29:21,454 --> 01:29:27,374
up a mint so trust in your mint or spreading your trust across many mints is really important in a

1154
01:29:27,374 --> 01:29:31,214
system like cashew but it gives you other properties that are really really beneficial

1155
01:29:31,214 --> 01:29:36,794
and so when people yell about like oh but it's custodial it's like yes fine but like everything

1156
01:29:36,794 --> 01:29:40,374
is trade-offs right like you are always going to trade something for something else and so

1157
01:29:40,374 --> 01:29:44,254
cashew makes the trade of yes this is custodial but like look at all this cool shit we get

1158
01:29:44,254 --> 01:29:51,614
yeah is this an example of that the meme that we are really early is that going to become some kind

1159
01:29:51,614 --> 01:29:57,554
of standard because i hope well why wouldn't it i mean like this is the dream is being able to

1160
01:29:57,554 --> 01:30:03,914
You cash you on a card and you have NFC tap everywhere and you pay just like you pay with your MasterCard or Visa.

1161
01:30:04,094 --> 01:30:06,454
Like the UX of Visa is incredible.

1162
01:30:06,794 --> 01:30:09,114
You just show up and go ding and you walk away.

1163
01:30:09,494 --> 01:30:09,874
That's it.

1164
01:30:09,874 --> 01:30:17,754
And so if we can get, you know, experiences like that, you know, everybody loves to like go and spend lightning off their own node.

1165
01:30:17,754 --> 01:30:23,494
But like how nice is it to sit there and wait for 12 seconds while Zeus spins and you're like, please work, please work.

1166
01:30:23,774 --> 01:30:25,074
And, you know, like it sucks.

1167
01:30:25,154 --> 01:30:25,634
It's terrible.

1168
01:30:25,634 --> 01:30:30,994
So we need to get to a place where our UX is as good or better than Visa before people are actually going to think this tech is useful.

1169
01:30:31,214 --> 01:30:35,254
But also with Cashew, there's a great privacy angle to that as well.

1170
01:30:35,254 --> 01:30:36,114
I mean, incredible privacy, yeah.

1171
01:30:36,194 --> 01:30:38,074
Because the Mint has no idea who you are.

1172
01:30:38,414 --> 01:30:43,894
And the token that went out just moves around however long in the community before it gets exchanged.

1173
01:30:44,194 --> 01:30:47,454
So, I mean, Cashew, I can see why they called it Cashew now.

1174
01:30:47,494 --> 01:30:49,774
Because it really is very similar to cash.

1175
01:30:49,774 --> 01:30:55,674
It's a bearer asset that you give to someone and you're private, providing Google's not looking at you through a telescope or whatever.

1176
01:30:56,314 --> 01:30:56,854
No one knows.

1177
01:30:56,934 --> 01:30:58,094
This is old tech too, right?

1178
01:30:58,154 --> 01:31:01,354
Like David Chaum came up with this e-cash concept in 82.

1179
01:31:01,614 --> 01:31:02,434
Yeah, Chaumian e-cash.

1180
01:31:02,814 --> 01:31:08,394
And the thing that stopped him was he didn't have a way to do it without having to go through the centralized financial system.

1181
01:31:08,854 --> 01:31:18,394
You know, he almost had it where he had deals with Visa and MasterCard and big banks to use this as kind of the cash that people would use in the digital world.

1182
01:31:18,394 --> 01:31:23,454
and it just kind of fell apart at the last second because the centralized players didn't want it or

1183
01:31:23,454 --> 01:31:30,314
didn't what year was that 2000 well 82 is the first paper 1982 and then you know in the 90s

1184
01:31:30,314 --> 01:31:34,574
he had a company that was trying to do this and it you know just didn't work what was he doing in

1185
01:31:34,574 --> 01:31:41,694
2009 i mean david chom's an interesting guy uh uh yeah we can go on but go and look into it i mean

1186
01:31:41,694 --> 01:31:46,754
it's just he's i think david chom was an incredibly intelligent person obviously he came up with

1187
01:31:46,754 --> 01:31:50,754
cash like the whole idea of the scheme is is pretty amazing he came up with this kind of

1188
01:31:50,754 --> 01:31:56,554
blinding scheme for you know and the math behind it i think on the business side i think he overplayed

1189
01:31:56,554 --> 01:32:01,814
his hand in a few cases where you know he tried to get too much and deals fell apart because of that

1190
01:32:01,814 --> 01:32:08,494
um and i think in bitcoin i don't know 100 like what his thoughts on but i think it was like it

1191
01:32:08,494 --> 01:32:13,254
took him a little while to come around to it in the sense that like maybe a sort of not made here

1192
01:32:13,254 --> 01:32:14,734
sort of stance or opinion.

1193
01:32:14,734 --> 01:32:15,294
Derangement syndrome.

1194
01:32:15,374 --> 01:32:16,054
Maybe, I don't know.

1195
01:32:16,054 --> 01:32:16,754
Bitcoin derangement system.

1196
01:32:16,754 --> 01:32:17,814
I think he's pretty into it now,

1197
01:32:17,894 --> 01:32:19,074
but like, you know, I don't know.

1198
01:32:19,214 --> 01:32:20,914
I think it's been used much better

1199
01:32:20,914 --> 01:32:23,194
with the recent Cashew and Fetty

1200
01:32:23,194 --> 01:32:23,794
and things like that

1201
01:32:23,794 --> 01:32:24,954
that are coming back to eCash,

1202
01:32:25,074 --> 01:32:26,554
but through a Bitcoin lens.

1203
01:32:26,934 --> 01:32:28,014
I think we get a lot more.

1204
01:32:28,354 --> 01:32:30,874
So just to get it straight for my listeners,

1205
01:32:31,394 --> 01:32:32,094
going forward,

1206
01:32:32,194 --> 01:32:37,074
if you were to sort of make some suggestions,

1207
01:32:37,394 --> 01:32:39,154
it would be Nostra, obviously.

1208
01:32:39,834 --> 01:32:41,334
It would be Bitcoin,

1209
01:32:41,334 --> 01:32:45,214
but with using a cashier version of Bitcoin would be a good idea.

1210
01:32:45,374 --> 01:32:46,874
Use layers of Bitcoin, right?

1211
01:32:46,934 --> 01:32:49,334
Like I think all of these things, again, comes back to trade-offs.

1212
01:32:49,434 --> 01:32:52,434
Like, you know, you hold your life savings somewhere super, super safe.

1213
01:32:52,534 --> 01:32:54,394
You're not walking around with your life savings constantly.

1214
01:32:54,834 --> 01:32:58,134
You know, you don't like you expect to have a little bit of cash in your wallet.

1215
01:32:58,234 --> 01:33:01,374
And if you get mugged for that little bit of cash, okay, not the end of the world.

1216
01:33:01,514 --> 01:33:04,194
Like, you know, if you get away with your safety, then you're fine.

1217
01:33:05,194 --> 01:33:07,254
And I think that's the way to think about these things as well.

1218
01:33:07,254 --> 01:33:12,794
Like your hardware wallet that's completely disconnected from the internet that holds your life savings doesn't go everywhere with you.

1219
01:33:13,154 --> 01:33:14,414
Your cash wallet does.

1220
01:33:14,814 --> 01:33:16,934
And, you know, maybe the mint rugs you at some point.

1221
01:33:17,234 --> 01:33:20,614
But your money in your cash wallet was even spread across 10 or 20 mints.

1222
01:33:20,954 --> 01:33:22,094
And so it's like nothing.

1223
01:33:22,254 --> 01:33:22,814
It's no big deal.

1224
01:33:23,534 --> 01:33:25,774
You know, it's not life-changing bad for you.

1225
01:33:26,294 --> 01:33:28,134
And so I think, you know, and lightning sits in the middle.

1226
01:33:28,394 --> 01:33:45,888
Lightning is sort of the lingua franca of you know high payments with Bitcoin You know it going to be used by you know the payment settlers all these you know different mints it going to be used by different you know layer two protocols So between Cashew and you know ARK or you know any of those things you going to

1227
01:33:45,888 --> 01:33:50,948
end up with this sort of lightning payments in the middle. And so yeah, for people, I would say,

1228
01:33:51,248 --> 01:33:54,048
you know, yeah, have a lightning wallet, have a Cashew wallet, play with these things,

1229
01:33:54,448 --> 01:33:58,328
you know, do it with a very small amount of money. You know, all of these things are still new tech.

1230
01:33:58,328 --> 01:34:01,868
I mean, I think Lightning is pretty solid at this point, but Cashew is definitely still new.

1231
01:34:03,088 --> 01:34:07,828
But get familiar with them and know which trade-off you're taking and why.

1232
01:34:08,048 --> 01:34:10,948
And also think about it for what application you're using it for.

1233
01:34:11,148 --> 01:34:14,428
And from a privacy standpoint, how would you operate in the world?

1234
01:34:14,748 --> 01:34:17,308
Social media, money, you know, those kind of aspects.

1235
01:34:17,468 --> 01:34:17,648
Yeah.

1236
01:34:17,728 --> 01:34:20,028
I mean, I think just more people need to think about it, right?

1237
01:34:20,088 --> 01:34:21,328
Like I'm doxxed.

1238
01:34:21,448 --> 01:34:22,708
Like I've been on the internet forever.

1239
01:34:23,048 --> 01:34:27,908
I do try to go back regularly and sort of nuke as much as I can, you know, from the past.

1240
01:34:27,908 --> 01:34:32,208
um but what kind of stuff like you mean just the internet's a copy machine though like yeah but

1241
01:34:32,208 --> 01:34:37,428
yeah social media stuff kids right yeah 12 10 and 9 whatever mine are basically the same age and

1242
01:34:37,428 --> 01:34:43,688
it's the same scenario of you know i've explained it's like privacy is a semi one-way door like once

1243
01:34:43,688 --> 01:34:47,488
you go through the door you're not going back and getting it like so what are you telling your kids

1244
01:34:47,488 --> 01:34:51,548
go on then what are you advising one is we i mean we're trying to keep them off of phones for as long

1245
01:34:51,548 --> 01:34:57,848
as possible um how old is your oldest is 12 right mine's pale she has and he doesn't have a smart

1246
01:34:57,848 --> 01:35:03,248
phone and he's killing us over it but like it's like hold the line you know i i was very

1247
01:35:03,248 --> 01:35:12,008
the same for a long time and then i i did i don't know i don't think i'm coping i i because i gave

1248
01:35:12,008 --> 01:35:18,048
her a graphene phone we we operate with signal she has a vpn on there she understands all this

1249
01:35:18,048 --> 01:35:24,968
stuff i figured and i use this as an example you know my wife's email address is hotmail still i

1250
01:35:24,968 --> 01:35:31,128
she's in her 40s and she right okay my bank account is still the hsbc from when i went to

1251
01:35:31,128 --> 01:35:38,268
university at 18 right yes i use vodafone which is 20 i started in the late 90s you know yeah so

1252
01:35:38,268 --> 01:35:45,528
if i start her on graphene proton mail is her email address you know probably it's gonna do well

1253
01:35:45,528 --> 01:35:49,848
that's not the part that matters the part that matters is what what is she posting online and

1254
01:35:49,848 --> 01:35:54,628
where you know what i mean like she's not on social media yeah well actually but there's a

1255
01:35:54,628 --> 01:36:01,068
backdoor social media that's whatsapp yes she's on whatsapp correct in a group and i i she understands

1256
01:36:01,068 --> 01:36:06,008
this now because i said alice that is a social network yeah don't think facebook instagram

1257
01:36:06,008 --> 01:36:11,648
tiktok there's no you're in a group chat there's even you can post what you've been up to today

1258
01:36:11,648 --> 01:36:15,928
it's the same thing but it's but i mean we shouldn't just demonize social media and you

1259
01:36:15,928 --> 01:36:21,008
know blanketly like i think there's there are you know things that are there are redeeming qualities

1260
01:36:21,008 --> 01:36:23,028
about certain social media platforms.

1261
01:36:23,268 --> 01:36:25,008
I think the problem is that like the social media

1262
01:36:25,008 --> 01:36:26,388
that most people look at and think at

1263
01:36:26,388 --> 01:36:28,068
is entirely ad driven.

1264
01:36:28,508 --> 01:36:29,748
And that means you are the product.

1265
01:36:29,888 --> 01:36:31,108
And so that means they will extract

1266
01:36:31,108 --> 01:36:34,248
as much as possible from you in pursuit of that.

1267
01:36:34,368 --> 01:36:37,048
And so to me, it's more helping my kids

1268
01:36:37,048 --> 01:36:38,928
use the right tools, like it sounds like you're doing.

1269
01:36:39,688 --> 01:36:41,308
It's understanding the models

1270
01:36:41,308 --> 01:36:44,128
and like why does a platform want you to be like this?

1271
01:36:44,128 --> 01:36:48,508
And why does this platform have this algorithmic feed

1272
01:36:48,508 --> 01:36:50,928
that its sole job is to keep you there?

1273
01:36:51,008 --> 01:36:56,808
um and if you arm people with those tools and that understanding my hope is that like at least they

1274
01:36:56,808 --> 01:37:01,088
are going in with eyes open and you know you can use these tools to your advantage but just know

1275
01:37:01,088 --> 01:37:04,468
that like just like a knife it's dangerous you know if you use it wrong it's going to cut you

1276
01:37:04,468 --> 01:37:09,728
how long do you think you'll keep your son off smartphone then i think he's probably gonna

1277
01:37:09,728 --> 01:37:15,108
13 is probably where it's i don't think we're going to be able to mine is 13 but she got it at

1278
01:37:15,108 --> 01:37:21,908
12 yeah and i i'd we had well i i had actually in our family liz doesn't care she's not that

1279
01:37:21,908 --> 01:37:27,048
bothered about all this stuff but i had thought because i had i felt the most strongly about it

1280
01:37:27,048 --> 01:37:33,008
it's like no no never and then and everyone said oh but reality is going to hit home and

1281
01:37:33,008 --> 01:37:36,728
the friends are going to do this and blah blah blah and and the one that got me actually in the

1282
01:37:36,728 --> 01:37:44,168
end was the inability to contact me when i was away like right now she's always sending me messages

1283
01:37:44,168 --> 01:37:50,588
all on signal videos this kind of stuff coming through and then you do have to give yourself a

1284
01:37:50,588 --> 01:37:56,808
reality check because it's not all bad and i think people like us we jump straight into the this is

1285
01:37:56,808 --> 01:38:03,148
terrible yes and it is in many ways but it is also not and you can keep your kids away from viruses

1286
01:38:03,148 --> 01:38:07,548
if you want and when they're an adult they'll be they'll get loads and loads of illnesses so

1287
01:38:07,548 --> 01:38:12,928
i'm kind of bringing myself to the middle as fast as i can and i figured this is the best way and

1288
01:38:12,928 --> 01:38:39,308
I agree. I mean, I wholeheartedly agree with that statement. I think, you know, the important thing is, to me at least, is, you know, that my kids are able to take personal responsibility and understand the tradeoffs that they make in certain things. And so like, you want to go on social media, fine. But like, know what you're doing, know what you're getting into, know why you want to do that. Don't just blindly do it because everybody else does it and not think about it. But yeah, there's plenty of amazing things about the internet.

1289
01:38:39,308 --> 01:38:46,888
In my case, we have actually banned her from being on legitimate social media websites.

1290
01:38:47,068 --> 01:38:55,428
But for me, and I've always framed it as when you're old enough, you'll understand why you don't necessarily want your image all over the Internet, basically.

1291
01:38:55,828 --> 01:38:57,928
Sure, it's in WhatsApp chats.

1292
01:38:58,068 --> 01:39:00,008
There's pictures of her in her group chats and stuff.

1293
01:39:00,068 --> 01:39:01,648
But that's the long and short of it.

1294
01:39:02,128 --> 01:39:03,788
And I don't know.

1295
01:39:03,928 --> 01:39:08,208
I'm probably overdoing it because I just don't know because I didn't have that.

1296
01:39:08,208 --> 01:39:15,208
I got on I got a mobile in the late 90s I was a journalist so we used them quite quickly I used

1297
01:39:15,208 --> 01:39:20,248
digital photos early I didn't adopt social media very quickly I remember looking at it thinking

1298
01:39:20,248 --> 01:39:27,308
it was weird and very intrusive yeah you know and and you know but obviously now I do but I use it

1299
01:39:27,308 --> 01:39:32,808
for a business I hardly tweet I tweet business things about the podcast I I've stopped using

1300
01:39:32,808 --> 01:39:38,708
on Facebook, obviously. But if I was going to, it's the most I would want to do is to show

1301
01:39:38,708 --> 01:39:43,788
five people the family photos. We don't need to do that on Facebook. We can do that on a group chat.

1302
01:39:43,788 --> 01:39:47,888
I actually have come full circle. I think I was just eternal tech optimist and loved technology.

1303
01:39:48,168 --> 01:39:53,588
And I was on all the platforms super early. I was still in university when Facebook was launching.

1304
01:39:53,728 --> 01:39:57,668
And so I was in there before most people. I was on Twitter very soon after it launched.

1305
01:39:57,668 --> 01:40:02,128
and for a while it was just like, oh yeah, this is great.

1306
01:40:02,248 --> 01:40:02,908
This is all good.

1307
01:40:03,088 --> 01:40:04,548
Whatever, you can post whatever you want

1308
01:40:04,548 --> 01:40:06,788
and it wasn't until much later that I was like,

1309
01:40:07,068 --> 01:40:07,828
hang on a sec

1310
01:40:07,828 --> 01:40:10,148
and as it started to go darker and darker,

1311
01:40:10,288 --> 01:40:13,528
it was like, okay, what I understood this to be

1312
01:40:13,528 --> 01:40:14,768
is not actually what this is

1313
01:40:14,768 --> 01:40:18,148
and to that one-way door point,

1314
01:40:18,328 --> 01:40:20,268
it's really hard to get rid of stuff on the internet.

1315
01:40:20,528 --> 01:40:21,408
It's a copy machine

1316
01:40:21,408 --> 01:40:23,068
and it's really good at copying things

1317
01:40:23,068 --> 01:40:25,228
and so you can try to delete accounts

1318
01:40:25,228 --> 01:40:32,468
But like, you know, 90% of SaaS startups, when you delete something, it's just a flag in a database that says delete it or not.

1319
01:40:32,548 --> 01:40:33,288
It's still in the database.

1320
01:40:33,408 --> 01:40:33,928
It's still there.

1321
01:40:34,188 --> 01:40:34,948
They still have it.

1322
01:40:35,708 --> 01:40:41,028
So, you know, to anybody in the younger generation, it's really just like, you know, guard it.

1323
01:40:41,068 --> 01:40:42,328
Because once you give it away, it's gone.

1324
01:40:43,408 --> 01:40:48,348
And, you know, be aware of what it actually is and why it is the way it is.

1325
01:40:48,448 --> 01:40:50,948
But I think explaining it like that to kids really works.

1326
01:40:51,028 --> 01:40:51,468
I think so too.

1327
01:40:51,468 --> 01:40:57,688
Because if you say nothing ever gets deleted, it just gets signaled to be not looked at.

1328
01:40:57,808 --> 01:40:57,948
Yeah.

1329
01:40:58,388 --> 01:41:02,468
And then you say, right, you've got a camera in your room looking at you at all times.

1330
01:41:02,468 --> 01:41:03,588
How do you feel about that?

1331
01:41:03,628 --> 01:41:05,028
It's like, I don't want that camera.

1332
01:41:05,188 --> 01:41:05,448
Exactly.

1333
01:41:05,548 --> 01:41:05,688
Right.

1334
01:41:05,748 --> 01:41:07,348
You don't want to do that stuff on the internet.

1335
01:41:07,348 --> 01:41:07,528
Right.

1336
01:41:07,528 --> 01:41:08,688
It's exactly the same thing.

1337
01:41:08,688 --> 01:41:09,808
And they get that.

1338
01:41:09,968 --> 01:41:10,088
Yeah.

1339
01:41:10,148 --> 01:41:12,268
We're private by nature.

1340
01:41:12,388 --> 01:41:19,348
And I really hate this kind of global push to normalize guilt around privacy and stuff.

1341
01:41:19,568 --> 01:41:20,168
Yes, I hate it.

1342
01:41:20,168 --> 01:41:46,508
But I think there's something deeply, deeply disturbing and driven in it. It's very much that we various, you know, interested entities around the world need you to not care about privacy in order to get their way. And you have to really assert it now. And the worst thing is, of course, that as less and less people assert it, you lose your, the people that are asserting it also lose their privacy.

1343
01:41:46,508 --> 01:41:46,908
Exactly.

1344
01:41:47,028 --> 01:41:48,388
Because now we've become targets.

1345
01:41:48,388 --> 01:41:49,488
It's a reverse network effect. Yeah. Yep.

1346
01:41:49,488 --> 01:41:52,688
So, I mean, just comment on that briefly before we go then.

1347
01:41:52,768 --> 01:41:55,508
How do you mitigate that scenario?

1348
01:41:56,148 --> 01:41:59,088
I mean, I think this is a two-pronged thing.

1349
01:41:59,088 --> 01:42:03,008
And it's easy to get blackpilled on privacy because people just don't care.

1350
01:42:03,508 --> 01:42:08,068
They will take convenience every day of the week over maintaining privacy.

1351
01:42:08,728 --> 01:42:16,248
However, I mean, I think we're getting to a shelling point where, you know, the powers that be are pushing things too far, right?

1352
01:42:16,248 --> 01:42:20,348
It will go too far at some point and more and more people will start to pay attention to this.

1353
01:42:21,508 --> 01:42:27,128
I think the other piece is, you know, the younger generations do actually think about this stuff.

1354
01:42:27,188 --> 01:42:33,328
And I think they get, you know, they get the dark half of their experience online pretty quickly.

1355
01:42:33,648 --> 01:42:38,028
And I wonder if we're going to go back towards, you know, less online time.

1356
01:42:38,028 --> 01:42:45,248
You know, people put in like I saw some random video that somebody had posted about, you know, 1985 and like what it was like to grow up in 1985 online.

1357
01:42:45,248 --> 01:42:47,788
And like all these kids were like, is this real?

1358
01:42:47,968 --> 01:43:04,522
Like could this could that actually have been the way things were You know you went over and like hung out with your friends at their house And like you know you had to call on pay phones like they were all in awe that like this could be and this looked so beautiful to them yeah yeah it was like we don go outside we don talk to people face to face

1359
01:43:04,522 --> 01:43:09,342
we don't hang out we don't you know everything has to be pre-programmed and pre you know so i think

1360
01:43:09,342 --> 01:43:16,302
younger the younger generation i think is going to turn faster on this stuff it my my kids are a good

1361
01:43:16,302 --> 01:43:23,362
anecdotal example then because i and i obviously it's us but we have managed to convince them

1362
01:43:23,362 --> 01:43:30,122
that it's uncool to be head in the phone all the time because that's kind of when whenever they hang

1363
01:43:30,122 --> 01:43:36,122
out with someone or if you know my daughter goes on some trip and she's with people she doesn't know

1364
01:43:36,122 --> 01:43:42,442
when we're discussing it afterwards it'll be oh yeah she was just all over you know and she doesn't

1365
01:43:42,442 --> 01:43:47,222
want to associate with that person yes and i don't know whether that one day she might go oh my god

1366
01:43:47,222 --> 01:43:53,462
my mom and dad hoodwinked me or whatever but also we just you know push like we we we push her in

1367
01:43:53,462 --> 01:43:59,962
directions of working outside playing outside right she's a horse rider so we are a great expense

1368
01:43:59,962 --> 01:44:06,122
yeah orchestrated a scenario in which she can ride a horse whenever she wants and all those things

1369
01:44:06,122 --> 01:44:12,982
when you add up make the idea of sitting down and looking at a phone a bad idea yes basically yes

1370
01:44:12,982 --> 01:44:19,782
but it's not a simple thing and i feel deeply for people who who work hey just living imagine

1371
01:44:19,782 --> 01:44:25,802
living in a city in some and not going outside much and having a job that keeps you out of your

1372
01:44:25,802 --> 01:44:31,422
kids lives and blah blah what else is going to happen right right it's it's it's miserable that

1373
01:44:31,422 --> 01:44:36,062
way right and like humans were not made to live like that and um i think we've tried to do the

1374
01:44:36,062 --> 01:44:40,542
exact same thing it's like no go and play like and well you're rock climbers as well so that must

1375
01:44:40,542 --> 01:44:44,402
yeah and we live in a small town so it's super safe and like you know there's all sorts of things

1376
01:44:44,402 --> 01:44:50,682
that you know add up to this and i think this is you know where hopefully you know my idea of the

1377
01:44:50,682 --> 01:44:54,462
promise of the internet was always like you can go anywhere and do anything now like you are not

1378
01:44:54,462 --> 01:45:00,582
constrained by geography anymore and you know i've been working remotely since like 2007 or so 2006

1379
01:45:00,582 --> 01:45:05,502
maybe. And it was so weird for a bunch of years for most people. And now it's become completely

1380
01:45:05,502 --> 01:45:10,162
normal. And, you know, but there's a dark side of that, which is you need to use that to your

1381
01:45:10,162 --> 01:45:13,142
advantage rather than just be like, I can work anywhere. So I'm just going to stay inside and

1382
01:45:13,142 --> 01:45:17,602
work and be on my computer all the time. It's like, no, go somewhere where you actually enjoy

1383
01:45:17,602 --> 01:45:22,222
the place that you get, you know, fulfillment from the things that are around you, interact with the

1384
01:45:22,222 --> 01:45:26,522
people that are around you, you know, and I think that's hard for a lot of people. But at the same

1385
01:45:26,522 --> 01:45:30,682
time like that's where all the contentment and fulfillment happens you know it's in the real world

1386
01:45:30,682 --> 01:45:35,342
and these things fundamentally are just tools like they are tools that are meant to serve us

1387
01:45:35,342 --> 01:45:41,182
not to serve facebook and google and we need to just make that the case amen to that and actually

1388
01:45:41,182 --> 01:45:44,562
we never got to talk about this but it doesn't matter that could happen another time but

1389
01:45:44,562 --> 01:45:51,822
what you're describing is basically where our foundation what we do is all about i think you

1390
01:45:51,822 --> 01:45:56,462
are in the business of dealing with digital freedom and i think we're in the business of

1391
01:45:56,462 --> 01:46:01,662
trying to it's much harder it's much more messier way messier yeah dealing with physical freedom and

1392
01:46:01,662 --> 01:46:07,742
yeah but they are very highly connected as we can see and we'll get more and more intricately

1393
01:46:07,742 --> 01:46:14,202
connected as we go on so yeah absolutely um i was gonna say many thanks for um well number one doing

1394
01:46:14,202 --> 01:46:20,222
what you're doing i think anyone who's doing open source tools for freedom lovers are on my christmas

1395
01:46:20,222 --> 01:46:25,882
list every year they're great people um so thanks for that but also thanks for you know i've never

1396
01:46:25,882 --> 01:46:30,102
met you before i just sent you a message on nostril and said will you come sit down with me and talk

1397
01:46:30,102 --> 01:46:34,962
and you went sure yeah so i love talking about stuff like this talking shop is is great fun uh

1398
01:46:34,962 --> 01:46:40,642
so it's uh i appreciate it right well great anyway and then yeah please please i'd love to reconnect

1399
01:46:40,642 --> 01:46:47,082
in an amount of time in the future next time we run physically in just to see how it's going but

1400
01:46:47,082 --> 01:46:52,402
yeah just for everyone it's it's white noise and it's a nostor thing so you've got to be

1401
01:46:52,402 --> 01:46:57,422
you've got to kind of yeah you can check out the site is just white noise.chat is the website and

1402
01:46:57,422 --> 01:47:00,342
we've got a little bit of information there and you can you know points you to download links and

1403
01:47:00,342 --> 01:47:05,102
things like that um you know if you're not on nostor i would say just uh download primal as

1404
01:47:05,102 --> 01:47:10,122
a starting point it's a great ux to get started you know you create a key pair the app does it for

1405
01:47:10,122 --> 01:47:14,122
you you know if you want to it's got a built-in lightning wallet you can set up really quickly

1406
01:47:14,122 --> 01:47:19,522
for zaps and you know that sort of thing um and just play around like i've heard so many people

1407
01:47:19,522 --> 01:47:24,502
say this is the you know first experience with social media that actually felt like wholesome

1408
01:47:24,502 --> 01:47:29,802
and positive rather than draining and soul destroying like a guy a guy on the panel we

1409
01:47:29,802 --> 01:47:34,282
were doing at baltic honeybeger the other day had a question but he first started out by saying

1410
01:47:34,282 --> 01:47:38,642
thank you all nostr is the only thing that supports my mental health rather than destroys it

1411
01:47:38,642 --> 01:47:39,942
That's like amazing.

1412
01:47:39,942 --> 01:47:42,442
It's the good mornings that I like, obviously.

1413
01:47:42,662 --> 01:47:43,982
One of the redeeming factors.

1414
01:47:44,162 --> 01:47:45,122
Everyone says good morning.

1415
01:47:45,262 --> 01:47:45,722
Yeah, yeah.

1416
01:47:45,862 --> 01:47:46,342
But it is.

1417
01:47:46,382 --> 01:47:48,962
It's like 1950s digital realm, isn't it?

1418
01:47:48,962 --> 01:47:50,002
It feels like that, yeah.

1419
01:47:50,142 --> 01:47:51,582
So it feels great.

1420
01:47:51,642 --> 01:47:55,002
I'm sure the spammers and the scammers and everybody else will show up eventually.

1421
01:47:55,142 --> 01:47:56,622
But for now, it feels like you're somewhere.

1422
01:47:57,142 --> 01:47:58,882
It feels like you're on the internet in 1994.

1423
01:47:58,882 --> 01:48:08,622
But hopefully, you would posit that by the time the scammers turn up, because of your ability on NOS to really tailor make your experience,

1424
01:48:08,642 --> 01:48:12,502
experience yes you should be good for that i mean obviously there's going to be you're going to come

1425
01:48:12,502 --> 01:48:17,302
up against some things but you should be good because the algorithm is completely within your

1426
01:48:17,302 --> 01:48:23,122
within your sphere of control absolutely 100 and like the web of trust aspect that you know we know

1427
01:48:23,122 --> 01:48:27,882
you know from a research point of view webs of trust work really well and it's how our brains

1428
01:48:27,882 --> 01:48:31,882
function when we operate in the real world like do i know this person through someone else how

1429
01:48:31,882 --> 01:48:37,402
much can i trust them um we haven't built out the digital systems to make that work really well yet

1430
01:48:37,402 --> 01:48:44,882
But I think that's part of what Nostra gives us is the ability to really kind of map that analog into the digital world.

1431
01:48:45,142 --> 01:48:47,622
So that, yeah, go ahead, spammers, like have at it.

1432
01:48:47,682 --> 01:48:48,542
Like we don't have to look at it.

1433
01:48:48,562 --> 01:48:49,122
We'll never see it.

1434
01:48:49,222 --> 01:48:49,562
We'll never know.

1435
01:48:49,682 --> 01:48:50,442
It's so exciting.

1436
01:48:50,562 --> 01:48:54,062
I'm so glad I spoke to you this morning because you're right.

1437
01:48:54,122 --> 01:48:55,962
You can get a bit blackbilled about all this stuff.

1438
01:48:55,962 --> 01:49:10,762
But actually, it's so exciting because a lot of things are arising right now, which are very much digital representations of really fundamental old school cultural things that we do.

1439
01:49:10,882 --> 01:49:12,202
Like exactly what you've just described.

1440
01:49:12,282 --> 01:49:13,982
I've never heard anyone describe Nostum like that.

1441
01:49:14,062 --> 01:49:18,282
I know there's the identity part is really important, but you're absolutely right.

1442
01:49:18,402 --> 01:49:18,502
Yeah.

1443
01:49:18,502 --> 01:49:27,182
You know, I knew you because of Max and I trust Max implicitly and I know he wouldn't be mates with you if you were a douchebag, you know.

1444
01:49:27,522 --> 01:49:27,822
Totally.

1445
01:49:27,922 --> 01:49:29,882
There is no way of working that out.

1446
01:49:29,942 --> 01:49:30,102
Right.

1447
01:49:30,362 --> 01:49:31,562
You're absolutely right.

1448
01:49:31,562 --> 01:49:31,982
Right, right.

1449
01:49:32,122 --> 01:49:40,362
And it's kind of interesting that the only decentralized social network, you know, decentralized system that's actually not a social network, Nostra, it's a generic publishing thing.

1450
01:49:40,622 --> 01:49:46,682
But we're sort of, you know, bootstrapping the network by shitposting, you know, like Matt O'Dell always says.

1451
01:49:46,682 --> 01:49:57,282
We are getting to that network effect of knowing who's who by virtue of a bunch of people just posting GM and sending random stuff in a Twitter-like scenario.

1452
01:49:57,902 --> 01:50:02,642
And a lot of other things, like PGP failed because they really couldn't build a network effect around it.

1453
01:50:03,522 --> 01:50:08,462
So having multiple modes to do that means that the entire network is better.

1454
01:50:09,122 --> 01:50:14,102
Any business that comes and builds on Nostra immediately has the entire user base at their disposal.

1455
01:50:14,102 --> 01:50:18,562
and every new business or new user or everything supports everything.

1456
01:50:18,682 --> 01:50:20,662
So it's like a network effect of network effects.

1457
01:50:20,902 --> 01:50:23,942
So I think it's probably the only thing that's in it, you know,

1458
01:50:24,142 --> 01:50:25,482
that has a real chance of functioning.

1459
01:50:26,002 --> 01:50:32,562
And also it's very useful in the sense that if you, a lot of apps now,

1460
01:50:32,562 --> 01:50:34,842
if you want to download a new app on Noster,

1461
01:50:35,262 --> 01:50:38,222
it'll say who out of your friend group already has.

1462
01:50:38,382 --> 01:50:38,762
That's brilliant.

1463
01:50:39,002 --> 01:50:43,402
ZapStore is the quintessential example of how cool Noster can be.

1464
01:50:43,402 --> 01:50:52,722
You can develop an app, you can sign it so that everybody knows that like, hey, me, the person that you're talking to over here is the developer and I've signed this app cryptographically.

1465
01:50:53,242 --> 01:50:56,242
And it's the same thing on the same, you know, on the server.

1466
01:50:56,382 --> 01:50:59,322
And you just freely download it without having to enter any KYC, anything.

1467
01:50:59,542 --> 01:51:03,002
And it shows you that, oh, these mates of yours have already downloaded it.

1468
01:51:03,002 --> 01:51:03,722
I've already downloaded it.

1469
01:51:03,902 --> 01:51:03,922
Cool.

1470
01:51:04,162 --> 01:51:06,582
So I know I'm on the right version and not some scam version.

1471
01:51:06,662 --> 01:51:07,022
Exactly.

1472
01:51:07,022 --> 01:51:22,122
Yeah, yeah. So there's like a million ways where this web of trust, this network, this kind of open publishing system means that we get the possibility of building a brand new internet that's like fully open. Yes. We hope, we hope. Long road though, it's a really long road.

1473
01:51:22,122 --> 01:51:26,182
Yeah, but you say we hope. It's already here. Like we're using it.

1474
01:51:26,302 --> 01:51:26,562
That's true.

1475
01:51:26,642 --> 01:51:28,562
I'm using Nostra every day, much more than Twitter.

1476
01:51:28,882 --> 01:51:29,002
Yeah.

1477
01:51:29,142 --> 01:51:39,582
It's weird that like this morning when I went to message you on Nostra, none of our messages were there. And then like 20 minutes later, they popped up. Yeah, that's weird and annoying, but I'm still using it.

1478
01:51:39,582 --> 01:51:42,842
It's early, but it does work. Yeah, most of the time.

1479
01:51:42,842 --> 01:51:43,942
anyway Jeff

1480
01:51:43,942 --> 01:51:44,702
many many many

1481
01:51:44,702 --> 01:51:45,462
many thanks indeed

1482
01:51:45,462 --> 01:51:45,962
for coming

1483
01:51:45,962 --> 01:51:47,062
and this has been awesome

1484
01:51:47,062 --> 01:51:47,842
I'm really happy

1485
01:51:47,842 --> 01:51:48,542
I really have

1486
01:51:48,542 --> 01:51:50,442
got wrapped my head

1487
01:51:50,442 --> 01:51:51,302
around some important

1488
01:51:51,302 --> 01:51:52,102
concepts here

1489
01:51:52,102 --> 01:51:53,242
so many thanks for coming

1490
01:51:53,242 --> 01:51:54,042
thank you for the invite

1491
01:51:54,042 --> 01:51:55,262
it's been super enjoyable

1492
01:51:55,262 --> 01:51:55,962
I'll see you next time

1493
01:51:55,962 --> 01:51:56,582
yeah thanks

1494
01:52:12,842 --> 01:52:42,822
Thank you.