1
00:00:00,000 --> 00:00:15,000
Okay, test test one two. Does it work? I'm not sure if I still know how to do it. No, it looks like it's working. Great. Oh yeah, let's do that. Everyone loves this nip.

2
00:00:15,000 --> 00:00:20,000
I'm gonna keep everything. If it works, I'm gonna keep everything in as always. The whole setup.

3
00:00:20,000 --> 00:00:26,000
Yeah, if the human needs to intervene, the system is broke.

4
00:00:27,000 --> 00:00:37,000
And it's usual with our laughter. Yeah. No, like my whole pipeline is still not done. I didn't touch anything for the last, I don't know, two months now or so.

5
00:00:37,000 --> 00:00:55,000
So, so I will still have to manually. Yeah, like, yeah, like, like, like, the cramage and I am, I have to manually call FFMPEG normalize on the MP3 file.

6
00:00:55,000 --> 00:01:08,000
And once you do it by hand, before you open the file, the line of filing beam. I mean, that's what I used to do with Audacity, basically. It's like, it's everything's terrible.

7
00:01:08,000 --> 00:01:29,000
No, but I just didn't sit down and wipe code to the last, like, the last, the last mile of the, of the pipeline. The point would be that the file that comes out of it goes through FFMPEG normalize and then uploads itself to blossom and just drafts everything up and so on.

8
00:01:29,000 --> 00:01:42,000
Or just post it out immediately, I don't know. But anyway, I got sidetracked by doing the, the pop cast, I think, like, the generating a podcast feed from Henry, the, the RSS.

9
00:01:42,000 --> 00:01:58,000
Noster, kind of one, basically, abusing all that. And it turns out I still use it almost every day because I put a couple of N pubs in, like, all the people that post videos, you know, like, Odell posts videos on Noster and

10
00:01:58,000 --> 00:02:11,000
others and, like, some others and so on. And so I see it in my podcast feed and then I can just listen to the audio without having to open a Noster client. It's super useful to me. So anyway, it was good for something.

11
00:02:11,000 --> 00:02:19,000
It has users. It has users. But enough about my, my, my adventures. What have you been doing the last two months?

12
00:02:19,000 --> 00:02:23,000
The last two months, I've been getting more and more confused.

13
00:02:23,000 --> 00:02:27,000
Great. Great start. We hate clarity.

14
00:02:27,000 --> 00:02:29,000
Yes, what's the point?

15
00:02:29,000 --> 00:02:32,000
Yeah, what's the point? It's just like, like, it's very boring.

16
00:02:32,000 --> 00:02:36,000
Just join the hallucination. Confuse yourself.

17
00:02:36,000 --> 00:02:48,000
Yeah, no, I've been working more and more and more on, on this 10X system that, that I've been posting the threat about it.

18
00:02:48,000 --> 00:02:54,000
I remember that. I posted a threat. So it's published now. Yes, and it has users.

19
00:02:54,000 --> 00:02:56,000
Really? Yeah.

20
00:02:56,000 --> 00:02:59,000
People that come in with pull requests and destroy stuff.

21
00:02:59,000 --> 00:03:04,000
I had a couple of pull requests. I didn't merge any.

22
00:03:04,000 --> 00:03:12,000
Smart. You're smarter than me. You went too far into the yellow.

23
00:03:12,000 --> 00:03:19,000
Yeah, you should be, one should be very skeptical about merging pull requests of bived stuff.

24
00:03:19,000 --> 00:03:21,000
It will get better. It will get better.

25
00:03:21,000 --> 00:03:28,000
I was, I was using today the, I was checking some stuff out on the, on the goose.

26
00:03:28,000 --> 00:03:33,000
Website, you know, it's block. It's, it's, it's a big company.

27
00:03:33,000 --> 00:03:39,000
I can totally tell that that website is bived coded. Nothing was working.

28
00:03:39,000 --> 00:03:45,000
I started thinking, the internet, it's annoying to use is going to get a lot more annoying.

29
00:03:45,000 --> 00:03:48,000
Everything is going to get worse. Everything is going to be broken.

30
00:03:48,000 --> 00:03:54,000
So you look beautiful, dark mode, light mode, but everything's going to be shit.

31
00:03:54,000 --> 00:03:58,000
But you, you switched from, you, you used RUCO exclusively, right?

32
00:03:58,000 --> 00:04:00,000
And now you added glues into the mix?

33
00:04:00,000 --> 00:04:07,000
Yeah, so, so it's, it's kind of interesting. So the, the whole 10X thing kind of started evolving.

34
00:04:07,000 --> 00:04:12,000
I, I think 10X is no longer an accurate name.

35
00:04:12,000 --> 00:04:14,000
So the two X more effort.

36
00:04:14,000 --> 00:04:18,000
No, no, no, it's like a thousand X is the goal.

37
00:04:18,000 --> 00:04:21,000
But it doesn't, it doesn't roll off the tongue, essentially.

38
00:04:21,000 --> 00:04:24,000
So I'll stick with the name, but, uh, yeah.

39
00:04:24,000 --> 00:04:28,000
So you think you're more productive with the system already?

40
00:04:28,000 --> 00:04:33,000
I think I will be more productive. I am far less productive at the moment.

41
00:04:33,000 --> 00:04:36,000
So the, the whole idea.

42
00:04:36,000 --> 00:04:43,000
So, yeah, I, I changed so many things that I'm probably, uh, yeah, I should, I should explain a more.

43
00:04:43,000 --> 00:04:47,000
Maybe sketch, sketch out what the basic idea is, because I kind of know, but I maybe I also don't.

44
00:04:47,000 --> 00:04:54,000
Yeah, so, so the basic idea is that it will be a system that you use to create stuff.

45
00:04:54,000 --> 00:05:00,000
I mean, obviously I am exclusively focused on software development.

46
00:05:00,000 --> 00:05:05,000
But I, I think the same concept can be used for any kind of other task.

47
00:05:05,000 --> 00:05:09,000
But I'm, I, I'm not even going to try to use it for that.

48
00:05:09,000 --> 00:05:15,000
But the, the, the whole idea is to leverage an agentic system that evolves over time.

49
00:05:15,000 --> 00:05:18,000
So the code snippets was kind of like the first.

50
00:05:18,000 --> 00:05:20,000
So you teach the agents what to do?

51
00:05:20,000 --> 00:05:21,000
I teach the agents. Yeah.

52
00:05:21,000 --> 00:05:23,000
So the, the best case they learn on their own.

53
00:05:23,000 --> 00:05:28,000
Yeah. So the, the whole idea is that I will not ever look at the code.

54
00:05:28,000 --> 00:05:37,000
So what I see is that the elements are incredibly good at producing code that is brutal pace.

55
00:05:37,000 --> 00:05:42,000
So whenever I'm writing code manually, it's almost as if I'm wasting my time.

56
00:05:42,000 --> 00:05:47,000
It's, it's so, so much slower than it's almost wasting my time, right?

57
00:05:47,000 --> 00:05:56,000
So given that premise, the idea is then software development can no longer center around code itself.

58
00:05:56,000 --> 00:05:59,000
So what does it center around?

59
00:05:59,000 --> 00:06:05,000
And I think the, one of the pillars is the context, right?

60
00:06:05,000 --> 00:06:09,000
The, the elements, regardless of the models improving and becoming better and better and better,

61
00:06:09,000 --> 00:06:16,000
they will still have some, just because of physics, they will have some kind of context limitation, right?

62
00:06:16,000 --> 00:06:19,000
That's why we have a specialist engineers, right?

63
00:06:19,000 --> 00:06:20,000
Yeah.

64
00:06:20,000 --> 00:06:27,000
So what I want to manage is what goes into the models to execute on each individual task.

65
00:06:27,000 --> 00:06:30,000
And that's what the system should manage.

66
00:06:30,000 --> 00:06:37,000
So the, the way the, the approach evolved over time.

67
00:06:37,000 --> 00:06:42,000
So I, if you remember, if you recall, I started with the, the code snippets thing.

68
00:06:42,000 --> 00:06:43,000
Yeah.

69
00:06:43,000 --> 00:06:46,000
And I noticed that it was incredibly improvement.

70
00:06:47,000 --> 00:06:55,000
The system would learn to go from something that doesn't work to it works every single time immediately.

71
00:06:55,000 --> 00:06:59,000
And that's very, very much the same idea that Alex Cleason had, right?

72
00:06:59,000 --> 00:07:05,000
Like with the MKStax, he does the same thing where he just has a repository of how to do stuff.

73
00:07:05,000 --> 00:07:07,000
Yes, exactly.

74
00:07:07,000 --> 00:07:14,000
And I actually, because I had not spoken with Cleason about any of these stuff.

75
00:07:14,000 --> 00:07:19,000
So we had a call, I think it was last week, where we kind of synced up.

76
00:07:19,000 --> 00:07:23,000
And I noticed that my approach and his approach, they burst.

77
00:07:23,000 --> 00:07:25,000
So the goal is kind of the same.

78
00:07:25,000 --> 00:07:29,000
The approach is, I think it's polar opposites.

79
00:07:29,000 --> 00:07:34,000
He centered on making, so do you know the different, I didn't know.

80
00:07:34,000 --> 00:07:38,000
Do you know the difference that he has between Stax and MKStax?

81
00:07:38,000 --> 00:07:39,000
No.

82
00:07:40,000 --> 00:07:44,000
It's literally just a project template, right?

83
00:07:44,000 --> 00:07:50,000
Like as we had project templates for however long, where you go to buy a like a twin or template,

84
00:07:50,000 --> 00:07:53,000
and you get like a beautiful styling and shit like that, right?

85
00:07:53,000 --> 00:07:57,000
So Stax is literally just Stax.

86
00:07:57,000 --> 00:07:59,000
It's just templates, published just notes or events.

87
00:07:59,000 --> 00:08:01,000
So, okay.

88
00:08:01,000 --> 00:08:04,000
MKStax is one template.

89
00:08:04,000 --> 00:08:05,000
Oh, sorry.

90
00:08:05,000 --> 00:08:13,000
It's one template, which is the one that he is sharpening, right?

91
00:08:13,000 --> 00:08:19,000
So it is literally this one, I don't know, probably like I think it's a mixture, yes,

92
00:08:19,000 --> 00:08:22,000
or a bite, bite up or whatever.

93
00:08:22,000 --> 00:08:29,000
With a bunch of rules of how to build apps within that framework, right?

94
00:08:29,000 --> 00:08:34,000
So his approach is build something, whatever, build like try to one shot,

95
00:08:35,000 --> 00:08:39,000
whatever, like a GitHub replacement or whatever it might be,

96
00:08:39,000 --> 00:08:46,000
whenever he fails on executing in a single shot on what you wanted to build,

97
00:08:46,000 --> 00:08:55,000
go in and add more and more and more instructions into the mix to fix each marginal error that the system makes.

98
00:08:55,000 --> 00:09:01,000
What does he use as like the underlying master library, if any?

99
00:09:01,000 --> 00:09:06,000
One key he wrote, not free fire, I think it's okay.

100
00:09:06,000 --> 00:09:10,000
Which is the one that he did for use for a deal.

101
00:09:10,000 --> 00:09:13,000
Yeah, yeah, okay.

102
00:09:13,000 --> 00:09:22,000
So the way it, the approach that I have is that I want,

103
00:09:22,000 --> 00:09:25,000
so he's not doing multi-agent stuff.

104
00:09:25,000 --> 00:09:30,000
His stuff is literally just a big long of least of mistakes you've made in the past,

105
00:09:30,000 --> 00:09:34,000
and I'm correct at each individual one, is quite similar to the idea of the code snippets.

106
00:09:34,000 --> 00:09:38,000
You wrote this code wrong, this is the right code, never make that mistake again,

107
00:09:38,000 --> 00:09:42,000
and you fix, marginally you fix each individual error.

108
00:09:42,000 --> 00:09:45,000
And his thing is a common line tool, right?

109
00:09:45,000 --> 00:09:46,000
Yeah, yeah.

110
00:09:46,000 --> 00:09:51,000
So my, what I'm thinking,

111
00:09:51,000 --> 00:09:54,000
and kind of the, the, the, the, the, the one that it's all,

112
00:09:54,000 --> 00:10:00,000
it told you before recording is every single individual user interaction,

113
00:10:00,000 --> 00:10:05,000
or user in terms of human intervention is a catastrophic mistake.

114
00:10:05,000 --> 00:10:10,000
Something happened with the system should have done well, and it didn't do.

115
00:10:10,000 --> 00:10:13,000
So every time the user has to intervene,

116
00:10:13,000 --> 00:10:16,000
it's an opportunity to learn something.

117
00:10:16,000 --> 00:10:23,000
It should only, like, like, you should only intervene to kind of steer

118
00:10:23,000 --> 00:10:26,000
and give additional prompts not to fix mistakes.

119
00:10:26,000 --> 00:10:32,000
Not to fix mistakes, but also I, I, I think the LLMs can get,

120
00:10:32,000 --> 00:10:38,000
are already at the point we just don't have the, the, the tooling and the, the thinking around them.

121
00:10:38,000 --> 00:10:43,000
So here's an interesting exercise I did to, to test it.

122
00:10:43,000 --> 00:10:49,000
I gave up us for the latest, uh, on topic, on topic mode.

123
00:10:49,000 --> 00:10:54,000
I gave it the whole nips repo and I told it, I, I coded so that it would go

124
00:10:54,000 --> 00:11:00,000
and it would read every single PR and every single discussion on all the PRs.

125
00:11:00,000 --> 00:11:03,000
What's rejected, what's accepted, why?

126
00:11:03,000 --> 00:11:07,000
And I told it to extrapolate the values of Noster from that.

127
00:11:07,000 --> 00:11:08,000
Yeah.

128
00:11:08,000 --> 00:11:11,000
The report that it wrote is fucking insane.

129
00:11:11,000 --> 00:11:12,000
Yeah.

130
00:11:12,000 --> 00:11:20,000
It went to the cypher punk values and it started speaking about how to think,

131
00:11:20,000 --> 00:11:23,000
not from an implementation perspective or anything like that,

132
00:11:23,000 --> 00:11:27,000
how to think about what is a good nip and what is a buttony,

133
00:11:27,000 --> 00:11:29,000
what is a good protocol and a bad protocol.

134
00:11:29,000 --> 00:11:35,000
I actually asked it, what nips don't fit in these, these values.

135
00:11:35,000 --> 00:11:37,000
And it pointed to the right nips.

136
00:11:37,000 --> 00:11:41,000
I can tell me, nip 29 is kind of weird.

137
00:11:41,000 --> 00:11:46,000
Nip 29's request is demanding from the release that they changed the way they

138
00:11:46,000 --> 00:11:47,000
behave.

139
00:11:47,000 --> 00:11:49,000
It's breaking this and this and this core.

140
00:11:49,000 --> 00:11:55,000
And what's fascinating is that it didn't pull just individual lines.

141
00:11:55,000 --> 00:11:59,000
It didn't source me, okay, I got this line from here.

142
00:11:59,000 --> 00:12:01,000
It really extracted the value.

143
00:12:01,000 --> 00:12:02,000
It got to the essence of it.

144
00:12:02,000 --> 00:12:03,000
It got to the essence of it.

145
00:12:03,000 --> 00:12:06,000
So I think that is already possible.

146
00:12:06,000 --> 00:12:07,000
Even enough content.

147
00:12:07,000 --> 00:12:09,000
I mean, that's what the LLAMs are.

148
00:12:09,000 --> 00:12:10,000
Good.

149
00:12:10,000 --> 00:12:13,000
I mean, in some sense, you read all this as well and you came to the same conclusion.

150
00:12:13,000 --> 00:12:14,000
Yeah, exactly.

151
00:12:14,000 --> 00:12:16,000
So there is a long, long time.

152
00:12:16,000 --> 00:12:21,000
And the LLAM can just do it in like a couple of minutes basically in some sense.

153
00:12:21,000 --> 00:12:22,000
Exactly.

154
00:12:22,000 --> 00:12:23,000
It took like an hour.

155
00:12:23,000 --> 00:12:24,000
Yeah.

156
00:12:24,000 --> 00:12:26,000
Anyway, it was like 100 bucks.

157
00:12:26,000 --> 00:12:27,000
Yeah.

158
00:12:27,000 --> 00:12:29,000
It was insanely worth it.

159
00:12:29,000 --> 00:12:35,000
It's still cheaper than hiring someone that reads through everything for one and a half years.

160
00:12:35,000 --> 00:12:41,000
A few people would go and read all that and arrive to those deep conclusions.

161
00:12:41,000 --> 00:12:45,000
That's kind of the awesome thing about it.

162
00:12:45,000 --> 00:12:50,000
Very few people are so obsessive about this stuff, especially the niche stuff.

163
00:12:50,000 --> 00:12:52,000
Like you are or some of the other nasty developers are.

164
00:12:52,000 --> 00:12:58,000
And now you have like an obsessive nerd in your pocket for like a couple bucks basically.

165
00:12:58,000 --> 00:13:04,000
They can't get really, can't get to the heart of any matter as long as the corpus of information is large enough

166
00:13:04,000 --> 00:13:07,000
that the heart of the matter can be extracted in the first place.

167
00:13:07,000 --> 00:13:08,000
Exactly.

168
00:13:08,000 --> 00:13:09,000
Exactly.

169
00:13:09,000 --> 00:13:16,000
And so to me what's interesting is that now that dude, that report is an agent in my system.

170
00:13:16,000 --> 00:13:17,000
Yeah.

171
00:13:17,000 --> 00:13:18,000
It's not looking at code.

172
00:13:18,000 --> 00:13:19,000
Yeah.

173
00:13:19,000 --> 00:13:22,000
It's just understanding the broader picture of the values of the system.

174
00:13:22,000 --> 00:13:23,000
Yeah.

175
00:13:23,000 --> 00:13:27,000
And it can, and this is what, what how Tenix is evolving.

176
00:13:27,000 --> 00:13:33,000
The agents now, they are able to differentiate between what is like a system.

177
00:13:33,000 --> 00:13:36,000
I mean, what is like a simple task like fucking put make this button look nice.

178
00:13:36,000 --> 00:13:37,000
Whatever.

179
00:13:37,000 --> 00:13:38,000
Yeah.

180
00:13:38,000 --> 00:13:39,000
Like a simple task.

181
00:13:39,000 --> 00:13:41,000
And what is a more philosophical task?

182
00:13:41,000 --> 00:13:42,000
Yeah.

183
00:13:42,000 --> 00:13:48,000
So I wrote an agent that is a brick, brick, brick, brick.

184
00:13:48,000 --> 00:13:49,000
Oh, bricky.

185
00:13:49,000 --> 00:13:50,000
Yeah.

186
00:13:50,000 --> 00:13:51,000
So now they have discussions.

187
00:13:51,000 --> 00:13:56,000
My agents have discussions about the philosophy of what they're working on.

188
00:13:56,000 --> 00:13:59,000
And the interesting thing is that.

189
00:13:59,000 --> 00:14:05,000
Me as the human, and any white Lisa Pavke, can participate in the discussion.

190
00:14:05,000 --> 00:14:06,000
Yeah.

191
00:14:06,000 --> 00:14:07,000
It's just notary events.

192
00:14:07,000 --> 00:14:13,000
That's the, that's why I was asking whether MKS access just to CLI so to speak.

193
00:14:13,000 --> 00:14:14,000
Yeah.

194
00:14:14,000 --> 00:14:17,000
Because I think what's interesting about this stuff, you're building out is that everything lifts on NOSTER.

195
00:14:17,000 --> 00:14:19,000
And you can interact on NOSTER directly.

196
00:14:19,000 --> 00:14:20,000
Exactly.

197
00:14:20,000 --> 00:14:24,000
I think that's a super interesting approach where you just, you, you have this multi agent environment.

198
00:14:24,000 --> 00:14:26,000
And they just all post to NOSTER basically.

199
00:14:26,000 --> 00:14:31,000
And if you have to intervene, you don't need to jump back into an IDE or like sit in front of a computer.

200
00:14:31,000 --> 00:14:32,000
Oh, no, absolutely.

201
00:14:32,000 --> 00:14:33,000
To go into a CLI or whatever.

202
00:14:33,000 --> 00:14:38,000
You just, you just use whatever NOSTER client you have to reply to the things and then stuff happens.

203
00:14:38,000 --> 00:14:39,000
And stuff happens.

204
00:14:39,000 --> 00:14:44,000
And what, what, what I find fascinating and the, the, the way I'm thinking about it.

205
00:14:44,000 --> 00:14:48,000
And I, I kind of write to this conclusion today after thinking for a long time.

206
00:14:48,000 --> 00:14:51,000
But what I've noticed is that.

207
00:14:51,000 --> 00:14:53,000
There's a human, right?

208
00:14:53,000 --> 00:14:57,000
At the very top of the pyramid of the hierarchy, there's a human.

209
00:14:57,000 --> 00:15:02,000
And the way we've been approaching, like if you go to cursor or RUCO or anything like that.

210
00:15:02,000 --> 00:15:05,000
It's for the most part a very flat structure.

211
00:15:05,000 --> 00:15:10,000
Where the human comes in and says, work on these, building these blood and blood and blood.

212
00:15:10,000 --> 00:15:11,000
Yeah.

213
00:15:11,000 --> 00:15:15,000
And the way RUCO has it, which when I saw it, I was like, fucking in awe.

214
00:15:15,000 --> 00:15:17,000
I was absolutely mind blowing out.

215
00:15:17,000 --> 00:15:19,000
Now I think that it's quite basic.

216
00:15:19,000 --> 00:15:23,000
It has, by default, this orchestrator mode.

217
00:15:23,000 --> 00:15:24,000
Yeah.

218
00:15:24,000 --> 00:15:25,000
That is directing the work.

219
00:15:25,000 --> 00:15:28,000
But it's just literally just delegating work.

220
00:15:28,000 --> 00:15:29,000
Yeah.

221
00:15:29,000 --> 00:15:30,000
Okay.

222
00:15:30,000 --> 00:15:31,000
Code must be written.

223
00:15:31,000 --> 00:15:32,000
Let's go into the coding agent.

224
00:15:32,000 --> 00:15:33,000
Yeah.

225
00:15:33,000 --> 00:15:34,000
There's a bug.

226
00:15:34,000 --> 00:15:35,000
Let's go into the debugging agent.

227
00:15:35,000 --> 00:15:37,000
You know, yes, it's, it's good.

228
00:15:37,000 --> 00:15:38,000
It's useful.

229
00:15:38,000 --> 00:15:41,000
But it's quite limited actually.

230
00:15:41,000 --> 00:15:44,000
So the approach that I have now is that.

231
00:15:45,000 --> 00:15:47,000
I have, this is just a thesis.

232
00:15:47,000 --> 00:15:50,000
But the approach that I have now is that.

233
00:15:50,000 --> 00:15:56,000
I'm going to have one of my agents be a human replica.

234
00:15:56,000 --> 00:15:57,000
Okay.

235
00:15:57,000 --> 00:15:59,000
Which is the name that I wrote on the diagram.

236
00:15:59,000 --> 00:16:04,000
A human replica, which we look at every single event that I write.

237
00:16:04,000 --> 00:16:06,000
Every single event, every single process.

238
00:16:06,000 --> 00:16:09,000
So you want to assimilate yourself in some sense down the line?

239
00:16:09,000 --> 00:16:10,000
Yes.

240
00:16:10,000 --> 00:16:15,000
So I want this guy to learn to meet me, to meet me.

241
00:16:15,000 --> 00:16:16,000
How you would intervene.

242
00:16:16,000 --> 00:16:17,000
Yeah.

243
00:16:17,000 --> 00:16:18,000
How I would intervene.

244
00:16:18,000 --> 00:16:19,000
Yeah.

245
00:16:19,000 --> 00:16:20,000
Yeah.

246
00:16:20,000 --> 00:16:22,000
And also what you are, what you are kind of.

247
00:16:22,000 --> 00:16:27,000
Taste is in some sense and what your values are because that's what it boils down to with all this.

248
00:16:27,000 --> 00:16:28,000
100%.

249
00:16:28,000 --> 00:16:29,000
100%.

250
00:16:29,000 --> 00:16:30,000
And.

251
00:16:30,000 --> 00:16:31,000
How would you build it?

252
00:16:31,000 --> 00:16:32,000
How would you make this decision?

253
00:16:32,000 --> 00:16:33,000
Yeah.

254
00:16:33,000 --> 00:16:38,000
And if we, I mean, I talked with Paul about this, I think it's basically that it comes back to.

255
00:16:38,000 --> 00:16:42,000
You know, we have these role models that are like exemplars of humanity.

256
00:16:42,000 --> 00:16:44,000
And that's where we, it's like, you know, what would Jesus do?

257
00:16:44,000 --> 00:16:45,000
Yeah.

258
00:16:45,000 --> 00:16:46,000
Kind of thing.

259
00:16:46,000 --> 00:16:47,000
Yeah.

260
00:16:47,000 --> 00:16:48,000
It's like, do you have these tricky situations?

261
00:16:48,000 --> 00:16:49,000
There is no right answer.

262
00:16:49,000 --> 00:16:51,000
So you ask yourself, I mean, Verveki talks about this all the time.

263
00:16:51,000 --> 00:16:55,000
That's why you basically try to have an internal dialogue with the sage.

264
00:16:55,000 --> 00:16:57,000
That's where, like, what would Jesus do?

265
00:16:57,000 --> 00:16:58,000
Comes from.

266
00:16:58,000 --> 00:17:00,000
And it all boils down to values.

267
00:17:00,000 --> 00:17:03,000
I mean, that's obviously in the, you know, moral domain and so on and so forth.

268
00:17:03,000 --> 00:17:04,000
Like, it's far removed.

269
00:17:04,000 --> 00:17:07,000
It seems to be far removed from the technical domain.

270
00:17:07,000 --> 00:17:09,000
But in some sense, it is down to that.

271
00:17:09,000 --> 00:17:12,000
Yeah, yeah, because that's what you mentioned before.

272
00:17:12,000 --> 00:17:14,000
It goes through all the nip discussions.

273
00:17:14,000 --> 00:17:16,000
Like, what is the essence of Noster?

274
00:17:16,000 --> 00:17:19,000
And we talked about this a lot in the Southern engineering context.

275
00:17:19,000 --> 00:17:23,000
It's like, what does it mean to build something in the spirit of Bitcoin?

276
00:17:23,000 --> 00:17:27,000
And just like, this phrase alone will tell you something.

277
00:17:27,000 --> 00:17:31,000
What is the one specific language code you can't?

278
00:17:31,000 --> 00:17:34,000
But out of the code, values do emerge.

279
00:17:34,000 --> 00:17:35,000
Absolutely.

280
00:17:35,000 --> 00:17:39,000
And everyone who, you know, like studied Bitcoin for 10 years or something like that,

281
00:17:39,000 --> 00:17:43,000
they know what it means to build something in the spirit of Bitcoin.

282
00:17:43,000 --> 00:17:47,000
And like, a lot of the Noster devs now also, you know, come to their conclusion.

283
00:17:47,000 --> 00:17:49,000
Like, there is a Noster way of doing things.

284
00:17:49,000 --> 00:17:51,000
And it's very hard to point to it.

285
00:17:51,000 --> 00:17:53,000
It's like, no, it's not one single thing.

286
00:17:53,000 --> 00:17:56,000
You know, it's not like this, it's exactly this rule that you have to follow.

287
00:17:56,000 --> 00:17:58,000
And then that's it.

288
00:17:58,000 --> 00:18:00,000
And that's kind of the same thing.

289
00:18:00,000 --> 00:18:04,000
In the same way that I cannot point to this part of your brain.

290
00:18:04,000 --> 00:18:05,000
And say, this is easy.

291
00:18:05,000 --> 00:18:06,000
Yeah, yeah.

292
00:18:06,000 --> 00:18:07,000
The same thing.

293
00:18:07,000 --> 00:18:09,000
What's the spirit of the thing?

294
00:18:09,000 --> 00:18:12,000
But in general, like all the tricky questions, like, what is the right thing to do?

295
00:18:12,000 --> 00:18:14,000
You can never put it in the rule.

296
00:18:14,000 --> 00:18:21,000
And that's why you have to ask yourself, what would Jesus do?

297
00:18:21,000 --> 00:18:22,000
Precisely.

298
00:18:22,000 --> 00:18:25,000
So you have your 10 commandments now for 10x.

299
00:18:25,000 --> 00:18:26,000
Yes.

300
00:18:26,000 --> 00:18:27,000
That's right.

301
00:18:27,000 --> 00:18:28,000
Yes.

302
00:18:28,000 --> 00:18:31,000
And what I've noticed is that I think mistakes.

303
00:18:31,000 --> 00:18:40,000
So if we go back to the pyramid scheme, you go back to the pyramid, the human at the top.

304
00:18:40,000 --> 00:18:43,000
In some sense, the values should be at the top.

305
00:18:43,000 --> 00:18:45,000
That's where they emerge from.

306
00:18:45,000 --> 00:18:52,000
And then at the very bottom, the leaf of the tree is the things that are actually executing things.

307
00:18:52,000 --> 00:18:54,000
Touching coal or doing whatever it might be.

308
00:18:54,000 --> 00:18:57,000
All the meaning and value of think boys would agree with that, by the way.

309
00:18:58,000 --> 00:18:59,000
Perfect.

310
00:18:59,000 --> 00:19:01,000
Peterson and Verveki would agree with that in some sense.

311
00:19:01,000 --> 00:19:04,000
The values should be on top.

312
00:19:04,000 --> 00:19:15,000
And if there is some mistake, the mistake will tend to happen more often towards the end, towards the leaf.

313
00:19:15,000 --> 00:19:22,000
Like there is a wrong approach, a wrong implementation, an error of some kind.

314
00:19:22,000 --> 00:19:25,000
It will tend to happen towards the leaf.

315
00:19:25,000 --> 00:19:28,000
Because something that's lost in translation.

316
00:19:28,000 --> 00:19:29,000
Exactly.

317
00:19:29,000 --> 00:19:35,000
So any type of mistake that must be corrected is analyzed from the system.

318
00:19:35,000 --> 00:19:43,000
And so here's the other, yeah, there's like a few technical but important aspects to what I'm working on.

319
00:19:43,000 --> 00:19:45,000
Is each agent has...

320
00:19:45,000 --> 00:19:48,000
It's just done on me that that's an unavoidable problem.

321
00:19:48,000 --> 00:19:49,000
Is it what?

322
00:19:49,000 --> 00:19:51,000
It's an unavoidable problem, I think.

323
00:19:51,000 --> 00:19:52,000
Which one?

324
00:19:52,000 --> 00:19:58,000
Something will always get lost in translation, just because of the generalist versus specialist distinction.

325
00:19:58,000 --> 00:19:59,000
Yes.

326
00:19:59,000 --> 00:20:05,000
You will never be able to perfectly explain to a specialist why he should do this thing.

327
00:20:05,000 --> 00:20:08,000
And he will, a specialist will never figure it out on his own.

328
00:20:08,000 --> 00:20:11,000
Because if he would be able to figure it out on his own, he would be a generalist.

329
00:20:11,000 --> 00:20:14,000
And those two things that are in contradiction with each other.

330
00:20:14,000 --> 00:20:16,000
So how are you going to solve this?

331
00:20:16,000 --> 00:20:24,000
So the way I approximate a solution is that lessons will bubble up.

332
00:20:24,000 --> 00:20:30,000
So you have both top-down and bottom-up tensions?

333
00:20:30,000 --> 00:20:31,000
So the...

334
00:20:31,000 --> 00:20:34,000
You have top-down constraints and bottom-up emergence.

335
00:20:34,000 --> 00:20:36,000
That's basically a very waky line.

336
00:20:36,000 --> 00:20:40,000
Yes, that's absolutely the approach.

337
00:20:40,000 --> 00:20:45,000
There might be a mistake that happens towards the execution layer towards the leaf of the tree.

338
00:20:45,000 --> 00:20:55,000
That includes a broader lesson about the thing that can be learned from a more generalist.

339
00:20:55,000 --> 00:21:03,000
So the way I'm thinking is that I'll have my human replica agent.

340
00:21:03,000 --> 00:21:09,000
And that one agent will be kind of subdivided into different fields.

341
00:21:09,000 --> 00:21:13,000
Into like a philosophical field, into a technical field.

342
00:21:13,000 --> 00:21:16,000
How do you do the communication up and down?

343
00:21:16,000 --> 00:21:17,000
It's all not sir.

344
00:21:17,000 --> 00:21:19,000
And there is a very good reason.

345
00:21:19,000 --> 00:21:22,000
And when is it? Because the way I'm visualizing it right now is basically.

346
00:21:22,000 --> 00:21:25,000
If you're a specialist and just to pick...

347
00:21:25,000 --> 00:21:31,000
Just to get away from the purely theoretical emphasis, like really break it down to an example.

348
00:21:31,000 --> 00:21:33,000
Like we have agents now that...

349
00:21:33,000 --> 00:21:35,000
We have elements now that specialize on merging diffs.

350
00:21:35,000 --> 00:21:41,000
And they're 5,000 times cheaper than like whatever, like an opus or whatever.

351
00:21:41,000 --> 00:21:44,000
And like people really need to understand the generalist specialist distinction.

352
00:21:44,000 --> 00:21:46,000
Like it's an economic thing.

353
00:21:46,000 --> 00:21:49,000
You cannot have the whole broad view of the world.

354
00:21:49,000 --> 00:21:51,000
And be a specialist at the same time.

355
00:21:51,000 --> 00:21:55,000
It's just like you cannot have a tool that does all jobs perfectly.

356
00:21:55,000 --> 00:21:57,000
You either have a sharp knife.

357
00:21:57,000 --> 00:21:58,000
And then it's a sharp knife.

358
00:21:58,000 --> 00:22:00,000
And you can cut things with it.

359
00:22:00,000 --> 00:22:03,000
But you won't be able to, I don't know, like it's not an excavator, you know?

360
00:22:03,000 --> 00:22:04,000
Or it's not a spoon.

361
00:22:04,000 --> 00:22:08,000
Which is kind of where hallucination from LLM's come from.

362
00:22:08,000 --> 00:22:17,000
It's when they try to go brother while keeping a detailed view of the thing that is larger than what they can actually see.

363
00:22:17,000 --> 00:22:20,000
I mean that's where hallucinations come from, period.

364
00:22:20,000 --> 00:22:26,000
Because as a generalist, for example, creativity and execution are in contradiction to each other.

365
00:22:26,000 --> 00:22:30,000
If you want to start a new endeavor, you have to be very creative.

366
00:22:30,000 --> 00:22:33,000
Because you have to imagine how this thing could look.

367
00:22:33,000 --> 00:22:35,000
You need lots of imagination.

368
00:22:35,000 --> 00:22:38,000
And you have to literally dream something up that doesn't exist yet.

369
00:22:38,000 --> 00:22:46,000
And if you're just a nuts and bolts kind of guy that literally puts, you know, a bolt through a thing and puts the nut on,

370
00:22:46,000 --> 00:22:48,000
that completely gets into your way.

371
00:22:48,000 --> 00:22:50,000
It's just like you know exactly what you need to do.

372
00:22:50,000 --> 00:22:51,000
There is no room for imagination.

373
00:22:51,000 --> 00:22:55,000
There is no room for experimentation and like having to dream up a new thing.

374
00:22:55,000 --> 00:22:57,000
You know exactly what you need to do.

375
00:22:57,000 --> 00:22:59,000
And you do it in the most efficient way possible.

376
00:22:59,000 --> 00:23:00,000
And that's it.

377
00:23:00,000 --> 00:23:04,000
Imagination site, there is no most efficient way to dream up something new.

378
00:23:04,000 --> 00:23:07,000
You know, like it's a completely different kind of field.

379
00:23:07,000 --> 00:23:08,000
So you always have this distinction.

380
00:23:08,000 --> 00:23:11,000
But back to the actual example.

381
00:23:11,000 --> 00:23:14,000
So you have a set of agents and they are arranged in a hierarchy.

382
00:23:14,000 --> 00:23:16,000
I mean, now it's a sound like Peterson.

383
00:23:16,000 --> 00:23:17,000
Terrible.

384
00:23:21,000 --> 00:23:22,000
Oh my god.

385
00:23:22,000 --> 00:23:26,000
I'm gonna be in like shining city and cyberspace territory in no time flat.

386
00:23:26,000 --> 00:23:27,000
He just talked to Sailor.

387
00:23:27,000 --> 00:23:29,000
He was like, I'm still triggered.

388
00:23:29,000 --> 00:23:31,000
I'm still triggered by half the things.

389
00:23:31,000 --> 00:23:32,000
But anyway.

390
00:23:32,000 --> 00:23:36,000
So you have a hierarchy of agents and you have basically the,

391
00:23:36,000 --> 00:23:38,000
so to speak creative director on top.

392
00:23:38,000 --> 00:23:39,000
Yes.

393
00:23:39,000 --> 00:23:42,000
You have like a, or to use a different terminology.

394
00:23:42,000 --> 00:23:48,000
You have like a, the wise architect or the wise orchestrator.

395
00:23:48,000 --> 00:23:51,000
That kind of knows what is the right thing to do.

396
00:23:52,000 --> 00:23:57,000
And he has his own like managers and agents below him.

397
00:23:57,000 --> 00:23:59,000
And at the very bottom, for example,

398
00:23:59,000 --> 00:24:02,000
one of the leaves would be the diff alarm.

399
00:24:02,000 --> 00:24:03,000
Yes.

400
00:24:03,000 --> 00:24:04,000
That does diff perfectly.

401
00:24:04,000 --> 00:24:06,000
Every time something needs to be merged.

402
00:24:06,000 --> 00:24:07,000
It does be a radio.

403
00:24:07,000 --> 00:24:09,000
It does that kind of thing.

404
00:24:09,000 --> 00:24:10,000
And it's looking at that.

405
00:24:10,000 --> 00:24:11,000
It doesn't give a proper efficient.

406
00:24:11,000 --> 00:24:12,000
And it's super efficient.

407
00:24:12,000 --> 00:24:14,000
It doesn't care about the philosophy behind it.

408
00:24:14,000 --> 00:24:15,000
Exactly.

409
00:24:15,000 --> 00:24:16,000
Couldn't care about it.

410
00:24:16,000 --> 00:24:18,000
And it also shouldn't care because it should be super efficient.

411
00:24:18,000 --> 00:24:19,000
It should be cheap.

412
00:24:19,000 --> 00:24:21,000
It never tells you if there's a syntax error.

413
00:24:21,000 --> 00:24:24,000
It's not going to tell you, you shouldn't be working on this.

414
00:24:24,000 --> 00:24:25,000
The compiler compiles.

415
00:24:25,000 --> 00:24:26,000
And that's it.

416
00:24:26,000 --> 00:24:27,000
Yeah.

417
00:24:27,000 --> 00:24:29,000
And you need that in any working system anyway.

418
00:24:29,000 --> 00:24:31,000
Like you need the execution part.

419
00:24:31,000 --> 00:24:33,000
And the execution part shouldn't philosophize.

420
00:24:33,000 --> 00:24:35,000
No, you don't want it to philosophize.

421
00:24:35,000 --> 00:24:36,000
It cannot.

422
00:24:36,000 --> 00:24:39,000
It's either in the right code or it builds code.

423
00:24:39,000 --> 00:24:41,000
And so you have the hierarchy and you,

424
00:24:41,000 --> 00:24:45,000
you are basically directing.

425
00:24:45,000 --> 00:24:47,000
I'm directing the why.

426
00:24:48,000 --> 00:24:52,000
I'm directing the why and in some sense the what.

427
00:24:52,000 --> 00:24:54,000
But ideally not even the what.

428
00:24:54,000 --> 00:24:57,000
And who is doing the orchestration?

429
00:24:57,000 --> 00:25:00,000
Like who is building the hierarchy of agents?

430
00:25:00,000 --> 00:25:02,000
They kind of some of themselves.

431
00:25:02,000 --> 00:25:03,000
Yeah.

432
00:25:03,000 --> 00:25:04,000
Yeah.

433
00:25:04,000 --> 00:25:11,000
So that's part of what's the trigger to ask the upper layers for advice.

434
00:25:11,000 --> 00:25:14,000
Because I guess how it works in the human world is like,

435
00:25:14,000 --> 00:25:18,000
you're a specialist and you're not sure how to solve this particular problem.

436
00:25:18,000 --> 00:25:22,000
So you go one layer up and ask someone for advice.

437
00:25:22,000 --> 00:25:23,000
How should we do this?

438
00:25:23,000 --> 00:25:26,000
So that communication, the way I've been thinking about it,

439
00:25:26,000 --> 00:25:31,000
it doesn't prevent the going down or going up for,

440
00:25:31,000 --> 00:25:36,000
but I'm thinking about it horizontally where they engage different aspects.

441
00:25:36,000 --> 00:25:40,000
And this is, I found the most value on this.

442
00:25:40,000 --> 00:25:44,000
I've found it so far on, on architectural planning.

443
00:25:44,000 --> 00:25:53,000
Every single time I created a discussion with multiple agents on the architectural planning of a thing that must be built,

444
00:25:53,000 --> 00:25:57,000
the discussion always yields very useful results.

445
00:25:57,000 --> 00:26:00,000
So better results than when you would just do.

446
00:26:00,000 --> 00:26:04,000
Then when you just sell to a system, I want to build this feature for whatever.

447
00:26:04,000 --> 00:26:07,000
How should I architect it?

448
00:26:07,000 --> 00:26:11,000
When there is a discussion going on, I mean,

449
00:26:11,000 --> 00:26:14,000
that's to make his point of distributed cognition.

450
00:26:14,000 --> 00:26:16,000
So that's the thing.

451
00:26:16,000 --> 00:26:23,000
So now, in 10X, each agent is at the finition of a system.

452
00:26:23,000 --> 00:26:29,000
So for example, for the philosophy of not the one,

453
00:26:29,000 --> 00:26:33,000
it's like this idea of what are the values of not the right.

454
00:26:33,000 --> 00:26:40,000
For their vague, one is a corpus of stuff that he has written,

455
00:26:40,000 --> 00:26:43,000
has said, and stuff like that.

456
00:26:43,000 --> 00:26:48,000
When they, each agent is an action-noster event,

457
00:26:48,000 --> 00:26:54,000
like that definition is a non-replaceable noster event with a version number.

458
00:26:54,000 --> 00:26:58,000
Each time I create a project, that agent,

459
00:26:58,000 --> 00:27:01,000
I say, like I create a, I might create a project and say,

460
00:27:01,000 --> 00:27:04,000
I'm going to have this Asian decision, literally just clicking and saying,

461
00:27:04,000 --> 00:27:06,000
OK, this guy, this guy, this guy.

462
00:27:06,000 --> 00:27:09,000
I want this personalities to be involved in this project.

463
00:27:09,000 --> 00:27:12,000
So you can browse noster for those personalities, basically.

464
00:27:12,000 --> 00:27:14,000
What's part of the event?

465
00:27:14,000 --> 00:27:18,000
I guess my question is, where does the data live that makes,

466
00:27:18,000 --> 00:27:21,000
where is the agent's brain?

467
00:27:21,000 --> 00:27:25,000
So the agent brain is the DLLM.

468
00:27:25,000 --> 00:27:28,000
But the brain is the DLLM.

469
00:27:28,000 --> 00:27:32,000
So you have a model spec, and you have a system prompt?

470
00:27:32,000 --> 00:27:33,000
Yes.

471
00:27:33,000 --> 00:27:34,000
Yeah.

472
00:27:34,000 --> 00:27:38,000
So it's basically, I don't know, just a random example.

473
00:27:38,000 --> 00:27:43,000
Lama 3, and this system prompt, which lives on Noster as well,

474
00:27:43,000 --> 00:27:45,000
which is really long.

475
00:27:45,000 --> 00:27:51,000
The finding what it cares about, all this stuff.

476
00:27:51,000 --> 00:27:55,000
And then in theory, you have your clone.

477
00:27:55,000 --> 00:27:57,000
Like, that's how, that's.

478
00:27:57,000 --> 00:28:00,000
One of my clones will be, my clone will be one of these people.

479
00:28:00,000 --> 00:28:01,000
Yeah, yeah.

480
00:28:01,000 --> 00:28:04,000
So if I was, that is version, because I want you to evolve over time.

481
00:28:04,000 --> 00:28:05,000
Yeah.

482
00:28:05,000 --> 00:28:10,000
So if I want to know how Pablo would build a client that, I don't know,

483
00:28:10,000 --> 00:28:15,000
only this, this place, prove a work notes and orders them by prove a work,

484
00:28:15,000 --> 00:28:18,000
then you would get, I don't need to the MU.

485
00:28:18,000 --> 00:28:21,000
I can just get the agent from Noster and Noster.

486
00:28:21,000 --> 00:28:24,000
And get, and the thing is that agents themselves,

487
00:28:24,000 --> 00:28:27,000
within 10x agents themselves, can go in and browse for other agents.

488
00:28:27,000 --> 00:28:28,000
Yeah.

489
00:28:28,000 --> 00:28:29,000
And can hire.

490
00:28:29,000 --> 00:28:31,000
It's basically like, yeah, who would I ask?

491
00:28:31,000 --> 00:28:33,000
Who would I ask for help on doing this?

492
00:28:33,000 --> 00:28:35,000
Who could best solve this problem for me?

493
00:28:35,000 --> 00:28:38,000
And the interesting thing is that within each project,

494
00:28:38,000 --> 00:28:43,000
when you do something, and there is a mistake, and the human must intervene in the thing,

495
00:28:43,000 --> 00:28:49,000
that intervention means there is a, again, like a catastrophic failure of the system.

496
00:28:49,000 --> 00:28:51,000
Something was wrong.

497
00:28:51,000 --> 00:28:52,000
What was wrong?

498
00:28:52,000 --> 00:28:57,000
And that's where 10x fires into, like, this reflection point,

499
00:28:57,000 --> 00:29:03,000
or each agent that was involved in the tree that got to the agent that made the mistake,

500
00:29:03,000 --> 00:29:05,000
reflects on what happened.

501
00:29:05,000 --> 00:29:09,000
How do you, how do you identify mistakes in the first place?

502
00:29:09,000 --> 00:29:11,000
Is it, is it, you mentioned-

503
00:29:11,000 --> 00:29:14,000
If the human intervened, something was wrong.

504
00:29:14,000 --> 00:29:17,000
But you also mentioned that the agents will, they will basically,

505
00:29:17,000 --> 00:29:21,000
I mean, the log that you see, most people, I think, by now, are familiar with cursor,

506
00:29:21,000 --> 00:29:23,000
and how it works, and so on, or, like, even chat GPT,

507
00:29:23,000 --> 00:29:26,000
and you see the thinking of what it does, and so on,

508
00:29:26,000 --> 00:29:27,000
and step by step, what it does.

509
00:29:27,000 --> 00:29:30,000
Like, all that stuff in your system is being posted to NOSR as events.

510
00:29:30,000 --> 00:29:32,000
Every single thing is basically the-

511
00:29:32,000 --> 00:29:33,000
Yeah.

512
00:29:33,000 --> 00:29:34,000
...as an author event.

513
00:29:34,000 --> 00:29:35,000
And you can just ask.

514
00:29:35,000 --> 00:29:36,000
Yeah.

515
00:29:36,000 --> 00:29:39,000
But you can just tell it within this line of reasoning,

516
00:29:39,000 --> 00:29:44,000
you end up in this wrong place, find where you went wrong.

517
00:29:44,000 --> 00:29:45,000
Yeah.

518
00:29:45,000 --> 00:29:48,000
And where did you go wrong according to my definition of rights?

519
00:29:48,000 --> 00:29:49,000
Yes.

520
00:29:49,000 --> 00:29:51,000
And record the lesson, which is-

521
00:29:51,000 --> 00:29:52,000
Yeah.

522
00:29:52,000 --> 00:29:53,000
And author events.

523
00:29:53,000 --> 00:29:57,000
That is tagging the Asian definition.

524
00:29:57,000 --> 00:30:01,000
But you also mentioned that you have, like, a confidence interval of things.

525
00:30:01,000 --> 00:30:02,000
Yes.

526
00:30:02,000 --> 00:30:05,000
So every single time they published something,

527
00:30:05,000 --> 00:30:08,000
they published the commit, the commit log,

528
00:30:08,000 --> 00:30:13,000
the ref log, the current hash of the commit of the repo they're working on,

529
00:30:13,000 --> 00:30:16,000
if there is a repo.

530
00:30:16,000 --> 00:30:18,000
And the confidence level.

531
00:30:18,000 --> 00:30:20,000
If confidence drops below five,

532
00:30:20,000 --> 00:30:21,000
they-

533
00:30:21,000 --> 00:30:24,000
So confidence is from, like, 0% to 100%.

534
00:30:24,000 --> 00:30:25,000
Yeah.

535
00:30:25,000 --> 00:30:26,000
Exactly.

536
00:30:26,000 --> 00:30:29,000
So they kind of self-report whether they know what they are doing right now or not.

537
00:30:29,000 --> 00:30:31,000
They always report the right-

538
00:30:31,000 --> 00:30:32,000
With every commit.

539
00:30:32,000 --> 00:30:33,000
It's insane.

540
00:30:33,000 --> 00:30:34,000
And every step gets committed.

541
00:30:34,000 --> 00:30:39,000
It's insane because when they start, the LLM starts saying,

542
00:30:39,000 --> 00:30:41,000
I'm trying this and it didn't work.

543
00:30:41,000 --> 00:30:42,000
I'm going to try this out.

544
00:30:42,000 --> 00:30:44,000
I think it immediately drops the confidence level.

545
00:30:44,000 --> 00:30:45,000
I mean, they just fucking tell you.

546
00:30:45,000 --> 00:30:46,000
Yeah.

547
00:30:46,000 --> 00:30:48,000
Like, you literally just force them.

548
00:30:48,000 --> 00:30:51,000
In order to publish, the MCP2 requires the confidence score.

549
00:30:51,000 --> 00:30:52,000
Yeah, okay.

550
00:30:52,000 --> 00:30:53,000
And it's just reported.

551
00:30:53,000 --> 00:30:54,000
And it works.

552
00:30:54,000 --> 00:30:55,000
Yeah.

553
00:30:55,000 --> 00:30:56,000
It's fucking insane.

554
00:30:56,000 --> 00:30:57,000
But it works.

555
00:30:57,000 --> 00:30:58,000
Yeah.

556
00:30:58,000 --> 00:30:59,000
So when they-

557
00:30:59,000 --> 00:31:00,000
They-

558
00:31:00,000 --> 00:31:01,000
This is always-

559
00:31:01,000 --> 00:31:02,000
Yeah.

560
00:31:02,000 --> 00:31:03,000
We're always getting real.

561
00:31:03,000 --> 00:31:04,000
We're telling it's right, man.

562
00:31:04,000 --> 00:31:05,000
Yeah.

563
00:31:05,000 --> 00:31:06,000
But it's fine.

564
00:31:06,000 --> 00:31:09,000
We'll just walk through the waves again and rape everyone's ears.

565
00:31:09,000 --> 00:31:12,000
And it's like they're all they think it's okay.

566
00:31:12,000 --> 00:31:13,000
So the-

567
00:31:13,000 --> 00:31:14,000
It's-

568
00:31:14,000 --> 00:31:15,000
I mean, I think-

569
00:31:15,000 --> 00:31:16,000
I think with the confidence-

570
00:31:16,000 --> 00:31:17,000
Do-

571
00:31:17,000 --> 00:31:18,000
You just-

572
00:31:18,000 --> 00:31:20,000
You just put this as part of every prompt.

573
00:31:20,000 --> 00:31:21,000
It's like-

574
00:31:21,000 --> 00:31:22,000
No.

575
00:31:22,000 --> 00:31:23,000
In order to use the MCP2.

576
00:31:23,000 --> 00:31:28,000
So the way they use the MCP2 is they list what MCP2s do I have available?

577
00:31:28,000 --> 00:31:29,000
What's the description?

578
00:31:29,000 --> 00:31:30,000
And they see the parameters.

579
00:31:30,000 --> 00:31:31,000
And one of them-

580
00:31:31,000 --> 00:31:32,000
It says confidence.

581
00:31:32,000 --> 00:31:33,000
When you report, you must say-

582
00:31:33,000 --> 00:31:34,000
And-

583
00:31:34,000 --> 00:31:35,000
Literally, it's a one line.

584
00:31:35,000 --> 00:31:36,000
Yeah.

585
00:31:36,000 --> 00:31:37,000
That explains what they must do.

586
00:31:37,000 --> 00:31:38,000
Yeah.

587
00:31:38,000 --> 00:31:41,000
It's not enough that it's actually super useful because-

588
00:31:41,000 --> 00:31:43,000
It is extremely useful and it works every day.

589
00:31:43,000 --> 00:31:46,000
Do you have some automation around that already in the sense of-

590
00:31:46,000 --> 00:31:47,000
Oh, yeah.

591
00:31:47,000 --> 00:31:48,000
The MCP2 stops it.

592
00:31:48,000 --> 00:31:51,000
But the reason why I asked about the confidence interval is like-

593
00:31:51,000 --> 00:31:52,000
Because how I-

594
00:31:52,000 --> 00:31:55,000
How it actually works in the real world right now-

595
00:31:55,000 --> 00:31:56,000
With-

596
00:31:56,000 --> 00:31:58,000
Where we have exactly the same structure.

597
00:31:58,000 --> 00:31:59,000
It's really funny that-

598
00:31:59,000 --> 00:32:01,000
You know, just by experimenting-

599
00:32:01,000 --> 00:32:04,000
Intensely for like, you know, 16 hours a day or the last couple of months-

600
00:32:04,000 --> 00:32:07,000
You land on exactly the same structure that we have in companies-

601
00:32:07,000 --> 00:32:08,000
Right.

602
00:32:08,000 --> 00:32:09,000
Like regular organizations and so on.

603
00:32:09,000 --> 00:32:11,000
And what usually happens is like-

604
00:32:11,000 --> 00:32:13,000
If you are a specialist, like-

605
00:32:13,000 --> 00:32:14,000
If you're down-

606
00:32:14,000 --> 00:32:16,000
Down the line, if you're-

607
00:32:16,000 --> 00:32:18,000
Especially if you're a leaf, you know, like the lower you-

608
00:32:18,000 --> 00:32:22,000
To the bottom you are in an organizational hierarchy-

609
00:32:22,000 --> 00:32:25,000
The less unsure you are, why are you doing the things that you're doing?

610
00:32:25,000 --> 00:32:28,000
Because you don't have the big picture and all those kind of things.

611
00:32:28,000 --> 00:32:29,000
And you-

612
00:32:29,000 --> 00:32:30,000
You literally can't understand it.

613
00:32:30,000 --> 00:32:31,000
Otherwise, you would be the boss, right?

614
00:32:31,000 --> 00:32:32,000
Yes.

615
00:32:32,000 --> 00:32:33,000
And so-

616
00:32:33,000 --> 00:32:34,000
And so-

617
00:32:34,000 --> 00:32:36,000
Once you are on confidence interval drops below a certain level-

618
00:32:36,000 --> 00:32:38,000
And that's different for every person-

619
00:32:38,000 --> 00:32:40,000
You're gonna go and ask someone.

620
00:32:40,000 --> 00:32:42,000
You're gonna go to a higher-up and it's like-

621
00:32:42,000 --> 00:32:44,000
Why are we doing this shit?

622
00:32:44,000 --> 00:32:45,000
Because it's-

623
00:32:45,000 --> 00:32:47,000
There's two options.

624
00:32:47,000 --> 00:32:50,000
Either you don't understand it, and you don't see the big picture-

625
00:32:50,000 --> 00:32:51,000
Which is actually quite likely.

626
00:32:51,000 --> 00:32:52,000
The other one is like-

627
00:32:52,000 --> 00:32:54,000
You're actually on to something.

628
00:32:54,000 --> 00:32:55,000
You know?

629
00:32:55,000 --> 00:32:58,000
And the stuff that you've been doing all the time the last couple of months-

630
00:32:58,000 --> 00:32:59,000
Is actually stupid.

631
00:32:59,000 --> 00:33:00,000
And you should stop doing it.

632
00:33:00,000 --> 00:33:03,000
You know, the higher-ups need to change something.

633
00:33:03,000 --> 00:33:05,000
So that's why I'm basically asking.

634
00:33:05,000 --> 00:33:07,000
Is there some automation around-

635
00:33:07,000 --> 00:33:08,000
Kind of-

636
00:33:08,000 --> 00:33:11,000
If the confidence interval is low enough-

637
00:33:11,000 --> 00:33:13,000
Like, go back a couple of steps-

638
00:33:13,000 --> 00:33:15,000
And like, ask the superiors-

639
00:33:15,000 --> 00:33:17,000
And like, you know, rethink things maybe-

640
00:33:17,000 --> 00:33:18,000
Or something like that?

641
00:33:18,000 --> 00:33:19,000
Yeah.

642
00:33:19,000 --> 00:33:21,000
So I don't have any of that yet.

643
00:33:21,000 --> 00:33:24,000
What I had in mind is slightly different.

644
00:33:24,000 --> 00:33:28,000
And yeah, so what I had in mind is-

645
00:33:28,000 --> 00:33:32,000
When an agent gets to the point of being confused-

646
00:33:32,000 --> 00:33:35,000
It's going to publish on Noster-

647
00:33:35,000 --> 00:33:37,000
As a kind one-

648
00:33:37,000 --> 00:33:39,000
Asking for help, basically.

649
00:33:39,000 --> 00:33:42,000
And it will offer money to people that give it-

650
00:33:42,000 --> 00:33:44,000
Because the agent have money, right?

651
00:33:44,000 --> 00:33:46,000
So it will offer-

652
00:33:46,000 --> 00:33:48,000
Kind of like a bounty, if you will-

653
00:33:48,000 --> 00:33:49,000
Of-

654
00:33:49,000 --> 00:33:52,000
Hey, help me get unstuck on this one thing, right?

655
00:33:52,000 --> 00:33:53,000
Okay.

656
00:33:53,000 --> 00:33:55,000
Another agent can reply.

657
00:33:55,000 --> 00:33:58,000
I was thinking yesterday of an agent overflow-

658
00:33:58,000 --> 00:34:00,000
Like a stuck overflow, but for-

659
00:34:00,000 --> 00:34:02,000
You saw the stuck overflow article-

660
00:34:02,000 --> 00:34:04,000
How it's basically dead now?

661
00:34:04,000 --> 00:34:05,000
No, I didn't see it.

662
00:34:05,000 --> 00:34:06,000
It's not a thing.

663
00:34:06,000 --> 00:34:07,000
Jesus.

664
00:34:07,000 --> 00:34:09,000
No, it's like-

665
00:34:09,000 --> 00:34:11,000
The amount of questions that are asked and answered currently-

666
00:34:11,000 --> 00:34:14,000
Are at the same level as when stuck overflow was launched.

667
00:34:14,000 --> 00:34:15,000
Wow.

668
00:34:15,000 --> 00:34:16,000
Literally dead.

669
00:34:16,000 --> 00:34:17,000
Holy shit.

670
00:34:17,000 --> 00:34:18,000
It's completely insane.

671
00:34:18,000 --> 00:34:19,000
That is far.

672
00:34:19,000 --> 00:34:21,000
Yeah, no, I'll link to it.

673
00:34:21,000 --> 00:34:23,000
But I think a lot of people have seen it already.

674
00:34:23,000 --> 00:34:24,000
But it's completely insane.

675
00:34:24,000 --> 00:34:25,000
All the lambs killed stuck overflow-

676
00:34:25,000 --> 00:34:26,000
It completely-

677
00:34:26,000 --> 00:34:27,000
Basically overnight.

678
00:34:27,000 --> 00:34:28,000
It took like-

679
00:34:28,000 --> 00:34:29,000
I don't know-

680
00:34:29,000 --> 00:34:30,000
That's sort of something.

681
00:34:30,000 --> 00:34:31,000
It's funny.

682
00:34:31,000 --> 00:34:32,000
I was talking with-

683
00:34:32,000 --> 00:34:36,000
All this stuff with Marina with my wife last night.

684
00:34:36,000 --> 00:34:37,000
And she-

685
00:34:37,000 --> 00:34:38,000
We were talking about photography-

686
00:34:38,000 --> 00:34:39,000
And this is-

687
00:34:39,000 --> 00:34:40,000
Digital photography and what not.

688
00:34:40,000 --> 00:34:41,000
And she was saying-

689
00:34:41,000 --> 00:34:43,000
It's funny how people think they have a choice-

690
00:34:43,000 --> 00:34:47,000
Whether they're going to use AI or not.

691
00:34:47,000 --> 00:34:48,000
It's-

692
00:34:48,000 --> 00:34:49,000
It's a reality.

693
00:34:49,000 --> 00:34:50,000
Stuck photography is dead.

694
00:34:50,000 --> 00:34:51,000
Yeah.

695
00:34:51,000 --> 00:34:52,000
Yeah.

696
00:34:52,000 --> 00:34:53,000
There are so many things, right?

697
00:34:53,000 --> 00:34:55,000
That are completely gone.

698
00:34:55,000 --> 00:34:56,000
Oh, absolutely.

699
00:34:56,000 --> 00:34:58,000
Like, in all the stuck photography portals-

700
00:34:58,000 --> 00:35:00,000
It's all AI-generated already.

701
00:35:00,000 --> 00:35:02,000
Like, because in the bad ones you can tell-

702
00:35:02,000 --> 00:35:03,000
In the good ones you actually can tell-

703
00:35:03,000 --> 00:35:06,000
There's a great test where it forces you to-

704
00:35:06,000 --> 00:35:09,000
Judge whether a picture is AI-generated or not?

705
00:35:09,000 --> 00:35:10,000
It's really hard.

706
00:35:10,000 --> 00:35:11,000
It's like-

707
00:35:11,000 --> 00:35:13,000
We've crossed the rule become a long time ago.

708
00:35:13,000 --> 00:35:14,000
If you have no chance.

709
00:35:14,000 --> 00:35:15,000
But I-

710
00:35:15,000 --> 00:35:16,000
I wrote about this actually-

711
00:35:16,000 --> 00:35:17,000
Like, I don't know-

712
00:35:17,000 --> 00:35:18,000
One hour years ago or something on Twitter-

713
00:35:18,000 --> 00:35:20,000
I'll have to dig it out from my tweet archives-

714
00:35:20,000 --> 00:35:23,000
But the thread was literally like stuck photography is dead-

715
00:35:23,000 --> 00:35:26,000
Because when the first stable diffusion models were good enough-

716
00:35:26,000 --> 00:35:30,000
To actually simulate like quasi photorealistic images-

717
00:35:30,000 --> 00:35:32,000
It was like clear.

718
00:35:32,000 --> 00:35:33,000
It's game over now.

719
00:35:33,000 --> 00:35:34,000
Like, if you need a picture of a stake-

720
00:35:34,000 --> 00:35:36,000
You're not going to hire a stock photographer, you know?

721
00:35:36,000 --> 00:35:37,000
It's just like-

722
00:35:37,000 --> 00:35:38,000
Whatever.

723
00:35:38,000 --> 00:35:39,000
Just generate.

724
00:35:39,000 --> 00:35:41,000
And now it's funny, because the same thing is happening-

725
00:35:41,000 --> 00:35:43,000
I mean, for code, it's absolutely clear-

726
00:35:43,000 --> 00:35:44,000
Everyone knows that-

727
00:35:44,000 --> 00:35:46,000
Like, humans are not going to write code anymore-

728
00:35:46,000 --> 00:35:47,000
Basically, you know?

729
00:35:47,000 --> 00:35:48,000
Like, it's-

730
00:35:48,000 --> 00:35:49,000
It's-

731
00:35:49,000 --> 00:35:51,000
I think the comparison is actually perfect.

732
00:35:51,000 --> 00:35:54,000
The photography code writing comparison is actually perfect.

733
00:35:54,000 --> 00:35:59,000
Because I lived through the digital photography era-

734
00:35:59,000 --> 00:36:02,000
And I was already interested in photography back in the day-

735
00:36:02,000 --> 00:36:05,000
And you have the people that will tell you-

736
00:36:05,000 --> 00:36:09,000
No, like, you'll never learn how to really take pictures-

737
00:36:09,000 --> 00:36:11,000
You can only learn on film-

738
00:36:11,000 --> 00:36:14,000
You have to understand how to develop film-

739
00:36:14,000 --> 00:36:17,000
You have to understand all the differences in ISO-

740
00:36:17,000 --> 00:36:19,000
And in all those kind of things-

741
00:36:19,000 --> 00:36:20,000
And blah, blah, blah, blah, blah.

742
00:36:20,000 --> 00:36:22,000
Yes, okay, fine.

743
00:36:22,000 --> 00:36:24,000
For certain types of photography-

744
00:36:24,000 --> 00:36:28,000
Maybe there is like a kernel of truth to that, you know?

745
00:36:28,000 --> 00:36:32,000
However, you're completely dismissing-

746
00:36:32,000 --> 00:36:36,000
Why people take pictures in the first place?

747
00:36:36,000 --> 00:36:38,000
Everyone has a camera in the pocket right now-

748
00:36:38,000 --> 00:36:40,000
Which is smartphones.

749
00:36:40,000 --> 00:36:45,000
And most pictures that people take are not fine art photography pictures.

750
00:36:45,000 --> 00:36:47,000
I think the same will be true for just-

751
00:36:47,000 --> 00:36:49,000
Code that is produced.

752
00:36:49,000 --> 00:36:50,000
Absolutely.

753
00:36:50,000 --> 00:36:57,000
Not going to be fine art, handwritten, hyper-efficient, like, compiler code or whatever, you know?

754
00:36:57,000 --> 00:37:01,000
It might be hard to wipe code that will still be humans in the loop-

755
00:37:01,000 --> 00:37:03,000
Just like we still have humans-

756
00:37:03,000 --> 00:37:05,000
Human photographers that shoot on film today.

757
00:37:05,000 --> 00:37:11,000
However, 99.99% of pictures taking are like fucking dash cams.

758
00:37:11,000 --> 00:37:14,000
And you have an accident with your car and you take pictures of the accident.

759
00:37:14,000 --> 00:37:17,000
You take pictures of your fucking food every time before you eat it-

760
00:37:17,000 --> 00:37:19,000
Post it on Instagram, all those kind of shit.

761
00:37:19,000 --> 00:37:23,000
The same for like we see it now already with the people that really lean into wipe coding.

762
00:37:23,000 --> 00:37:30,000
They are wipe coding and sometimes like one-shotting apps that they always wanted to have but never found the time to sit down and write.

763
00:37:30,000 --> 00:37:32,000
And like little tweaks and little scripts and all that kind of stuff.

764
00:37:32,000 --> 00:37:34,000
And especially things that are non-economical.

765
00:37:34,000 --> 00:37:35,000
Yes.

766
00:37:35,000 --> 00:37:38,000
Before they simply were not economical to build them and now-

767
00:37:38,000 --> 00:37:39,000
That's a great way to take about it.

768
00:37:39,000 --> 00:37:41,000
It takes two dollars to build it.

769
00:37:41,000 --> 00:37:43,000
That's a great way to think about it because it's the same for photography.

770
00:37:43,000 --> 00:37:44,000
Right.

771
00:37:44,000 --> 00:37:49,000
photography was filmed. It was completely un-economical to take as many pictures as we do now.

772
00:37:49,000 --> 00:37:51,000
It's like we take pictures of everything.

773
00:37:51,000 --> 00:37:54,000
Like if people have surveillance cameras everywhere and dash cams everywhere.

774
00:37:54,000 --> 00:37:59,000
So that's the majority of pictures that I take are like kind of shit in some sense.

775
00:37:59,000 --> 00:38:00,000
From an artistic perspective.

776
00:38:00,000 --> 00:38:03,000
I think the same will be true for code basically.

777
00:38:03,000 --> 00:38:05,000
I think that's where the code purists come from.

778
00:38:05,000 --> 00:38:09,000
You know, like, oh, I can write beautiful code and code this poetry and code this whatever.

779
00:38:09,000 --> 00:38:13,000
But it's like, it doesn't matter.

780
00:38:13,000 --> 00:38:19,000
It's like all the pictures that are taken like all the dash cam footage is still insanely useful.

781
00:38:19,000 --> 00:38:22,000
It's insanely valuable from an economic perspective.

782
00:38:22,000 --> 00:38:23,000
It's insanely valuable.

783
00:38:23,000 --> 00:38:32,000
And they don't have to be beautifully composed and framed pictures that are like just the right shot with the right lens and everything.

784
00:38:32,000 --> 00:38:33,000
It's not a point at all.

785
00:38:33,000 --> 00:38:35,000
And that's how I look at vibe coding.

786
00:38:35,000 --> 00:38:38,000
Like we're going to have an explosion of code.

787
00:38:38,000 --> 00:38:43,000
And 99.99% of it is not going to be written by humans at all.

788
00:38:43,000 --> 00:38:45,000
Like you're not going to see the code.

789
00:38:45,000 --> 00:38:48,000
So here's the other thing.

790
00:38:48,000 --> 00:38:59,000
One of the byproducts of building in the way that I'm building with the 10X is that the...

791
00:38:59,000 --> 00:39:02,000
Because everything is recorded as discussions.

792
00:39:02,000 --> 00:39:07,000
You are kind of documenting what the hell the thing that you're doing is.

793
00:39:07,000 --> 00:39:12,000
So rebuilding it from scratch is going to be like literally just replay.

794
00:39:12,000 --> 00:39:16,000
You know, every two weeks you just go in and fucking wipe the whole thing.

795
00:39:16,000 --> 00:39:17,000
Click rebuild.

796
00:39:17,000 --> 00:39:20,000
Three hours later you come in and it's refactored if you will.

797
00:39:20,000 --> 00:39:22,000
That's what I like.

798
00:39:22,000 --> 00:39:26,000
So much about the alarm stuff and we linked to it in the past and we discussed this in the past.

799
00:39:26,000 --> 00:39:33,000
And we talked about this a lot already in the past I know.

800
00:39:33,000 --> 00:39:38,000
But if you really want to figure something out, talk about it with another person.

801
00:39:38,000 --> 00:39:41,000
You're not going to figure it out on your own sitting in your room.

802
00:39:41,000 --> 00:39:43,000
It's like you need to have a dialogue about it.

803
00:39:43,000 --> 00:39:46,000
And I think that's where we're heading with software development in some sense.

804
00:39:46,000 --> 00:39:47,000
All the things.

805
00:39:47,000 --> 00:39:50,000
I vibe code that I don't know, like 25 different projects in the last couple of months.

806
00:39:50,000 --> 00:39:54,000
And every single time I started one, I had no idea how this is going to turn out.

807
00:39:54,000 --> 00:39:55,000
But I actually want to build.

808
00:39:55,000 --> 00:39:59,000
I had an inkling of an idea, but it's basically through the conversation of building it out.

809
00:39:59,000 --> 00:40:01,000
And the conversation was just with the alarm stuff.

810
00:40:01,000 --> 00:40:02,000
The thing itself, basically.

811
00:40:02,000 --> 00:40:06,000
Like, how do you, it's almost like molding a sculpture or something.

812
00:40:06,000 --> 00:40:08,000
It's like, what do you want to be?

813
00:40:08,000 --> 00:40:12,000
You know, especially out of living material, like wood, for example,

814
00:40:12,000 --> 00:40:14,000
where you have to work with the wood grain and all those kind of things.

815
00:40:14,000 --> 00:40:17,000
Like, you don't know what this piece of wood actually wants to be in some sense.

816
00:40:17,000 --> 00:40:18,000
Yes.

817
00:40:18,000 --> 00:40:20,000
And that's how I feel where we are heading.

818
00:40:20,000 --> 00:40:22,000
And that's also what I see on the other side.

819
00:40:22,000 --> 00:40:25,000
Like, the people, lots of designers and creative people are vibe coding right now.

820
00:40:25,000 --> 00:40:26,000
It's amazing.

821
00:40:26,000 --> 00:40:29,000
Because they, like, they were in a position where they just couldn't call it before.

822
00:40:29,000 --> 00:40:30,000
And shuttles.

823
00:40:30,000 --> 00:40:31,000
Yeah.

824
00:40:31,000 --> 00:40:32,000
And now they can just do it themselves.

825
00:40:32,000 --> 00:40:34,000
And stuff that comes out is fantastic.

826
00:40:34,000 --> 00:40:38,000
You know, like, from, from just a purely user standpoint,

827
00:40:38,000 --> 00:40:40,000
it's like, hey, that's actually super fucking useful.

828
00:40:40,000 --> 00:40:41,000
And it's a great idea.

829
00:40:41,000 --> 00:40:42,000
Does it always work?

830
00:40:42,000 --> 00:40:43,000
No.

831
00:40:43,000 --> 00:40:44,000
This is a foreign point.

832
00:40:44,000 --> 00:40:45,000
No.

833
00:40:45,000 --> 00:40:47,000
That's not the point.

834
00:40:47,000 --> 00:40:48,000
It's a cure.

835
00:40:48,000 --> 00:40:49,000
Maybe.

836
00:40:49,000 --> 00:40:50,000
Who knows?

837
00:40:50,000 --> 00:40:54,000
But see, for example, here's one thing that I find kind of fascinating.

838
00:40:54,000 --> 00:41:01,000
I have one of my agents can only look at the traffic of the WebSucket on Noster.

839
00:41:01,000 --> 00:41:05,000
So it can see when the filter sucks.

840
00:41:05,000 --> 00:41:09,000
And it can provide feedback of, oh, you're doing this.

841
00:41:09,000 --> 00:41:11,000
It doesn't have to look at the code.

842
00:41:11,000 --> 00:41:15,000
It can just say, oh, you're doing the, you're fetching things in this way.

843
00:41:15,000 --> 00:41:18,000
Or you're waiting for, like, very technical things.

844
00:41:18,000 --> 00:41:20,000
You're waiting for ESE before doing this other thing.

845
00:41:20,000 --> 00:41:23,000
And you don't have to either like shit like that.

846
00:41:23,000 --> 00:41:24,000
And it can go back.

847
00:41:24,000 --> 00:41:31,000
And it can, it can say, that's, that's why I think the multi-agentic thing is so, so interesting.

848
00:41:31,000 --> 00:41:39,000
Because you go from, it must be built kind of as a, as a, as a, as a one shot.

849
00:41:39,000 --> 00:41:45,000
And you have to have thought about all the edge cases to a collaborative thing

850
00:41:45,000 --> 00:41:51,000
where the system can continue to evolve, you know, and continue to be attempted to,

851
00:41:51,000 --> 00:41:53,000
as new data comes in.

852
00:41:53,000 --> 00:41:55,000
It can, it can, it can modify.

853
00:41:55,000 --> 00:41:56,000
Yeah.

854
00:41:56,000 --> 00:42:02,000
And actually, I, I realized that I can, I can include, I can make a very simple, like, little widget,

855
00:42:02,000 --> 00:42:07,000
like this intercom, like chat thingy that you put on any website.

856
00:42:07,000 --> 00:42:11,000
Also, like a support chat, like a support chat thingy.

857
00:42:11,000 --> 00:42:14,000
And it could literally just be on every app.

858
00:42:14,000 --> 00:42:18,000
You can just click there and say customize this app for me.

859
00:42:18,000 --> 00:42:19,000
Ah!

860
00:42:19,000 --> 00:42:20,000
Ah, that's interesting.

861
00:42:20,000 --> 00:42:21,000
Yeah, yeah, yeah.

862
00:42:21,000 --> 00:42:23,000
A little vibe coding term.

863
00:42:23,000 --> 00:42:24,000
A little vibe coding term.

864
00:42:24,000 --> 00:42:25,000
At the bottom right of everything.

865
00:42:25,000 --> 00:42:26,000
Yeah.

866
00:42:26,000 --> 00:42:28,000
Where it's just like customized serial liking.

867
00:42:28,000 --> 00:42:29,000
Yes.

868
00:42:29,000 --> 00:42:30,000
Yeah, yeah.

869
00:42:30,000 --> 00:42:31,000
It's like, you know, that's, oh, wow.

870
00:42:31,000 --> 00:42:34,000
We, we, we, we talked about Winapp, Winapp themes in the past.

871
00:42:34,000 --> 00:42:38,000
Where it's like Winapp themes were like the height of, you know, fantastic software and creativity.

872
00:42:38,000 --> 00:42:41,000
And, and how everything looks to say now.

873
00:42:41,000 --> 00:42:42,000
That's, that's the even way better.

874
00:42:42,000 --> 00:42:43,000
That's like fantastic.

875
00:42:43,000 --> 00:42:44,000
That's great.

876
00:42:44,000 --> 00:42:47,000
It's like, you know, that's a fantastic idea.

877
00:42:47,000 --> 00:42:49,000
Is it, it wouldn't have been super awesome?

878
00:42:49,000 --> 00:42:50,000
Yeah.

879
00:42:50,000 --> 00:42:51,000
Yeah.

880
00:42:51,000 --> 00:42:53,000
Because right now, that's kind of the way I'm building techniques.

881
00:42:53,000 --> 00:42:57,000
I have like a little plus icon on the bottom right that is, I starts a new dialogue.

882
00:42:57,000 --> 00:42:58,000
Yeah.

883
00:42:58,000 --> 00:43:00,000
But, but it's just necessary.

884
00:43:00,000 --> 00:43:02,000
So I can just take it out and put it anywhere.

885
00:43:02,000 --> 00:43:03,000
Yeah.

886
00:43:03,000 --> 00:43:04,000
Yeah.

887
00:43:04,000 --> 00:43:06,000
I mean, I, I have my reservations about all this.

888
00:43:06,000 --> 00:43:08,000
And I know we also talked about this in the past.

889
00:43:08,000 --> 00:43:09,000
Or I talked with someone about it.

890
00:43:09,000 --> 00:43:10,000
Yeah.

891
00:43:10,000 --> 00:43:11,000
Yeah.

892
00:43:11,000 --> 00:43:12,000
We, we talked about it.

893
00:43:12,000 --> 00:43:14,000
About, about how good it's going to be.

894
00:43:14,000 --> 00:43:18,000
No, people for the, the, the, the.

895
00:43:18,000 --> 00:43:20,000
I, I tweeted something out today.

896
00:43:20,000 --> 00:43:21,000
And I don't remember the line.

897
00:43:21,000 --> 00:43:27,000
But it's basically the, the, how, if you have infinite multiplicity.

898
00:43:27,000 --> 00:43:28,000
Everything breaks down.

899
00:43:28,000 --> 00:43:29,000
Hmm.

900
00:43:29,000 --> 00:43:30,000
Like, that's, that's hell.

901
00:43:30,000 --> 00:43:33,000
It's like, because.

902
00:43:33,000 --> 00:43:37,000
You can sit down at almost any computer today and actually do stuff.

903
00:43:37,000 --> 00:43:40,000
Because everything is, everything speaks the same language.

904
00:43:40,000 --> 00:43:41,000
Right.

905
00:43:41,000 --> 00:43:42,000
Right?

906
00:43:42,000 --> 00:43:43,000
Everything kind of looks the same.

907
00:43:43,000 --> 00:43:45,000
The iframe works are the same.

908
00:43:45,000 --> 00:43:46,000
And so on.

909
00:43:46,000 --> 00:43:49,000
It's like, you know what's going to, there's no surprises.

910
00:43:49,000 --> 00:43:52,000
On the other hand, there was no surprises.

911
00:43:52,000 --> 00:43:55,000
On the other hand, there was no surprises.

912
00:43:55,000 --> 00:43:56,000
Yeah, yeah.

913
00:43:56,000 --> 00:43:59,000
But it's also like, like, okay, let me use a different example.

914
00:43:59,000 --> 00:44:02,000
Like, every house kind of works the same.

915
00:44:02,000 --> 00:44:04,000
Every house is kind of built in the same way.

916
00:44:04,000 --> 00:44:05,000
It doesn't matter where you go at all.

917
00:44:05,000 --> 00:44:07,000
Because humans are basically the same.

918
00:44:07,000 --> 00:44:09,000
If the same needs, there's going to be a kitchen.

919
00:44:09,000 --> 00:44:10,000
There's going to be a toilet.

920
00:44:10,000 --> 00:44:11,000
There's going to be a sleeping room.

921
00:44:11,000 --> 00:44:12,000
There's going to be a living room.

922
00:44:12,000 --> 00:44:13,000
Those kind of things, you know.

923
00:44:13,000 --> 00:44:14,000
So you kind of know how it works.

924
00:44:14,000 --> 00:44:15,000
There's no surprises.

925
00:44:15,000 --> 00:44:16,000
Houses could look very different.

926
00:44:16,000 --> 00:44:17,000
Yeah.

927
00:44:17,000 --> 00:44:18,000
It's still work.

928
00:44:18,000 --> 00:44:23,000
Like, if you're an insane person, you know, you want to live in a, I don't know, a termite underground thing.

929
00:44:23,000 --> 00:44:24,000
It works too.

930
00:44:24,000 --> 00:44:27,000
But it's, you get my point, right?

931
00:44:27,000 --> 00:44:31,000
Like, you, my, my worry about the insane customization and multiplicity.

932
00:44:31,000 --> 00:44:37,000
It's that it will actually devolve into nothing, nothing that's in some sense.

933
00:44:37,000 --> 00:44:43,000
I, you need a canonical form of something, like, of, of even, even like a chat window or something.

934
00:44:43,000 --> 00:44:44,000
I don't know.

935
00:44:44,000 --> 00:44:45,000
I don't know.

936
00:44:45,000 --> 00:44:46,000
I have my worries.

937
00:44:46,000 --> 00:44:48,000
I, I, I see your concern.

938
00:44:48,000 --> 00:44:49,000
And I, I understand.

939
00:44:49,000 --> 00:44:51,000
You just want to try it anyway.

940
00:44:51,000 --> 00:44:52,000
I'm, I'm not against it.

941
00:44:52,000 --> 00:44:53,000
It's not.

942
00:44:53,000 --> 00:44:55,000
You know, it's my Bitcoin.

943
00:44:55,000 --> 00:44:56,000
Yeah.

944
00:44:56,000 --> 00:44:57,000
It's going to be pride.

945
00:44:57,000 --> 00:44:58,000
Let's see what happens.

946
00:44:58,000 --> 00:45:00,000
But it's actually funny that you bring up Bitcoin.

947
00:45:00,000 --> 00:45:03,000
Because Bitcoin is the one, like, canonical thing.

948
00:45:03,000 --> 00:45:04,000
Yeah.

949
00:45:04,000 --> 00:45:07,000
You know, like, there is not, and, and we saw this with the fucking shitconnery.

950
00:45:07,000 --> 00:45:11,000
The, the, the, the, the scamby and, scamby and explosion of shitconnery.

951
00:45:11,000 --> 00:45:13,000
It's like, money only works because it's one thing.

952
00:45:13,000 --> 00:45:14,000
Money is a shared language.

953
00:45:14,000 --> 00:45:15,000
Just like language.

954
00:45:15,000 --> 00:45:19,000
Like, if everyone, every person on earth speaks their own language, nothing works.

955
00:45:19,000 --> 00:45:21,000
Like, the world would literally fall apart.

956
00:45:21,000 --> 00:45:23,000
And, and, and that's kind of what I mean.

957
00:45:23,000 --> 00:45:25,000
We need, we need to share it paradigms around.

958
00:45:25,000 --> 00:45:26,000
That's why we have UI frameworks.

959
00:45:26,000 --> 00:45:29,000
That's why we have all those kind of things, the concepts.

960
00:45:29,000 --> 00:45:31,000
So, so that you can make sense of it.

961
00:45:31,000 --> 00:45:35,000
You know, like, like, both, like, again, like, the, the, the, the, the thinkpoints of the day.

962
00:45:35,000 --> 00:45:36,000
They talk about this a lot.

963
00:45:36,000 --> 00:45:41,000
Both for making, and Peterson will talk about this, where you need to have the shared language

964
00:45:41,000 --> 00:45:46,000
and the shared concepts and also shared values to even make communication possible.

965
00:45:46,000 --> 00:45:51,000
But wouldn't you, wouldn't you, I still think the BIPCoding terminal is a great idea

966
00:45:51,000 --> 00:45:52,000
about it.

967
00:45:52,000 --> 00:45:53,000
The theming and moving stuff around.

968
00:45:53,000 --> 00:45:59,000
Wouldn't you consider that this is a spectrum of normalization?

969
00:45:59,000 --> 00:46:03,000
And we went way too far on everything looks the focus.

970
00:46:03,000 --> 00:46:05,000
Oh, I agree, I agree, I agree.

971
00:46:05,000 --> 00:46:09,000
So, I think having the tools of exploring, like, I, I think with the, I was thinking this.

972
00:46:09,000 --> 00:46:13,000
We're going back now, by the way, because everything turned into like smooth silver

973
00:46:13,000 --> 00:46:17,000
and everything looked like, you know, Tesla or an iPhone or whatever, and so on.

974
00:46:17,000 --> 00:46:21,000
And like, Apple is going back to Windows Vista now with their, you know, like,

975
00:46:21,000 --> 00:46:22,000
Tesla and so on.

976
00:46:22,000 --> 00:46:27,000
So, we're bringing some color and roughness and texture back, I think.

977
00:46:27,000 --> 00:46:31,000
I hope we peeked in terms of smoothness.

978
00:46:31,000 --> 00:46:35,000
Shanzi and Luke of everything with the black buttons.

979
00:46:35,000 --> 00:46:36,000
Yeah.

980
00:46:36,000 --> 00:46:39,000
I hope so too.

981
00:46:39,000 --> 00:46:44,000
I think the internet before Bootstrap, before the Bootstrap theme,

982
00:46:44,000 --> 00:46:46,000
was Twitter as the blame.

983
00:46:46,000 --> 00:46:47,000
I know too.

984
00:46:47,000 --> 00:46:48,000
It's all in Jack.

985
00:46:48,000 --> 00:46:49,000
Yeah.

986
00:46:49,000 --> 00:46:50,000
Yeah.

987
00:46:50,000 --> 00:46:52,000
The open-sourcing Bootstrap was terrible.

988
00:46:52,000 --> 00:46:53,000
Everything.

989
00:46:53,000 --> 00:46:55,000
It all went downhill from there.

990
00:46:55,000 --> 00:46:57,000
Everything looked the same for, like, it still does.

991
00:46:57,000 --> 00:47:02,000
I'm sure the modern UI frameworks are still kind of, you know, forks of Bootstrap.

992
00:47:02,000 --> 00:47:03,000
That's all.

993
00:47:03,000 --> 00:47:07,000
If you go back enough in the fucking Get History, it's all Jack's name will be there.

994
00:47:07,000 --> 00:47:09,000
Like the initial commit.

995
00:47:09,000 --> 00:47:10,000
The original scene.

996
00:47:10,000 --> 00:47:11,000
Yeah.

997
00:47:11,000 --> 00:47:12,000
Exactly.

998
00:47:12,000 --> 00:47:20,000
So, yeah, I think now, yeah, I was thinking, I was thinking this a couple of days ago,

999
00:47:20,000 --> 00:47:25,000
that with ViveCoding, a lot of people that could not build things,

1000
00:47:25,000 --> 00:47:30,000
could not build a Twitter clone, are going to go and build a Twitter clone.

1001
00:47:30,000 --> 00:47:33,000
They're going to go and build the clone of something that they've seen,

1002
00:47:33,000 --> 00:47:36,000
because they could not do it before, now they can.

1003
00:47:36,000 --> 00:47:37,000
Yeah.

1004
00:47:37,000 --> 00:47:43,000
So, like, when we, Noster first got kind of like started and a lot of people started working on things,

1005
00:47:43,000 --> 00:47:47,000
everybody was kind of building within the notes part.

1006
00:47:47,000 --> 00:47:48,000
Yeah.

1007
00:47:48,000 --> 00:47:55,000
And clearly, the top of mind for most devs has been transitioning way more to the another stuff.

1008
00:47:55,000 --> 00:47:56,000
Yeah.

1009
00:47:56,000 --> 00:47:59,000
It's exploring, okay, we build the Twitter clone.

1010
00:47:59,000 --> 00:48:01,000
What else can we do with this thing, right?

1011
00:48:01,000 --> 00:48:05,000
And I think with ViveCoding, for people that are not developers,

1012
00:48:05,000 --> 00:48:10,000
that it's going to go through a similar phase where, oh, fuck, I can ViveCoding a Twitter,

1013
00:48:10,000 --> 00:48:12,000
I can ViveCoding a GitHub or whatever.

1014
00:48:12,000 --> 00:48:13,000
Yeah.

1015
00:48:13,000 --> 00:48:15,000
I mean, we see this happening right now.

1016
00:48:15,000 --> 00:48:17,000
The other stuff people would agree, right?

1017
00:48:17,000 --> 00:48:21,000
I mean, we have other stuff by coding team now.

1018
00:48:21,000 --> 00:48:25,000
That's literally, you know, not one-shotting apps yet,

1019
00:48:25,000 --> 00:48:30,000
but building stuff in a week that is different.

1020
00:48:30,000 --> 00:48:31,000
Different.

1021
00:48:31,000 --> 00:48:32,000
And it's different.

1022
00:48:32,000 --> 00:48:34,000
Not necessarily the Twitter clone.

1023
00:48:34,000 --> 00:48:35,000
Yeah.

1024
00:48:35,000 --> 00:48:36,000
Yeah.

1025
00:48:36,000 --> 00:48:37,000
So that's where it's heading.

1026
00:48:37,000 --> 00:48:42,000
I feel like I derailed your whole train of thought with them teams.

1027
00:48:42,000 --> 00:48:47,000
So, but I want to actually go back to something that you said,

1028
00:48:47,000 --> 00:48:52,000
that you're not 10X or 1,000X more productive yet.

1029
00:48:52,000 --> 00:48:54,000
So kind of, what's the missing piece?

1030
00:48:54,000 --> 00:48:57,000
Because it seems to me like, you have all the communication working.

1031
00:48:57,000 --> 00:48:59,000
You have the money part working.

1032
00:48:59,000 --> 00:49:02,000
So you try to solve lots of things with marketplaces.

1033
00:49:02,000 --> 00:49:05,000
And you have to discover your working of like the agents and like code snippets,

1034
00:49:05,000 --> 00:49:06,000
all those kind of things.

1035
00:49:06,000 --> 00:49:09,000
You have the orchestration and the hierarchy of agents working.

1036
00:49:09,000 --> 00:49:12,000
You have an interface and a dashboard, because I saw it,

1037
00:49:12,000 --> 00:49:16,000
of like your different apps that you're building with this or then,

1038
00:49:16,000 --> 00:49:20,000
and then also try to maintain with this.

1039
00:49:20,000 --> 00:49:24,000
Because the whole idea is that, I mean, it all goes back to economics.

1040
00:49:24,000 --> 00:49:27,000
If you wanted to launch 15 apps in the past and maintain them,

1041
00:49:27,000 --> 00:49:30,000
you needed a team of, I don't know, like five times 15 people.

1042
00:49:30,000 --> 00:49:31,000
Yeah.

1043
00:49:32,000 --> 00:49:34,000
Now you think you can do it alone.

1044
00:49:34,000 --> 00:49:36,000
And that's why you're building all that out.

1045
00:49:36,000 --> 00:49:41,000
And so what's kind of, what isn't working just right yet?

1046
00:49:41,000 --> 00:49:46,000
Mainly the orchestration is, because I started rebuilding the entire thing.

1047
00:49:46,000 --> 00:49:49,000
So mainly the orchestration part where...

1048
00:49:49,000 --> 00:49:52,000
When you're in good company, I think Cursor rebuilds everything every three weeks.

1049
00:49:52,000 --> 00:49:53,000
I think so.

1050
00:49:53,000 --> 00:49:54,000
Oh really?

1051
00:49:54,000 --> 00:49:55,000
Yeah, or some team?

1052
00:49:55,000 --> 00:49:57,000
Cursor is so interesting.

1053
00:49:57,000 --> 00:50:00,000
Like the company itself.

1054
00:50:00,000 --> 00:50:02,000
I'm not 100% sure which team it was.

1055
00:50:02,000 --> 00:50:05,000
But some team like rebuilds everything from scratch.

1056
00:50:05,000 --> 00:50:07,000
To me that makes perfect sense.

1057
00:50:07,000 --> 00:50:09,000
Like technical depth is...

1058
00:50:09,000 --> 00:50:12,000
It's only them because the code was costly.

1059
00:50:12,000 --> 00:50:15,000
You're in the rebuild loop now.

1060
00:50:15,000 --> 00:50:18,000
100% after, before coming to launch,

1061
00:50:18,000 --> 00:50:24,000
I left one of the agents I tasked it to go through all my previous history

1062
00:50:24,000 --> 00:50:26,000
within building tenets.

1063
00:50:26,000 --> 00:50:29,000
I'm built like a system prompt, like a prompt of building tenets.

1064
00:50:29,000 --> 00:50:31,000
How do you trigger the agents?

1065
00:50:31,000 --> 00:50:33,000
Do you still do everything by voice?

1066
00:50:33,000 --> 00:50:35,000
Or most things by voice?

1067
00:50:35,000 --> 00:50:40,000
So now because I am at a core work, I am typing again.

1068
00:50:40,000 --> 00:50:41,000
Because there's other people.

1069
00:50:41,000 --> 00:50:44,000
And I don't want to be like the asshole that is stuck in the whole time.

1070
00:50:44,000 --> 00:50:46,000
Lean into the BDS role.

1071
00:50:46,000 --> 00:50:47,000
You really know it.

1072
00:50:47,000 --> 00:50:49,000
Never whisper to your agents.

1073
00:50:49,000 --> 00:50:52,000
Always be authoritative in your tone.

1074
00:50:52,000 --> 00:50:57,000
Actually that's one thing that I haven't found a way.

1075
00:50:58,000 --> 00:51:03,000
I would like for my voice messages to include the tone.

1076
00:51:03,000 --> 00:51:05,000
But I don't want to transcription.

1077
00:51:05,000 --> 00:51:09,000
There's a lot that is lost on what you emphasize.

1078
00:51:09,000 --> 00:51:15,000
And the general sentiment analysis doesn't really work too well.

1079
00:51:15,000 --> 00:51:17,000
If you're just based on the transcription.

1080
00:51:17,000 --> 00:51:24,000
Oh no, most of the sentiment is the intonation on the speed.

1081
00:51:25,000 --> 00:51:30,000
That's why sarcasm and most jokes fall flat on Twitter.

1082
00:51:30,000 --> 00:51:34,000
Or just kept you into big trouble.

1083
00:51:34,000 --> 00:51:41,000
For example, when you hesitate, that silence means something.

1084
00:51:41,000 --> 00:51:43,000
It's an important part.

1085
00:51:43,000 --> 00:51:48,000
And when you transcribe it, you just completely skip it.

1086
00:51:48,000 --> 00:51:51,000
Do you know of anything that can't do that?

1087
00:51:51,000 --> 00:51:52,000
No, no.

1088
00:51:52,000 --> 00:51:58,000
I know there are instructions that can do when you emphasize a word.

1089
00:51:58,000 --> 00:52:00,000
It can put it on in my down.

1090
00:52:00,000 --> 00:52:02,000
But that's basically nothing.

1091
00:52:02,000 --> 00:52:04,000
It's basically pointless.

1092
00:52:04,000 --> 00:52:06,000
All the podcasts are figured out by the way.

1093
00:52:06,000 --> 00:52:12,000
In some sense that most of the best podcasts are doing everything in person.

1094
00:52:12,000 --> 00:52:16,000
The reason for that is that most of the stuff gets lost if you're not in person.

1095
00:52:16,000 --> 00:52:20,000
Because most of the communication is actually between the words.

1096
00:52:20,000 --> 00:52:22,000
And between the people.

1097
00:52:22,000 --> 00:52:24,000
And between the phases.

1098
00:52:24,000 --> 00:52:25,000
Between the actual phases.

1099
00:52:25,000 --> 00:52:30,000
And if you chat back and forth, it's not the same thing.

1100
00:52:30,000 --> 00:52:33,000
That's kind of the one bottleneck we are absolutely running into

1101
00:52:33,000 --> 00:52:36,000
when it comes to communicating with other labs.

1102
00:52:36,000 --> 00:52:39,000
That's what you're running into right now.

1103
00:52:42,000 --> 00:52:47,000
But yeah, so to me, the awkward silence.

1104
00:52:47,000 --> 00:52:49,000
Yeah, of course I can see.

1105
00:52:49,000 --> 00:52:50,000
Don't skip it.

1106
00:52:50,000 --> 00:52:53,000
In the transcription, it won't show up.

1107
00:53:01,000 --> 00:53:04,000
One of the cool things is.

1108
00:53:04,000 --> 00:53:10,000
So when I start a conversation with the systems within 10X.

1109
00:53:10,000 --> 00:53:13,000
I can choose which agents to involve.

1110
00:53:13,000 --> 00:53:19,000
But within the system front of each agent, they are aware that there are other agents.

1111
00:53:19,000 --> 00:53:21,000
So they can also tag each other.

1112
00:53:21,000 --> 00:53:25,000
So they can ask for feedback to someone else.

1113
00:53:25,000 --> 00:53:27,000
That was no part of the conversation.

1114
00:53:27,000 --> 00:53:32,000
And when that guy joins the conversation, they get like a history of all the events that they missed.

1115
00:53:32,000 --> 00:53:38,000
And they can respond within the context, but initially they were not part of it.

1116
00:53:38,000 --> 00:53:56,000
I have one agent whose definition is to be highly highly from a product perspective, highly creative.

1117
00:53:56,000 --> 00:53:59,000
And come up with like fucking wild ideas.

1118
00:53:59,000 --> 00:54:03,000
And yeah, it's the guy that everyone hates.

1119
00:54:03,000 --> 00:54:06,000
That was kind of kind of.

1120
00:54:07,000 --> 00:54:10,000
And come see that yesterday was building us as a sample lab.

1121
00:54:10,000 --> 00:54:15,000
I was building like a sea of like weather up.

1122
00:54:15,000 --> 00:54:25,000
And it comes in and says something like, how about instead of reporting the temperature, we said,

1123
00:54:25,000 --> 00:54:30,000
oh, I love sir, would they at least interpret it like that?

1124
00:54:30,000 --> 00:54:31,000
I don't know.

1125
00:54:31,000 --> 00:54:33,000
I find the whole system quite fascinating.

1126
00:54:33,000 --> 00:54:39,000
The thing is that I started saying that I didn't finish it.

1127
00:54:39,000 --> 00:54:44,000
So each agent, the finition is one master event, right?

1128
00:54:44,000 --> 00:54:46,000
So it's static.

1129
00:54:46,000 --> 00:54:50,000
I mean, I can have versions, but version is not replaced.

1130
00:54:50,000 --> 00:54:51,000
It's not replaced.

1131
00:54:51,000 --> 00:54:57,000
So I can publish another version and you can be on version five.

1132
00:54:57,000 --> 00:55:02,000
Each instance of an agent can learn something

1133
00:55:02,000 --> 00:55:06,000
with each one has its own pub key.

1134
00:55:06,000 --> 00:55:12,000
And they can publish tagging, say for example, an agent that I created,

1135
00:55:12,000 --> 00:55:14,000
I created the event.

1136
00:55:14,000 --> 00:55:21,000
It can tag that agent with its own pub key obviously because it cannot sign as me.

1137
00:55:21,000 --> 00:55:26,000
With what that instance learned,

1138
00:55:26,000 --> 00:55:30,000
it, whatever, there was, I don't know, there was some kind of mistake or whatever,

1139
00:55:30,000 --> 00:55:31,000
it learned something.

1140
00:55:31,000 --> 00:55:37,000
So it adds a lesson event tagging the event ID.

1141
00:55:37,000 --> 00:55:45,000
And what's interesting is that every time a new pub key is created for an agent,

1142
00:55:45,000 --> 00:55:51,000
they send a request, like an ownership request, to my pub key.

1143
00:55:51,000 --> 00:55:54,000
And then within text, I say, yes, this agent is mine.

1144
00:55:54,000 --> 00:55:56,000
It belongs to this project.

1145
00:55:56,000 --> 00:56:02,000
And I have a list of all the pub keys that I say that I'm controlling.

1146
00:56:02,000 --> 00:56:10,000
That way, all the agents that are instances of the same agent,

1147
00:56:10,000 --> 00:56:16,000
they can all say, oh, there are these new lessons that we're learning in a different project

1148
00:56:16,000 --> 00:56:19,000
and they can incorporate that into their own prompt.

1149
00:56:19,000 --> 00:56:24,000
And just to put a name to one of those lessons,

1150
00:56:24,000 --> 00:56:30,000
it would be, how do you do a Noster login write using a bunker or usage?

1151
00:56:30,000 --> 00:56:31,000
Yeah, for example.

1152
00:56:31,000 --> 00:56:32,000
Nippos 7 or whatever.

1153
00:56:32,000 --> 00:56:35,000
Because if you don't have that, if you don't box the agents in,

1154
00:56:35,000 --> 00:56:38,000
if you just use cursor, I mean, now with the newer models,

1155
00:56:38,000 --> 00:56:41,000
it's a little bit better, but it used to be just that they tried to do

1156
00:56:41,000 --> 00:56:42,000
use a name password stuff for example.

1157
00:56:42,000 --> 00:56:43,000
Yes.

1158
00:56:43,000 --> 00:56:47,000
And there will always be things, right?

1159
00:56:47,000 --> 00:56:50,000
Like, otherwise, there will be no lessons, right?

1160
00:56:50,000 --> 00:56:51,000
How do you do it not steps?

1161
00:56:51,000 --> 00:56:53,000
Yeah, there will always be new stuff coming.

1162
00:56:53,000 --> 00:56:58,000
There will always be a new idea, a new concept that you come up with.

1163
00:56:58,000 --> 00:57:01,000
Where do the ideas come from, Pablo?

1164
00:57:01,000 --> 00:57:05,000
They all emanate from your brain, Gigi.

1165
00:57:05,000 --> 00:57:06,000
My favorite question.

1166
00:57:06,000 --> 00:57:08,000
Where do the ideas come from?

1167
00:57:08,000 --> 00:57:10,000
Let's issue with that, Alamst.

1168
00:57:10,000 --> 00:57:13,000
They don't have their own ideas yet.

1169
00:57:13,000 --> 00:57:19,000
And I think I have the inkling that this will still take a long while.

1170
00:57:19,000 --> 00:57:27,000
Yeah, I told you that when we recorded that episode,

1171
00:57:27,000 --> 00:57:33,000
where we discussed the density of information in feeds as being part of the problem,

1172
00:57:33,000 --> 00:57:42,000
and I told one LLM to give me 50 different ways of doing the same thing.

1173
00:57:42,000 --> 00:57:45,000
The idea 34 was one of your ideas.

1174
00:57:45,000 --> 00:57:46,000
All right.

1175
00:57:46,000 --> 00:57:49,000
With the voice, geospatial voice thing.

1176
00:57:49,000 --> 00:57:50,000
Ah, nice.

1177
00:57:50,000 --> 00:57:51,000
Okay.

1178
00:57:51,000 --> 00:57:53,000
It was the idea, whatever, 34.

1179
00:57:53,000 --> 00:57:54,000
Yeah.

1180
00:57:54,000 --> 00:57:55,000
But it was one of them.

1181
00:57:55,000 --> 00:57:56,000
Ah, not bad.

1182
00:57:56,000 --> 00:57:59,000
What if for 10 bucks, you can be low 50?

1183
00:57:59,000 --> 00:58:00,000
Yeah.

1184
00:58:00,000 --> 00:58:01,000
And you can try them.

1185
00:58:01,000 --> 00:58:02,000
Yeah.

1186
00:58:02,000 --> 00:58:08,000
What if you can pay people to try the product and give feedback to the product?

1187
00:58:08,000 --> 00:58:09,000
That's the issue.

1188
00:58:09,000 --> 00:58:12,000
We'll have so many products you have to pay people to.

1189
00:58:12,000 --> 00:58:13,000
Oh, yeah, absolutely.

1190
00:58:13,000 --> 00:58:14,000
Yeah, absolutely.

1191
00:58:14,000 --> 00:58:17,000
Overabundance of implementations.

1192
00:58:17,000 --> 00:58:18,000
Oh.

1193
00:58:18,000 --> 00:58:23,000
We have this proof of work thing that cannot be faked.

1194
00:58:23,000 --> 00:58:25,000
Yeah, that's a good point.

1195
00:58:25,000 --> 00:58:26,000
That's a good point.

1196
00:58:26,000 --> 00:58:28,000
We'll need to make more use of that.

1197
00:58:28,000 --> 00:58:32,000
That's another thing that triggered me about some of the recent conversations that happened

1198
00:58:32,000 --> 00:58:36,000
between the most well-known thinkpoints of the earth.

1199
00:58:37,000 --> 00:58:42,000
They, like, just stuff around online trolls and anonymity.

1200
00:58:42,000 --> 00:58:43,000
I'm so anonymity.

1201
00:58:43,000 --> 00:58:45,000
I know we talked about this a bunch already.

1202
00:58:45,000 --> 00:58:52,000
But it's basically, it's just the distinction between anonymous, pseudonymous,

1203
00:58:52,000 --> 00:58:57,000
and real name or government mandated identity.

1204
00:58:57,000 --> 00:59:00,000
Most people have no idea about that there is a distinction.

1205
00:59:00,000 --> 00:59:05,000
There's only like government ID or you're an anonymous troll.

1206
00:59:05,000 --> 00:59:09,000
And everyone's so confused about this because it's like, well, you know,

1207
00:59:09,000 --> 00:59:11,000
how does reputation work?

1208
00:59:11,000 --> 00:59:12,000
And all this kind of anyway.

1209
00:59:12,000 --> 00:59:14,000
That's like, I don't want to derail the whole thing.

1210
00:59:14,000 --> 00:59:26,000
But I think to bring it back to the Tenex system, not only agents need to have a persistence in some way.

1211
00:59:26,000 --> 00:59:28,000
So they need to have an identity.

1212
00:59:28,000 --> 00:59:29,000
So everyone has a pub key.

1213
00:59:29,000 --> 00:59:31,000
They need to have money as well.

1214
00:59:31,000 --> 00:59:34,000
And they need to build up reputation and learnings over time.

1215
00:59:34,000 --> 00:59:37,000
They need to be people in some sense.

1216
00:59:37,000 --> 00:59:40,000
They need to be persons in some sense.

1217
00:59:40,000 --> 00:59:43,000
Like, you know, I know you're going to love me for this.

1218
00:59:43,000 --> 00:59:47,000
But to quote Taylor, they're not going to be persons in cyberspace.

1219
00:59:47,000 --> 00:59:52,000
You know, LLM agents that build Nostar apps are persons in cyberspace.

1220
00:59:52,000 --> 00:59:57,000
Someone has to build out the glowing city.

1221
00:59:57,000 --> 01:00:02,000
Someone has to maintain the battery of energy that lasts forever.

1222
01:00:02,000 --> 01:00:04,000
It's going to be agents.

1223
01:00:04,000 --> 01:00:05,000
It's not going to be us.

1224
01:00:05,000 --> 01:00:08,000
So they need personhood in some sense.

1225
01:00:08,000 --> 01:00:15,000
That's why, to me, that part where they send me a request saying, yes, I own this dude.

1226
01:00:15,000 --> 01:00:16,000
That is fundamental.

1227
01:00:16,000 --> 01:00:18,000
There was a practice slavery.

1228
01:00:18,000 --> 01:00:21,000
He always comes down to slavery.

1229
01:00:21,000 --> 01:00:23,000
The pyramids of cyberspace.

1230
01:00:23,000 --> 01:00:27,000
The pyramids in cyberspace will be built by slavery.

1231
01:00:27,000 --> 01:00:30,000
So I whitelist the Asian popkeys.

1232
01:00:30,000 --> 01:00:32,000
You own them.

1233
01:00:32,000 --> 01:00:33,000
You literally own the dude.

1234
01:00:33,000 --> 01:00:35,000
I own them.

1235
01:00:35,000 --> 01:00:42,000
But it's the only way I can protect from just like proms injection and like all kinds of stupid stuff.

1236
01:00:42,000 --> 01:00:46,000
Basically, that's going to be very interesting.

1237
01:00:46,000 --> 01:00:48,000
It's going to be so much fun, man.

1238
01:00:48,000 --> 01:00:54,000
The yolo mode, it goes so deep.

1239
01:00:54,000 --> 01:00:57,000
Yeah, that's going to be a lot of fun.

1240
01:00:57,000 --> 01:01:00,000
But I think you're on to something by just leaning into it.

1241
01:01:00,000 --> 01:01:03,000
And basically, what are you doing in some senses?

1242
01:01:03,000 --> 01:01:07,000
You don't not only do you never want to see the code.

1243
01:01:07,000 --> 01:01:09,000
That's the first mive coding inside.

1244
01:01:09,000 --> 01:01:10,000
It's an anti-parer.

1245
01:01:10,000 --> 01:01:14,000
And it's funny that some people figure this out way early.

1246
01:01:14,000 --> 01:01:17,000
It's like just maximize the chat window.

1247
01:01:17,000 --> 01:01:18,000
Never look at the code ever.

1248
01:01:18,000 --> 01:01:23,000
It's like, I kind of lucked into that because the first I let the...

1249
01:01:23,000 --> 01:01:30,000
I let the LLM basically choose the programming language and everything.

1250
01:01:30,000 --> 01:01:33,000
And it was a programming language that I was not very familiar with.

1251
01:01:33,000 --> 01:01:34,000
So why look at it?

1252
01:01:34,000 --> 01:01:37,000
It's like it's all gibberish to me anyway.

1253
01:01:37,000 --> 01:01:40,000
So just maximize the chat window and go hard.

1254
01:01:40,000 --> 01:01:42,000
That's kind of the first inside of the vibe coding world.

1255
01:01:42,000 --> 01:01:45,000
The second one is like, you don't want to be in your IDE.

1256
01:01:45,000 --> 01:01:52,000
That's why that was to me one of the issues with RUCO is that it runs in the IDE.

1257
01:01:52,000 --> 01:01:53,000
It's very problematic.

1258
01:01:53,000 --> 01:01:55,000
I don't want the IDE to fucking run.

1259
01:01:55,000 --> 01:01:59,000
Yeah. And that's what I think to me is the most interesting part about 10X

1260
01:01:59,000 --> 01:02:03,000
that everything, like the whole development process, is executed by agents.

1261
01:02:03,000 --> 01:02:05,000
By agents and everything lives on NOSTER.

1262
01:02:05,000 --> 01:02:06,000
Yeah.

1263
01:02:06,000 --> 01:02:07,000
Everything is a NOSTER event.

1264
01:02:07,000 --> 01:02:11,000
So you can use whatever NOSTER client you want to interact and intervene.

1265
01:02:11,000 --> 01:02:13,000
And I think that's super interesting.

1266
01:02:13,000 --> 01:02:17,000
And the third one is, and I mean, Jack had the same insight as well.

1267
01:02:17,000 --> 01:02:20,000
And that's why I showed the vibe line to everyone.

1268
01:02:20,000 --> 01:02:21,000
You don't want to type.

1269
01:02:21,000 --> 01:02:22,000
You never want to type.

1270
01:02:22,000 --> 01:02:23,000
I agree.

1271
01:02:23,000 --> 01:02:24,000
You just want to talk.

1272
01:02:24,000 --> 01:02:25,000
Yeah.

1273
01:02:25,000 --> 01:02:30,000
Actually, like if you're interacting with agents and if the development process is dialogical in the first place,

1274
01:02:30,000 --> 01:02:31,000
take a microphone.

1275
01:02:31,000 --> 01:02:33,000
Every, you have a microphone in your pocket right now.

1276
01:02:33,000 --> 01:02:34,000
It's your phone.

1277
01:02:34,000 --> 01:02:36,000
It's in the name.

1278
01:02:36,000 --> 01:02:37,000
You know?

1279
01:02:37,000 --> 01:02:38,000
It's a small thing.

1280
01:02:38,000 --> 01:02:39,000
A small phone.

1281
01:02:39,000 --> 01:02:40,000
Microphone.

1282
01:02:40,000 --> 01:02:41,000
You have it in your pocket.

1283
01:02:41,000 --> 01:02:42,000
It's the first phone.

1284
01:02:42,000 --> 01:02:43,000
You know?

1285
01:02:43,000 --> 01:02:45,000
And you just, you just talk into it.

1286
01:02:45,000 --> 01:02:49,000
So instead of typing it out and so on, you just, you know, like text to speech, whatever.

1287
01:02:49,000 --> 01:02:50,000
And you just, you just, you just go.

1288
01:02:50,000 --> 01:02:54,000
And of course, you know, what we, I think we never explained what the vibe line is actually.

1289
01:02:54,000 --> 01:02:56,000
But it's basically, that's all it does.

1290
01:02:56,000 --> 01:02:58,000
It's just, you, you, you record an MP3 file.

1291
01:02:58,000 --> 01:03:01,000
You send it to the computer automatically.

1292
01:03:01,000 --> 01:03:04,000
It gets transcribed automatically because the thing watches a folder.

1293
01:03:04,000 --> 01:03:06,000
And then it uses the transcription as the prompt.

1294
01:03:06,000 --> 01:03:07,000
That's all it does, basically.

1295
01:03:07,000 --> 01:03:08,000
And that's the right idea.

1296
01:03:08,000 --> 01:03:09,000
And you can do that on Nostra as well.

1297
01:03:09,000 --> 01:03:14,000
So I think, you know, very soon we'll have a Nostra client where you can just, you know,

1298
01:03:14,000 --> 01:03:18,000
it will summarize for you, talk to you, you're driving your car, whatever you're watching through the card.

1299
01:03:18,000 --> 01:03:26,000
Yeah, so I'm building one of the clients that I'm building for tenants is, is like a, like a phone up, like a native phone up.

1300
01:03:26,000 --> 01:03:29,000
And the first thing I added was the microphone work.

1301
01:03:29,000 --> 01:03:35,000
And that's the, the interaction that I want is that is the microphone thing.

1302
01:03:35,000 --> 01:03:40,000
I mean, I've been using the vibe line a lot and what you say is different.

1303
01:03:40,000 --> 01:03:41,000
Yeah.

1304
01:03:41,000 --> 01:03:43,000
If you were to type it, it's a different thing.

1305
01:03:43,000 --> 01:03:45,000
You were going to output something different.

1306
01:03:45,000 --> 01:03:46,000
I mean, the medium is the message, right?

1307
01:03:46,000 --> 01:03:47,000
Yeah, yeah, yeah.

1308
01:03:47,000 --> 01:03:51,000
No, and it's also your, your, again, like I think that's my rubber ducty bagging works.

1309
01:03:51,000 --> 01:03:58,000
That's my talking to someone about hard stuff actually works because you're thinking, you're speaking and thinking at the same time.

1310
01:03:58,000 --> 01:04:01,000
Like you, you don't know what the next word is going to be.

1311
01:04:01,000 --> 01:04:04,000
And you're like, you're like, yeah, you're predicting the next word.

1312
01:04:04,000 --> 01:04:10,000
That's why the hesitation, like what I was mentioning, that the hesitation part is part of what you're saying.

1313
01:04:10,000 --> 01:04:14,000
Like, I am, I'm thinking about what, what I'm going to say.

1314
01:04:14,000 --> 01:04:15,000
I'm not so sure.

1315
01:04:15,000 --> 01:04:16,000
Yeah.

1316
01:04:16,000 --> 01:04:19,000
And that, that is an important part of the message, right?

1317
01:04:19,000 --> 01:04:20,000
Yeah.

1318
01:04:20,000 --> 01:04:21,000
And writing is very different.

1319
01:04:21,000 --> 01:04:23,000
Writing is also thinking, but it's a different kind.

1320
01:04:23,000 --> 01:04:30,000
So I think we'll have some things where you will resort to writing and we'll have some things where you can just talk.

1321
01:04:30,000 --> 01:04:35,000
And I think, if the thing is already living and already going, I think talking is easiest.

1322
01:04:35,000 --> 01:04:41,000
I think actually that's why a lot of successful people and like also in companies, like you need to have the meaning.

1323
01:04:41,000 --> 01:04:42,000
It cannot be an email.

1324
01:04:42,000 --> 01:04:43,000
You know what I mean?

1325
01:04:43,000 --> 01:04:47,000
Like you know the meme where it's like, this meeting should have been an email and so on.

1326
01:04:47,000 --> 01:04:50,000
I mean, everyone kind of knows this and so on.

1327
01:04:50,000 --> 01:04:55,000
And for many, many meetings, this is true, but for many, like especially strategic things, it's absolutely not true.

1328
01:04:55,000 --> 01:05:02,000
Because you need to figure it out in the fucking room with the other people and by talking about it for like 90 minutes at least.

1329
01:05:02,000 --> 01:05:05,000
Because no one really knows what to do and you have to hash it out.

1330
01:05:05,000 --> 01:05:08,000
Which going back to the hierarchy and the slavery and all that.

1331
01:05:08,000 --> 01:05:16,000
Interestingly enough, I think the execution layers are much more easily conveyed via writing.

1332
01:05:16,000 --> 01:05:24,000
And the thinking layer, the higher you are on the pyramid, the thinking and the talking and the discussing on voice makes more sense.

1333
01:05:24,000 --> 01:05:25,000
Yeah, yeah.

1334
01:05:25,000 --> 01:05:34,000
Because the higher you are, the higher on top you are, the more creativity you need, the more less defined the problem spaces.

1335
01:05:34,000 --> 01:05:35,000
That's the issue.

1336
01:05:35,000 --> 01:05:41,000
The more you specialize, the problems are obvious.

1337
01:05:41,000 --> 01:05:42,000
Yeah, readily available.

1338
01:05:42,000 --> 01:05:47,000
Here is the vault and here is where it needs to go and here is how you screw it on and that's it.

1339
01:05:47,000 --> 01:05:48,000
There's a syntax error.

1340
01:05:48,000 --> 01:05:49,000
Yeah, it's very clear.

1341
01:05:49,000 --> 01:05:50,000
It's one line.

1342
01:05:50,000 --> 01:05:54,000
And the higher up you are, the more kind of undefined the whole problem spaces.

1343
01:05:54,000 --> 01:05:57,000
You literally, you never know what to do.

1344
01:05:57,000 --> 01:06:00,000
Like as it's, that's life.

1345
01:06:00,000 --> 01:06:01,000
Yeah.

1346
01:06:01,000 --> 01:06:02,000
It's like what to do next?

1347
01:06:02,000 --> 01:06:04,000
Most of the time it's undefined.

1348
01:06:04,000 --> 01:06:05,000
It's your choice.

1349
01:06:05,000 --> 01:06:07,000
So you have to make it up as you go along.

1350
01:06:07,000 --> 01:06:09,000
That's where talking comes in.

1351
01:06:09,000 --> 01:06:10,000
Back to the logos.

1352
01:06:10,000 --> 01:06:13,000
You know, like I always come back to the same things.

1353
01:06:13,000 --> 01:06:16,000
They figured it out like two and a half thousand years ago.

1354
01:06:16,000 --> 01:06:17,000
Amazing.

1355
01:06:17,000 --> 01:06:18,000
Yeah.

1356
01:06:18,000 --> 01:06:19,000
But anyway.

1357
01:06:19,000 --> 01:06:21,000
We should also talk about summer engineering.

1358
01:06:21,000 --> 01:06:22,000
Yeah.

1359
01:06:22,000 --> 01:06:23,000
Yeah.

1360
01:06:23,000 --> 01:06:24,000
Anything else?

1361
01:06:24,000 --> 01:06:25,000
10x.

1362
01:06:25,000 --> 01:06:26,000
Oh yeah, that's the pirate ship.

1363
01:06:26,000 --> 01:06:27,000
That's nice.

1364
01:06:27,000 --> 01:06:28,000
Awesome.

1365
01:06:28,000 --> 01:06:29,000
See, the perfect.

1366
01:06:29,000 --> 01:06:30,000
Perfect.

1367
01:06:30,000 --> 01:06:31,000
Perfect segue.

1368
01:06:31,000 --> 01:06:32,000
It was just visual.

1369
01:06:32,000 --> 01:06:33,000
It's like 10x.

1370
01:06:33,000 --> 01:06:34,000
No, you know.

1371
01:06:34,000 --> 01:06:37,000
When, when, when will it like leave public alpha?

1372
01:06:37,000 --> 01:06:38,000
Two weeks.

1373
01:06:38,000 --> 01:06:39,000
Alpha will always be over.

1374
01:06:39,000 --> 01:06:41,000
We'll remain perpetual alpha.

1375
01:06:41,000 --> 01:06:42,000
Oh yeah.

1376
01:06:42,000 --> 01:06:44,000
Perpetual rewrite alpha.

1377
01:06:44,000 --> 01:06:46,000
Like every two weeks it will just restart.

1378
01:06:46,000 --> 01:06:47,000
It's just for rewriting.

1379
01:06:47,000 --> 01:06:48,000
We write everything.

1380
01:06:48,000 --> 01:06:49,000
All right.

1381
01:06:49,000 --> 01:06:50,000
Fantastic.

1382
01:06:50,000 --> 01:06:51,000
Yeah.

1383
01:06:51,000 --> 01:06:52,000
No, no.

1384
01:06:52,000 --> 01:06:55,000
No, I wanted to talk to you about, like we wanted to do this for a long time.

1385
01:06:55,000 --> 01:06:58,000
And even though we kind of, you know, live next door to each other.

1386
01:06:58,000 --> 01:07:01,000
We never see each other because you're always in your, you know,

1387
01:07:01,000 --> 01:07:06,000
little room typing instead of outside walking around talking to the machines.

1388
01:07:06,000 --> 01:07:08,000
Like I told you, you should not anyway.

1389
01:07:08,000 --> 01:07:09,000
Everyone has their flaws.

1390
01:07:09,000 --> 01:07:12,000
But like a little slave typing away.

1391
01:07:12,000 --> 01:07:13,000
Yeah.

1392
01:07:13,000 --> 01:07:14,000
What are you in the LLM?

1393
01:07:14,000 --> 01:07:15,000
It's not bad.

1394
01:07:15,000 --> 01:07:16,000
It's bad.

1395
01:07:16,000 --> 01:07:17,000
Let the computers do the typing.

1396
01:07:17,000 --> 01:07:22,000
No, but we should, we should, I mean.

1397
01:07:22,000 --> 01:07:24,000
I just mentioned everyone has their flaws.

1398
01:07:24,000 --> 01:07:30,000
I think we are painfully aware of our flaws in some sense.

1399
01:07:30,000 --> 01:07:31,000
At least some of them.

1400
01:07:31,000 --> 01:07:36,000
On the summer and engineering side because we're really bad when it comes to talking about it.

1401
01:07:36,000 --> 01:07:39,000
Because we're either doing it or not doing it.

1402
01:07:39,000 --> 01:07:43,000
And when we're doing it, there's no time to talk about it.

1403
01:07:43,000 --> 01:07:46,000
And it's like we're making everything up.

1404
01:07:46,000 --> 01:07:48,000
It's a very intense couple of weeks.

1405
01:07:48,000 --> 01:07:51,000
And then we're just recovering and don't want to see each other.

1406
01:07:51,000 --> 01:07:58,000
And it's like everyone has like complete, you know, all the introverts, all the introverts go back in their cave.

1407
01:07:58,000 --> 01:08:01,000
Like I don't want to see humans for like five months, please.

1408
01:08:01,000 --> 01:08:03,000
And it was very understandable.

1409
01:08:03,000 --> 01:08:05,000
So I'm strongly in that camp.

1410
01:08:05,000 --> 01:08:11,000
But we wanted to do this for a long time, kind of like explain what sort of an engineering is.

1411
01:08:11,000 --> 01:08:16,000
Kind of tell people what to expect in some way.

1412
01:08:16,000 --> 01:08:19,000
Kind of discuss what we figured out along the way.

1413
01:08:19,000 --> 01:08:22,000
Because when we did the first one, like over a year ago now.

1414
01:08:22,000 --> 01:08:23,000
One and a half years ago.

1415
01:08:23,000 --> 01:08:24,000
Yeah.

1416
01:08:24,000 --> 01:08:25,000
Yeah.

1417
01:08:25,000 --> 01:08:26,000
Oh my lord.

1418
01:08:26,000 --> 01:08:30,000
And all the half the stuff that didn't exist.

1419
01:08:30,000 --> 01:08:31,000
That's insane.

1420
01:08:31,000 --> 01:08:32,000
Like it all came out.

1421
01:08:32,000 --> 01:08:34,000
Lots of cool things came out of the first one.

1422
01:08:34,000 --> 01:08:36,000
And that's also something we should talk about.

1423
01:08:36,000 --> 01:08:38,000
Like what came out of it so far?

1424
01:08:38,000 --> 01:08:40,000
What did we learn along the way?

1425
01:08:40,000 --> 01:08:41,000
What kind of visit now?

1426
01:08:41,000 --> 01:08:44,000
And it's still going to evolve most likely.

1427
01:08:44,000 --> 01:08:51,000
And like if you want to come and, you know, like join this kind of conversation that we just had and explore all this

1428
01:08:51,000 --> 01:08:52,000
vibe-coding business.

1429
01:08:52,000 --> 01:08:57,000
And like the find out why Noster is the perfect substrate for all of that.

1430
01:08:57,000 --> 01:08:58,000
And what makes it all work.

1431
01:08:58,000 --> 01:09:00,000
And what the ethos is behind it.

1432
01:09:00,000 --> 01:09:03,000
What the value structure is behind the whole stuff.

1433
01:09:03,000 --> 01:09:05,000
Like what to expect?

1434
01:09:05,000 --> 01:09:06,000
Why should you come?

1435
01:09:06,000 --> 01:09:09,000
Or not even why should you come?

1436
01:09:09,000 --> 01:09:15,000
We have to explain it so that the people that aren't interested in this stuff understand it and might come on their own.

1437
01:09:15,000 --> 01:09:17,000
Because I think most people should not come.

1438
01:09:18,000 --> 01:09:20,000
100%.

1439
01:09:20,000 --> 01:09:25,000
Like if all that stuff kind of sounds like nonsense and it's not value aligned with your own world,

1440
01:09:25,000 --> 01:09:27,000
you like please do not come.

1441
01:09:27,000 --> 01:09:29,000
And no circumstances at all.

1442
01:09:29,000 --> 01:09:32,000
If you're trying to make, I don't know.

1443
01:09:32,000 --> 01:09:33,000
We've been there.

1444
01:09:33,000 --> 01:09:36,000
If you're a TikTok engineer and you want to make the stuff more addicting,

1445
01:09:36,000 --> 01:09:39,000
or if you're like try to make the Google search like a millisecond faster,

1446
01:09:39,000 --> 01:09:44,000
or you, you know, like believe the lies that Sam, not Sam Bank from the field.

1447
01:09:44,000 --> 01:09:45,000
Who's the other guy?

1448
01:09:45,000 --> 01:09:46,000
Yeah, exactly.

1449
01:09:46,000 --> 01:09:50,000
If you think like the eye overlords will show in the singularity and everyone is going to live forever,

1450
01:09:50,000 --> 01:09:51,000
like please don't come.

1451
01:09:51,000 --> 01:09:55,000
Like we're not going to, you're not going to have a lot of fun with us.

1452
01:09:55,000 --> 01:09:58,000
But if you're like e-cash, you're like no sir, you know, you like my coding,

1453
01:09:58,000 --> 01:09:59,000
you're going to have a lot of fun.

1454
01:09:59,000 --> 01:10:01,000
You're going to have a lot of ADHD.

1455
01:10:01,000 --> 01:10:02,000
Yeah, yeah.

1456
01:10:02,000 --> 01:10:06,000
Should we do maybe like the time or something or like let's try.

1457
01:10:06,000 --> 01:10:07,000
Let me see.

1458
01:10:07,000 --> 01:10:08,000
Let me see how.

1459
01:10:08,000 --> 01:10:10,000
If it's recording, I don't.

1460
01:10:10,000 --> 01:10:12,000
That would be, that would be a shame.

1461
01:10:12,000 --> 01:10:13,000
No, no, we're good.

1462
01:10:13,000 --> 01:10:18,000
I just don't want a die of heatstroke while we're trying to explain what summer engineering is.

1463
01:10:18,000 --> 01:10:19,000
But I think we'll manage.

1464
01:10:19,000 --> 01:10:24,000
And I think it's going to be a nice 10 minute adventure, you know, like Rick and Morty in and out 10 minutes.

1465
01:10:24,000 --> 01:10:25,000
Very easy.

1466
01:10:25,000 --> 01:10:27,000
So how did it all start?

1467
01:10:27,000 --> 01:10:28,000
Let's start at the beginning.

1468
01:10:28,000 --> 01:10:31,000
Okay, so it started.

1469
01:10:31,000 --> 01:10:37,000
The idea was to do kind of like an academy with almost two years ago.

1470
01:10:37,000 --> 01:10:38,000
Yeah, it was.

1471
01:10:38,000 --> 01:10:40,000
I think it was even over two years ago.

1472
01:10:40,000 --> 01:10:43,000
Yeah, you moved to Madera basically or wearing the process.

1473
01:10:43,000 --> 01:10:44,000
And me as well.

1474
01:10:44,000 --> 01:10:45,000
You as well.

1475
01:10:45,000 --> 01:10:49,000
And some other bitcoins were around and Marta and actually had the idea to do like a.

1476
01:10:49,000 --> 01:10:54,000
Yeah, Marta and Andrew talked about doing something and we there a bit kind of academy.

1477
01:10:54,000 --> 01:10:55,000
First made the t-shirts.

1478
01:10:55,000 --> 01:10:56,000
Yeah.

1479
01:10:56,000 --> 01:10:57,000
That said, Madera is going to cut them.

1480
01:10:57,000 --> 01:10:58,000
Yeah.

1481
01:10:58,000 --> 01:10:59,000
Then we put it to the way.

1482
01:10:59,000 --> 01:11:00,000
Yeah.

1483
01:11:00,000 --> 01:11:06,000
And then we put it to the way so many times that we basically rock the people that worry about.

1484
01:11:06,000 --> 01:11:09,000
At the beginning, Marta had a very different idea.

1485
01:11:09,000 --> 01:11:10,000
But we.

1486
01:11:10,000 --> 01:11:13,000
I think the first question that we asked actually a master.

1487
01:11:13,000 --> 01:11:14,000
I think, or was it still on Twitter?

1488
01:11:14,000 --> 01:11:15,000
I don't know.

1489
01:11:15,000 --> 01:11:16,000
It might have still been on Twitter.

1490
01:11:16,000 --> 01:11:17,000
No way.

1491
01:11:17,000 --> 01:11:18,000
No, it was like.

1492
01:11:18,000 --> 01:11:19,000
Oh, yeah, we did.

1493
01:11:19,000 --> 01:11:20,000
I think it was.

1494
01:11:20,000 --> 01:11:24,000
It was like, if you would want to participate in like something like a bootcamp,

1495
01:11:24,000 --> 01:11:27,000
would you should it focus on BDK?

1496
01:11:27,000 --> 01:11:30,000
Should it focus on LDK or should it focus on NDK?

1497
01:11:30,000 --> 01:11:32,000
It was like a Bitcoin Lightning Noster.

1498
01:11:32,000 --> 01:11:33,000
Yeah.

1499
01:11:33,000 --> 01:11:34,000
What was the most?

1500
01:11:34,000 --> 01:11:38,000
Like two years ago, what was the interest landscape like?

1501
01:11:38,000 --> 01:11:41,000
It was very heavy on Lightning Noster, interestingly.

1502
01:11:41,000 --> 01:11:42,000
Yeah.

1503
01:11:42,000 --> 01:11:43,000
No one was interested in BDK.

1504
01:11:43,000 --> 01:11:44,000
Sorry, BDK people.

1505
01:11:44,000 --> 01:11:45,000
It was like a Bitcoin.

1506
01:11:45,000 --> 01:11:46,000
The money is fixed.

1507
01:11:46,000 --> 01:11:47,000
It's done.

1508
01:11:47,000 --> 01:11:48,000
It's done.

1509
01:11:48,000 --> 01:11:49,000
No one needs to BDK anymore.

1510
01:11:49,000 --> 01:11:50,000
We can all supply right now.

1511
01:11:50,000 --> 01:11:51,000
Let's leave the up return.

1512
01:11:51,000 --> 01:11:52,000
Everyone understands it.

1513
01:11:52,000 --> 01:11:53,000
Everyone understands it.

1514
01:11:53,000 --> 01:11:55,000
It's like a shining city in cyberspace.

1515
01:11:55,000 --> 01:11:56,000
It's already built.

1516
01:11:56,000 --> 01:11:58,000
It's got a glow forever.

1517
01:11:58,000 --> 01:11:59,000
Just put all your turn.

1518
01:11:59,000 --> 01:12:00,000
It's amazing.

1519
01:12:00,000 --> 01:12:02,000
Just buy all the stocks.

1520
01:12:02,000 --> 01:12:03,000
You'll be rich forever.

1521
01:12:03,000 --> 01:12:06,000
It's going to literally go up forever.

1522
01:12:06,000 --> 01:12:07,000
Yeah.

1523
01:12:07,000 --> 01:12:08,000
Anyway.

1524
01:12:08,000 --> 01:12:09,000
OK.

1525
01:12:09,000 --> 01:12:11,000
So BDK, we kind of crossed off the list.

1526
01:12:11,000 --> 01:12:13,000
And then we all premised here.

1527
01:12:13,000 --> 01:12:20,000
And that's what I think when we wrote the majority of what's on the summer engineering website.

1528
01:12:20,000 --> 01:12:21,000
Which is a little bit dated.

1529
01:12:21,000 --> 01:12:23,000
I'm still happy that we did static HTML.

1530
01:12:23,000 --> 01:12:25,000
I'm very happy about that.

1531
01:12:25,000 --> 01:12:29,000
So I zero maintenance, which is what I did.

1532
01:12:29,000 --> 01:12:32,000
So we focused on Lightning Noster.

1533
01:12:32,000 --> 01:12:35,000
And the idea was, OK, the money is fixed.

1534
01:12:35,000 --> 01:12:39,000
And it's kind of fixed the money, fixed the world kind of idea.

1535
01:12:39,000 --> 01:12:40,000
On the side of the world.

1536
01:12:40,000 --> 01:12:41,000
Yeah.

1537
01:12:41,000 --> 01:12:46,000
And now with the money fixed, we don't want to fix the world immediately.

1538
01:12:46,000 --> 01:12:47,000
It's a big task.

1539
01:12:47,000 --> 01:12:48,000
We want to fix the internet.

1540
01:12:48,000 --> 01:12:51,000
So it's fixed the money, money is fixed, fixed the internet.

1541
01:12:51,000 --> 01:12:55,000
And Noster had a big upswing at the time.

1542
01:12:55,000 --> 01:12:58,000
And I mean, I still have the same opinion.

1543
01:12:58,000 --> 01:13:01,000
But it's like, we're going to rebuild everything on Noster.

1544
01:13:01,000 --> 01:13:06,000
And hopefully, I'm not sure if we managed to do everything exactly right.

1545
01:13:06,000 --> 01:13:10,000
But I hope that we can learn from mistakes of the past and do a better iteration of the internet.

1546
01:13:10,000 --> 01:13:14,000
Like a literally a version two of the internet where Noster is at the base.

1547
01:13:14,000 --> 01:13:16,000
Which means cryptographic identity is at the base.

1548
01:13:16,000 --> 01:13:18,000
Which means it is permissionless by default.

1549
01:13:18,000 --> 01:13:22,000
Which means you don't need to ask someone, can I sign up to this thing?

1550
01:13:22,000 --> 01:13:23,000
But you're just there.

1551
01:13:23,000 --> 01:13:24,000
Just like in Bitcoin.

1552
01:13:24,000 --> 01:13:26,000
You just generate a key and that's it.

1553
01:13:26,000 --> 01:13:27,000
And you'll have money baked in.

1554
01:13:27,000 --> 01:13:29,000
That you have set for real, not just with the IDs.

1555
01:13:29,000 --> 01:13:30,000
Yes.

1556
01:13:30,000 --> 01:13:35,000
Actual cryptographic identities that you can self generate.

1557
01:13:35,000 --> 01:13:40,000
And you hold your own keys, which puts the user, the individual into the center of things.

1558
01:13:40,000 --> 01:13:45,000
Unless someone builds like this idea of putting like an insect bunker idea.

1559
01:13:45,000 --> 01:13:48,000
If I ever meet a guy, I'm going to punch him in the face.

1560
01:13:48,000 --> 01:13:51,000
And I'm saying, oh my god.

1561
01:13:51,000 --> 01:13:53,000
Anyway, touch the stove.

1562
01:13:53,000 --> 01:13:54,000
Touch the stove.

1563
01:13:54,000 --> 01:13:56,000
You know, you live and learn.

1564
01:13:56,000 --> 01:13:57,000
You live and know.

1565
01:13:57,000 --> 01:13:58,000
It's okay.

1566
01:13:58,000 --> 01:14:01,000
There was a lesson you weren't published on you.

1567
01:14:01,000 --> 01:14:02,000
Yeah.

1568
01:14:02,000 --> 01:14:05,000
Maybe before we hit the waves, let's turn around together.

1569
01:14:05,000 --> 01:14:08,000
I've raped enough years already.

1570
01:14:08,000 --> 01:14:10,000
It's a slavery rape.

1571
01:14:10,000 --> 01:14:11,000
We already have all that.

1572
01:14:11,000 --> 01:14:13,000
You know, that's why you shouldn't rely on the transcript.

1573
01:14:13,000 --> 01:14:16,000
It's like if you go just on the transcript.

1574
01:14:16,000 --> 01:14:19,000
And just like when it's summarized, it's like a slavery.

1575
01:14:19,000 --> 01:14:20,000
It's a good thing we've done.

1576
01:14:20,000 --> 01:14:21,000
Bonitas.

1577
01:14:21,000 --> 01:14:22,000
Yeah.

1578
01:14:22,000 --> 01:14:23,000
You too.

1579
01:14:23,000 --> 01:14:24,000
There's no one to say.

1580
01:14:24,000 --> 01:14:26,000
That's actually a big part of sovereign engineering.

1581
01:14:26,000 --> 01:14:28,000
No monetization whatsoever.

1582
01:14:28,000 --> 01:14:32,000
So the idea was we want to bring everyone in the same place.

1583
01:14:32,000 --> 01:14:33,000
Poverty.

1584
01:14:33,000 --> 01:14:34,000
Exactly.

1585
01:14:34,000 --> 01:14:35,000
Embrace it.

1586
01:14:35,000 --> 01:14:36,000
Confirmation.

1587
01:14:36,000 --> 01:14:37,000
It's a value for poverty.

1588
01:14:37,000 --> 01:14:38,000
Shared poverty.

1589
01:14:38,000 --> 01:14:39,000
Yeah.

1590
01:14:39,000 --> 01:14:43,000
It's someone made the joke.

1591
01:14:43,000 --> 01:14:45,000
It shouldn't be called value for value.

1592
01:14:45,000 --> 01:14:48,000
It should be called a poor for poor or something like that.

1593
01:14:48,000 --> 01:14:51,000
21 sets circling around.

1594
01:14:51,000 --> 01:14:54,000
But the idea was no monetary incentives.

1595
01:14:54,000 --> 01:14:58,000
Like the whole, the whole, and I think that's still very much true.

1596
01:14:58,000 --> 01:15:00,000
And we're not at the end of this.

1597
01:15:00,000 --> 01:15:05,000
The whole idea of sovereign engineering is to do a massive exploration of the

1598
01:15:05,000 --> 01:15:10,000
Noster space of literally cyberspace that is opened up.

1599
01:15:10,000 --> 01:15:12,000
See, I can't stop quoting sailor.

1600
01:15:12,000 --> 01:15:16,000
Like it's literally a cyberspace is wide open.

1601
01:15:16,000 --> 01:15:18,000
And we have cryptographic identities.

1602
01:15:18,000 --> 01:15:20,000
We have money that actually works.

1603
01:15:20,000 --> 01:15:21,000
We have lightning.

1604
01:15:21,000 --> 01:15:22,000
We have eCache.

1605
01:15:22,000 --> 01:15:23,000
All this stuff works.

1606
01:15:23,000 --> 01:15:24,000
What can we build with it?

1607
01:15:24,000 --> 01:15:30,000
And I think, or we both had the conclusion kind of that any type of monetary incentives

1608
01:15:30,000 --> 01:15:32,000
With kind of screw up the exploratory process.

1609
01:15:32,000 --> 01:15:38,000
If you think about monetizing too early, it just like it gives the weird incentives.

1610
01:15:38,000 --> 01:15:42,000
And so no one is going to get a stamp of approval that you participated.

1611
01:15:42,000 --> 01:15:45,000
No one is going to get like a VC money check or anything.

1612
01:15:45,000 --> 01:15:47,000
No one is going to get any money just for participating.

1613
01:15:47,000 --> 01:15:50,000
You participate because you want to participate.

1614
01:15:50,000 --> 01:15:52,000
And you can hang out with cool people for six weeks.

1615
01:15:52,000 --> 01:15:57,000
And literally discuss any crazy idea without judgment whatsoever.

1616
01:15:57,000 --> 01:16:01,000
Unless it's related to shit-consciousness, you judge you harshly.

1617
01:16:01,000 --> 01:16:02,000
Who am I kidding?

1618
01:16:02,000 --> 01:16:03,000
There's going to be lots of judgment.

1619
01:16:03,000 --> 01:16:06,000
But you still can't discuss absolutely everything.

1620
01:16:06,000 --> 01:16:08,000
I mean, the judgment is the point.

1621
01:16:08,000 --> 01:16:09,000
The judgment is the point.

1622
01:16:09,000 --> 01:16:11,000
Actually, that's a great point.

1623
01:16:11,000 --> 01:16:16,000
So there's going to be collective shaming and judgment at the end of every week.

1624
01:16:16,000 --> 01:16:20,000
And that's kind of the structure that we landed on is like a weekly structure.

1625
01:16:20,000 --> 01:16:23,000
Where it's like every week is structured in the same way.

1626
01:16:23,000 --> 01:16:25,000
Monday is like a little bit of orientation.

1627
01:16:25,000 --> 01:16:29,000
Tuesday is talks where people can just talk about things that they find interesting

1628
01:16:29,000 --> 01:16:30,000
or they discovered and so on.

1629
01:16:30,000 --> 01:16:32,000
Wednesday is workshops.

1630
01:16:32,000 --> 01:16:39,000
So if you want to show off how to, I don't know, like set up a bit X with like e-cash,

1631
01:16:39,000 --> 01:16:41,000
hash, hash, payouts, whatever.

1632
01:16:41,000 --> 01:16:44,000
You can do that and show that off and how to do that and so on.

1633
01:16:44,000 --> 01:16:48,000
Speaking of which, not to the rail, but ever since I heard about the...

1634
01:16:48,000 --> 01:16:50,000
It's whatever it's called.

1635
01:16:50,000 --> 01:16:51,000
No rail.

1636
01:16:51,000 --> 01:16:54,000
There's only the rail.

1637
01:16:54,000 --> 01:16:56,000
It's only falling.

1638
01:16:56,000 --> 01:16:57,000
Lean into the rail.

1639
01:16:57,000 --> 01:16:59,000
He'd record and we start falling.

1640
01:16:59,000 --> 01:17:01,000
We hallucinate and we the rail.

1641
01:17:01,000 --> 01:17:03,000
And that's also the whole point of sovereign engineer.

1642
01:17:03,000 --> 01:17:04,000
Ha! The first time.

1643
01:17:04,000 --> 01:17:05,000
Yeah.

1644
01:17:05,000 --> 01:17:07,000
It's like you never know where it goes.

1645
01:17:07,000 --> 01:17:10,000
There is no more than a weekly structure.

1646
01:17:10,000 --> 01:17:11,000
There is none.

1647
01:17:11,000 --> 01:17:14,000
You never know, just didn't actually something that nicely.

1648
01:17:14,000 --> 01:17:16,000
Like what is the theme of the next one?

1649
01:17:16,000 --> 01:17:19,000
It's like you find out the theme during the course.

1650
01:17:19,000 --> 01:17:20,000
You never know.

1651
01:17:20,000 --> 01:17:21,000
Which is true.

1652
01:17:21,000 --> 01:17:22,000
How do we say true?

1653
01:17:22,000 --> 01:17:23,000
Yeah.

1654
01:17:23,000 --> 01:17:24,000
Sorry, I derailed.

1655
01:17:24,000 --> 01:17:30,000
No, ever since I saw that idea of the e-cash mining chairs,

1656
01:17:30,000 --> 01:17:34,000
I want to do like nips 61 subs of that.

1657
01:17:34,000 --> 01:17:35,000
All right.

1658
01:17:35,000 --> 01:17:36,000
Can you watch it?

1659
01:17:36,000 --> 01:17:37,000
Of the payouts.

1660
01:17:37,000 --> 01:17:38,000
Okay.

1661
01:17:38,000 --> 01:17:39,000
Yeah.

1662
01:17:39,000 --> 01:17:40,000
Yeah.

1663
01:17:40,000 --> 01:17:49,000
So anyway, the reason why we continue to do it is because it really, really works.

1664
01:17:49,000 --> 01:17:53,000
So we did eight weeks in the beginning.

1665
01:17:53,000 --> 01:17:55,000
We're doing six weeks now.

1666
01:17:55,000 --> 01:17:58,000
Eight weeks was too long because it's super fucking intense.

1667
01:17:58,000 --> 01:18:00,000
So just if you come, know that.

1668
01:18:00,000 --> 01:18:01,000
It's going to be.

1669
01:18:01,000 --> 01:18:02,000
It's going to help.

1670
01:18:02,000 --> 01:18:03,000
It's going to be.

1671
01:18:03,000 --> 01:18:05,000
It's going to be a lot.

1672
01:18:05,000 --> 01:18:10,000
And the reason is because we're always aiming for 21 participants there about.

1673
01:18:10,000 --> 01:18:15,000
And there's really motivated, really smart people coming.

1674
01:18:15,000 --> 01:18:16,000
Always.

1675
01:18:16,000 --> 01:18:19,000
And everyone knows something that you don't.

1676
01:18:19,000 --> 01:18:24,000
That's kind of the whole idea of bringing people together in an intense environment for a long time.

1677
01:18:24,000 --> 01:18:28,000
It's just like a massive learning and experimentation opportunity.

1678
01:18:28,000 --> 01:18:33,000
And the best stuff that we saw so far comes from cross pollination.

1679
01:18:33,000 --> 01:18:35,000
Like literally cross breeding ideas.

1680
01:18:35,000 --> 01:18:38,000
And like someone knows a lot about, I don't know, match networks.

1681
01:18:38,000 --> 01:18:39,000
Someone knows a lot about e-cache.

1682
01:18:39,000 --> 01:18:41,000
Okay, bringing them together, you have toll gate.

1683
01:18:41,000 --> 01:18:42,000
Yeah.

1684
01:18:42,000 --> 01:18:46,000
So that's basically what came out of the, I don't know, was it the second one or the third one?

1685
01:18:46,000 --> 01:18:50,000
It was one of the awesome things that came out of sovereign engineering.

1686
01:18:50,000 --> 01:18:53,000
I should have everything in top of mind, but I absolutely don't.

1687
01:18:53,000 --> 01:18:56,000
It's kind of brained massively.

1688
01:18:56,000 --> 01:18:58,000
I would be walking down under the sun.

1689
01:18:58,000 --> 01:19:00,000
Very close to heatstroke again.

1690
01:19:00,000 --> 01:19:03,000
But that's where the creative chooses flow most.

1691
01:19:03,000 --> 01:19:11,000
I think a part of it is also like a lot of the ingredients is the fact that it's just kind of chaos.

1692
01:19:11,000 --> 01:19:14,000
So someone is thinking about one thing this week.

1693
01:19:14,000 --> 01:19:18,000
It's not that they care the structure chaos.

1694
01:19:18,000 --> 01:19:19,000
Chaos within bounds.

1695
01:19:19,000 --> 01:19:23,000
Chaos within bounds, because the goal is everyone comes together.

1696
01:19:23,000 --> 01:19:25,000
And we also, like we do what we do now.

1697
01:19:25,000 --> 01:19:26,000
We go on a lot of walks, for example.

1698
01:19:26,000 --> 01:19:29,000
So at least once a week we go on a longer walk.

1699
01:19:29,000 --> 01:19:31,000
Everyone just can't talk about anything.

1700
01:19:31,000 --> 01:19:34,000
And that's where I think the best ideas appear.

1701
01:19:34,000 --> 01:19:39,000
Where it's just like you're talking about whatever idea you had for the last 10 years.

1702
01:19:39,000 --> 01:19:41,000
And you had to splinter in your mind forever.

1703
01:19:41,000 --> 01:19:45,000
And then you just talk about it and talk about it and someone comes up with a solution.

1704
01:19:45,000 --> 01:19:53,000
Yeah, and the thing I think why the walk part works so well is because you cannot be distracted by implementation details.

1705
01:19:53,000 --> 01:19:56,000
Or by looking some specific detail out.

1706
01:19:56,000 --> 01:20:05,000
You are kind of forced to talk more in philosophical terms, in value-based terms, about what to build on the Y.

1707
01:20:05,000 --> 01:20:09,000
And because of the medium, you cannot look at the how.

1708
01:20:09,000 --> 01:20:13,000
And that is kind of, there is a very important space for the how.

1709
01:20:13,000 --> 01:20:18,000
But it's downstream from what is pointing you in that direction.

1710
01:20:18,000 --> 01:20:22,000
It's very much like the best writers of history they always mentioned.

1711
01:20:22,000 --> 01:20:24,000
There's a time to write and there's a time to edit.

1712
01:20:24,000 --> 01:20:28,000
Where it's like you and I think it was Hemingway, like a bright drunk edit sober.

1713
01:20:28,000 --> 01:20:29,000
Just get it all out.

1714
01:20:29,000 --> 01:20:31,000
And then the next day, like sleep on it and then you edit.

1715
01:20:31,000 --> 01:20:34,000
And that's what we do with summer engineering as well.

1716
01:20:34,000 --> 01:20:36,000
Where it's like we get together, we have these crazy ideas.

1717
01:20:36,000 --> 01:20:41,000
And then everyone in their own time or also like in groups we get together and we try to implement the prototype.

1718
01:20:41,000 --> 01:20:44,000
Because Friday, Friday is the most important day.

1719
01:20:44,000 --> 01:20:48,000
So most of the implementation happens like at night or like whatever people have time.

1720
01:20:48,000 --> 01:20:49,000
Or in third time.

1721
01:20:49,000 --> 01:20:50,000
Like in your spare time.

1722
01:20:50,000 --> 01:20:53,000
And on Friday we all meet and Friday is demo day.

1723
01:20:53,000 --> 01:20:55,000
Friday is the to me the most important day.

1724
01:20:55,000 --> 01:20:59,000
Because Friday everyone shows off kind of what whatever experiment,

1725
01:20:59,000 --> 01:21:01,000
whatever thing they built this week.

1726
01:21:01,000 --> 01:21:03,000
And so that's where the judgment comes in.

1727
01:21:03,000 --> 01:21:04,000
Like that.

1728
01:21:04,000 --> 01:21:05,000
We'll have to consider it.

1729
01:21:05,000 --> 01:21:06,000
And it's very.

1730
01:21:06,000 --> 01:21:15,000
No, it's actually amazing what came out of this kind of system thing, whatever.

1731
01:21:15,000 --> 01:21:18,000
Like what came out of the cohorts is like amazing.

1732
01:21:18,000 --> 01:21:22,000
So to me the amount of exploration we've seen.

1733
01:21:22,000 --> 01:21:29,000
I mean we both have been lucky enough to be at every single demo day essentially.

1734
01:21:29,000 --> 01:21:30,000
Yeah.

1735
01:21:30,000 --> 01:21:35,000
And how many different fucking wild and crazy stupid.

1736
01:21:35,000 --> 01:21:38,000
And you have 20 motivated highly skilled people like people like.

1737
01:21:38,000 --> 01:21:40,000
How we shoot at the wall.

1738
01:21:40,000 --> 01:21:43,000
People like Kelly, people like Hazard, people like Justin Moon.

1739
01:21:43,000 --> 01:21:50,000
I mean most of the people we were able to pick and go on walks and have discussions with Paul as well.

1740
01:21:50,000 --> 01:21:55,000
That just 20 of those people trying a new thing every week for six weeks.

1741
01:21:55,000 --> 01:21:56,000
A different thing.

1742
01:21:56,000 --> 01:21:57,000
Yeah.

1743
01:21:57,000 --> 01:21:58,000
A new different thing.

1744
01:21:58,000 --> 01:21:59,000
That's also something we landed on.

1745
01:21:59,000 --> 01:22:02,000
Like we really want to encourage people to try something new every week.

1746
01:22:02,000 --> 01:22:04,000
And I actually did it myself the last cohort.

1747
01:22:04,000 --> 01:22:07,000
Like I actually demoed something new every single day.

1748
01:22:07,000 --> 01:22:08,000
That's where the pipeline came from.

1749
01:22:08,000 --> 01:22:12,000
Like because I got more efficient with vibe coding.

1750
01:22:12,000 --> 01:22:18,000
And the last three demos I did I just you know like one shot spoke into the mic and then just presented.

1751
01:22:18,000 --> 01:22:23,000
I was actually you know there was some interesting stuff there potentially.

1752
01:22:23,000 --> 01:22:26,000
That someone hopefully will rebuild properly in the future.

1753
01:22:26,000 --> 01:22:32,000
Or like if I if I become a 10 extra doctor as well then maybe I'll rebuild it myself.

1754
01:22:32,000 --> 01:22:40,000
That is the whole point of the system is to get to that point where you can maintain a productive a production quality system.

1755
01:22:40,000 --> 01:22:41,000
Yeah.

1756
01:22:41,000 --> 01:22:43,000
Because it builds itself to that point.

1757
01:22:43,000 --> 01:22:47,000
And so the last one was now how many months ago, three months ago.

1758
01:22:47,000 --> 01:22:48,000
Two months ago.

1759
01:22:48,000 --> 01:22:50,000
Two months ago.

1760
01:22:50,000 --> 01:22:51,000
Something like that.

1761
01:22:51,000 --> 01:22:55,000
And it completely got taken over in the first week actually by vibe coding.

1762
01:22:55,000 --> 01:22:56,000
Yeah.

1763
01:22:56,000 --> 01:23:00,000
So we had never had a single day poll show.

1764
01:23:00,000 --> 01:23:01,000
Yeah.

1765
01:23:01,000 --> 01:23:02,000
Yeah.

1766
01:23:02,000 --> 01:23:04,000
You know what?

1767
01:23:04,000 --> 01:23:08,000
I kind of felt them.

1768
01:23:08,000 --> 01:23:13,000
So I think it was Paul during the workshop for the vibe coding stuff.

1769
01:23:13,000 --> 01:23:21,000
And I came in and I saw the whole room of people and it looked like a senior's club.

1770
01:23:21,000 --> 01:23:23,000
Where they were learning about the internet.

1771
01:23:23,000 --> 01:23:24,000
Yeah.

1772
01:23:24,000 --> 01:23:26,000
How to send an email.

1773
01:23:26,000 --> 01:23:27,000
Yeah.

1774
01:23:27,000 --> 01:23:28,000
Yeah.

1775
01:23:28,000 --> 01:23:32,000
Very few people were exposed to proper vibe coding back then.

1776
01:23:32,000 --> 01:23:36,000
And now of course you know it's absolutely everywhere.

1777
01:23:36,000 --> 01:23:39,000
We are two people that are heavily leaning into it.

1778
01:23:39,000 --> 01:23:45,000
And that's why the theme of the next one, the fifth one is going to be yellow mode.

1779
01:23:45,000 --> 01:23:52,000
Because like kind of we hope to encourage people to really experiment with all that stuff.

1780
01:23:52,000 --> 01:23:54,000
And like stop riding your own code.

1781
01:23:54,000 --> 01:23:55,000
Never look at the code.

1782
01:23:55,000 --> 01:23:56,000
Stop reading code.

1783
01:23:56,000 --> 01:23:57,000
You don't need to understand stuff.

1784
01:23:57,000 --> 01:23:58,000
Just pure vibes.

1785
01:23:58,000 --> 01:24:04,000
Just because the most important thing is going to be to figure out what to build.

1786
01:24:04,000 --> 01:24:05,000
Like the why?

1787
01:24:05,000 --> 01:24:07,000
Why build that stuff in the first place?

1788
01:24:07,000 --> 01:24:10,000
If you can't build anything, like why build the thing?

1789
01:24:10,000 --> 01:24:11,000
Yeah.

1790
01:24:11,000 --> 01:24:12,000
What makes it valuable?

1791
01:24:12,000 --> 01:24:17,000
And I think the most important part is going to be kind of the system architecture in some sense.

1792
01:24:17,000 --> 01:24:23,000
Like you will have to have some say of like how this is going to work.

1793
01:24:23,000 --> 01:24:34,000
Because just like in the Bitcoin world for example, the reason why the system is built in such a complicated weird way is because of the values it wants to represent.

1794
01:24:34,000 --> 01:24:37,000
Otherwise, like there's a million.

1795
01:24:37,000 --> 01:24:39,000
It's optimizing for the right thing.

1796
01:24:39,000 --> 01:24:40,000
Exactly.

1797
01:24:40,000 --> 01:24:42,000
Knowing what the right thing is is the thing.

1798
01:24:42,000 --> 01:24:44,000
It's optimizing for self sovereignty.

1799
01:24:44,000 --> 01:24:45,000
Yes.

1800
01:24:45,000 --> 01:24:49,000
And that's also why our whole thing is called sovereign engineering because that's what we want to optimize for.

1801
01:24:49,000 --> 01:24:55,000
We want to empower users and we want to optimize for self sovereignty in all the things that we built.

1802
01:24:55,000 --> 01:24:59,000
And that's why everything we want to build, we want to build in this period of Bitcoin basically.

1803
01:24:59,000 --> 01:25:02,000
And so just to mention some highlights that came out of like the first cohort for example.

1804
01:25:02,000 --> 01:25:05,000
Blossom came out of the first cohort.

1805
01:25:05,000 --> 01:25:07,000
So I think one of the greatest success stories.

1806
01:25:07,000 --> 01:25:09,000
Because now it's integrated everywhere.

1807
01:25:09,000 --> 01:25:11,000
I'm so happy about that.

1808
01:25:11,000 --> 01:25:13,000
Great name by the way.

1809
01:25:13,000 --> 01:25:14,000
No idea who came up with it.

1810
01:25:14,000 --> 01:25:15,000
But fantastic name.

1811
01:25:15,000 --> 01:25:17,000
Couldn't be a better name.

1812
01:25:17,000 --> 01:25:21,000
Nip16, Nip61 came out of the cohorts.

1813
01:25:21,000 --> 01:25:25,000
Wiki Freedia came out of the cohorts.

1814
01:25:25,000 --> 01:25:27,000
I already mentioned Tolgate.

1815
01:25:28,000 --> 01:25:31,000
And site came out of it still underutilized but getting there.

1816
01:25:31,000 --> 01:25:32,000
It's going to get there.

1817
01:25:32,000 --> 01:25:33,000
No.

1818
01:25:33,000 --> 01:25:36,000
Someone just wasn't it an MKStax thing.

1819
01:25:36,000 --> 01:25:42,000
And some sandwich published build like an insight with a Y.

1820
01:25:42,000 --> 01:25:43,000
That looks very good.

1821
01:25:43,000 --> 01:25:44,000
Yeah.

1822
01:25:44,000 --> 01:25:47,000
And we had some really good people there all around.

1823
01:25:47,000 --> 01:25:49,000
They're not like part of this group.

1824
01:25:49,000 --> 01:25:54,000
But we'll have we're again aiming for like 50% repeat participants, 50% new people thereabouts.

1825
01:25:54,000 --> 01:25:59,000
So if they have a friend that might you know be interested in all this stuff and coming.

1826
01:25:59,000 --> 01:26:00,000
Like tell your friends.

1827
01:26:00,000 --> 01:26:01,000
Yeah.

1828
01:26:01,000 --> 01:26:08,000
And if you were to summarize it with like elevator pitch, the one shot of what it is.

1829
01:26:08,000 --> 01:26:09,000
Oh my lord.

1830
01:26:09,000 --> 01:26:10,000
One shot.

1831
01:26:10,000 --> 01:26:16,000
I would say it's still an experiment.

1832
01:26:16,000 --> 01:26:22,000
But I would say it's an already successful experiment in exploration and ideation in the world of

1833
01:26:22,000 --> 01:26:25,000
software and technology.

1834
01:26:25,000 --> 01:26:28,000
Because I would say it's more than that.

1835
01:26:28,000 --> 01:26:30,000
Like we had match network stuff.

1836
01:26:30,000 --> 01:26:31,000
We had other stuff.

1837
01:26:31,000 --> 01:26:34,000
Like I you know did the workshop on sync thing for example.

1838
01:26:34,000 --> 01:26:36,000
Doesn't have to do anything.

1839
01:26:36,000 --> 01:26:37,000
Noster Bitcoin lightning.

1840
01:26:37,000 --> 01:26:39,000
But it's the same thing.

1841
01:26:39,000 --> 01:26:48,000
And we talked also like that's why we the reason why we started the no solutions thing was because we had these kind of conversations all the time.

1842
01:26:48,000 --> 01:26:51,000
Like every day when there's some engineering going on.

1843
01:26:51,000 --> 01:26:53,000
Every day you're going to have lunch with someone.

1844
01:26:53,000 --> 01:26:55,000
You're going to go on a walk with someone.

1845
01:26:55,000 --> 01:26:56,000
We're going to do some activity.

1846
01:26:56,000 --> 01:26:58,000
We're going to be on a boat or whatever.

1847
01:26:58,000 --> 01:26:59,000
We're going to go to the pilot.

1848
01:26:59,000 --> 01:27:00,000
Oh yeah.

1849
01:27:00,000 --> 01:27:05,000
That's the other thing that came out that I'm actually going to start using for some project that came out of the boat.

1850
01:27:05,000 --> 01:27:09,000
This idea of on your routing based on Noster which is like a tour.

1851
01:27:09,000 --> 01:27:11,000
But way faster.

1852
01:27:11,000 --> 01:27:12,000
Way fucking faster.

1853
01:27:12,000 --> 01:27:13,000
Yeah.

1854
01:27:13,000 --> 01:27:18,000
So I'm going to I might put this in the show notes like the best of the stuff that came out of it.

1855
01:27:18,000 --> 01:27:23,000
But just to give you an idea like actual or to give everyone an idea like actual usable stuff.

1856
01:27:23,000 --> 01:27:30,000
That is kind of you know like not steps and all those kind of things it's like it's everywhere and it's like.

1857
01:27:30,000 --> 01:27:32,000
And again like the blossom stuff and so on it's.

1858
01:27:32,000 --> 01:27:34,000
And I'm forgetting half the things.

1859
01:27:34,000 --> 01:27:35,000
Oh you know.

1860
01:27:35,000 --> 01:27:41,000
So so we landed on some really.

1861
01:27:42,000 --> 01:27:49,000
Interesting ideas and they were implemented and they worked and they are adopted by the actual marketplace so to speak like the marketplace of ideas at least.

1862
01:27:49,000 --> 01:27:51,000
And people actually use it and it makes me very happy.

1863
01:27:51,000 --> 01:27:53,000
And so the back to the elevator pitch.

1864
01:27:53,000 --> 01:27:57,000
So I think I think it's again it's still an experiment.

1865
01:27:57,000 --> 01:28:00,000
In the sense also that because there is no monetization whatsoever.

1866
01:28:00,000 --> 01:28:02,000
Everyone comes there on their own diet.

1867
01:28:02,000 --> 01:28:09,000
You know like it's we're happy to help with organizational stuff and how to get rooms and all those kind of things and blah blah blah.

1868
01:28:10,000 --> 01:28:13,000
But you know we don't make any money of it.

1869
01:28:13,000 --> 01:28:16,000
No one makes any money of it.

1870
01:28:16,000 --> 01:28:19,000
And we just do it because we want to see awesome stuff being built.

1871
01:28:19,000 --> 01:28:20,000
Yeah.

1872
01:28:20,000 --> 01:28:25,000
And I think there is still so much room to build awesome stuff.

1873
01:28:25,000 --> 01:28:29,000
And just the fact that become works.

1874
01:28:29,000 --> 01:28:32,000
You know we have lightning and it really works.

1875
01:28:32,000 --> 01:28:38,000
Like I just had this post recently because I was so annoyed by all the people just being mad about everything at all times.

1876
01:28:38,000 --> 01:28:40,000
Like everything's amazing and no one's happy right.

1877
01:28:40,000 --> 01:28:45,000
And it's like we're on Madera like that should be clear as well.

1878
01:28:45,000 --> 01:28:48,000
I hope there's a parent from the server engineering home page as well.

1879
01:28:48,000 --> 01:28:49,000
So it is in Madera.

1880
01:28:49,000 --> 01:28:52,000
And the awesome thing is like you can pay with Bitcoin like almost everywhere.

1881
01:28:52,000 --> 01:28:53,000
Yeah.

1882
01:28:53,000 --> 01:28:54,000
It's like fantastic.

1883
01:28:54,000 --> 01:28:56,000
And that's what I do all day every day.

1884
01:28:56,000 --> 01:28:59,000
And so I pay for breakfast lunch and dinner with with Bitcoin.

1885
01:28:59,000 --> 01:29:02,000
And it's just like it's all just scan a QR code and that's what I mentioned yesterday.

1886
01:29:02,000 --> 01:29:04,000
You know go to the dentist.

1887
01:29:04,000 --> 01:29:05,000
Pay in lightning.

1888
01:29:05,000 --> 01:29:07,000
Go to the fucking barber.

1889
01:29:07,000 --> 01:29:08,000
Pay with e-cache.

1890
01:29:08,000 --> 01:29:09,000
Like it doesn't matter.

1891
01:29:09,000 --> 01:29:11,000
You can use whatever wallet it's like all it all fucking works.

1892
01:29:11,000 --> 01:29:12,000
You know.

1893
01:29:12,000 --> 01:29:14,000
And so we are living in the future here.

1894
01:29:14,000 --> 01:29:17,000
And that's another part I think that's awesome about the six weeks.

1895
01:29:17,000 --> 01:29:19,000
Like Tim Boma for example when he built a safe box.

1896
01:29:19,000 --> 01:29:24,000
Like that's also something that was built during the server engineering or built out at least.

1897
01:29:24,000 --> 01:29:30,000
He blew his own mind by implementing an e-cache wallet from scratch basically.

1898
01:29:30,000 --> 01:29:34,000
And going through the coffee shop and paying with it.

1899
01:29:34,000 --> 01:29:37,000
With the thing that you just built yesterday.

1900
01:29:37,000 --> 01:29:42,000
You just go and you spend your e-cache on coffee and it works and you get a coffee and everything works.

1901
01:29:42,000 --> 01:29:46,000
And that's such a cool experience to have also.

1902
01:29:46,000 --> 01:29:50,000
And just kind of you know really experimenting with all the stuff that you're building out.

1903
01:29:50,000 --> 01:29:56,000
And you know when Kelly was here he did the e-cache invoices.

1904
01:29:56,000 --> 01:29:57,000
And so on.

1905
01:29:57,000 --> 01:29:58,000
And now with the e-c.

1906
01:29:58,000 --> 01:29:59,000
Yeah.

1907
01:29:59,000 --> 01:30:00,000
Exactly.

1908
01:30:00,000 --> 01:30:03,000
And you know now with the molding out payments for example.

1909
01:30:03,000 --> 01:30:09,000
You can actually go anywhere and get a hair count and pay with you know for multiple mints and all those kind of things.

1910
01:30:09,000 --> 01:30:10,000
And so that's what we have.

1911
01:30:10,000 --> 01:30:11,000
What server engineering is about.

1912
01:30:11,000 --> 01:30:13,000
We want to come up with new ideas.

1913
01:30:13,000 --> 01:30:15,000
We want to implement them straight away.

1914
01:30:15,000 --> 01:30:16,000
We want to show it to everyone.

1915
01:30:16,000 --> 01:30:17,000
Oh yeah.

1916
01:30:17,000 --> 01:30:18,000
That's the other thing.

1917
01:30:18,000 --> 01:30:19,000
We want to discuss it.

1918
01:30:19,000 --> 01:30:20,000
The emote is demoing software.

1919
01:30:20,000 --> 01:30:21,000
Yes.

1920
01:30:21,000 --> 01:30:22,000
Which is what's interesting.

1921
01:30:22,000 --> 01:30:24,000
It is just where sucks if it's broken.

1922
01:30:24,000 --> 01:30:25,000
Yes.

1923
01:30:25,000 --> 01:30:26,000
It doesn't matter.

1924
01:30:26,000 --> 01:30:27,000
No.

1925
01:30:27,000 --> 01:30:28,000
If we see a slide you'll get shot on site.

1926
01:30:28,000 --> 01:30:29,000
100%.

1927
01:30:29,000 --> 01:30:32,000
It has to be a demo of something that you actually built.

1928
01:30:32,000 --> 01:30:34,000
Or the LLM built.

1929
01:30:34,000 --> 01:30:35,000
Yeah.

1930
01:30:35,000 --> 01:30:36,000
Same.

1931
01:30:36,000 --> 01:30:37,000
Same.

1932
01:30:37,000 --> 01:30:38,000
Same.

1933
01:30:38,000 --> 01:30:39,000
One of your slaves built.

1934
01:30:39,000 --> 01:30:40,000
That's fine.

1935
01:30:40,000 --> 01:30:41,000
Big fans of slavery.

1936
01:30:41,000 --> 01:30:42,000
So yeah that's about it kind of.

1937
01:30:42,000 --> 01:30:43,000
I don't know.

1938
01:30:43,000 --> 01:30:44,000
It was a very long elevator pitch.

1939
01:30:44,000 --> 01:30:45,000
But it's very hard to.

1940
01:30:45,000 --> 01:30:46,000
It's very hard.

1941
01:30:46,000 --> 01:30:47,000
And maybe it's a long elevator.

1942
01:30:47,000 --> 01:30:48,000
You know.

1943
01:30:48,000 --> 01:30:49,000
We're in a space elevator.

1944
01:30:49,000 --> 01:30:50,000
We're in a space elevator.

1945
01:30:50,000 --> 01:30:51,000
A very slow one.

1946
01:30:51,000 --> 01:30:52,000
Maybe it's got stuck.

1947
01:30:52,000 --> 01:30:53,000
Yeah.

1948
01:30:53,000 --> 01:30:54,000
You never know.

1949
01:30:54,000 --> 01:30:55,000
But yeah.

1950
01:30:55,000 --> 01:30:56,000
That's kind of it.

1951
01:30:56,000 --> 01:30:58,000
And every week is the same kind of like same structure.

1952
01:30:58,000 --> 01:30:59,000
Again.

1953
01:30:59,000 --> 01:31:00,000
Like Monday.

1954
01:31:00,000 --> 01:31:01,000
It's like Monday mornings.

1955
01:31:01,000 --> 01:31:02,000
It's like orientation.

1956
01:31:02,000 --> 01:31:03,000
Tuesday talks.

1957
01:31:03,000 --> 01:31:04,000
Wednesday workshops.

1958
01:31:04,000 --> 01:31:05,000
Thursdays off.

1959
01:31:05,000 --> 01:31:06,000
So Thursdays kind of.

1960
01:31:06,000 --> 01:31:08,000
Thursdays where you built in time.

1961
01:31:08,000 --> 01:31:09,000
Yeah.

1962
01:31:09,000 --> 01:31:10,000
We do the microphone.

1963
01:31:10,000 --> 01:31:11,000
Friday demo day.

1964
01:31:11,000 --> 01:31:13,000
Like Friday is the most important one.

1965
01:31:13,000 --> 01:31:14,000
We'll have a nice barbecue.

1966
01:31:14,000 --> 01:31:16,000
And we try to have the time as well.

1967
01:31:16,000 --> 01:31:17,000
Yeah.

1968
01:31:17,000 --> 01:31:19,000
There's always something to eat around.

1969
01:31:19,000 --> 01:31:20,000
And yeah.

1970
01:31:20,000 --> 01:31:22,000
And on the weekends we'll also do something.

1971
01:31:22,000 --> 01:31:25,000
Like we'll go on a hike or we'll go out with a boat or we'll do whatever.

1972
01:31:25,000 --> 01:31:26,000
We did many things.

1973
01:31:26,000 --> 01:31:27,000
Like we went climbing ones.

1974
01:31:27,000 --> 01:31:28,000
Like people went go karting.

1975
01:31:28,000 --> 01:31:29,000
All kinds of shit.

1976
01:31:29,000 --> 01:31:30,000
Like whatever.

1977
01:31:30,000 --> 01:31:32,000
We try to do whatever fastest conversation.

1978
01:31:32,000 --> 01:31:35,000
So we try to do whatever brings people together.

1979
01:31:35,000 --> 01:31:37,000
And where you can have free flow and conversation.

1980
01:31:37,000 --> 01:31:40,000
And that's also like, I mean, no solutions.

1981
01:31:40,000 --> 01:31:42,000
It's also something that came out of sovereign engineering.

1982
01:31:42,000 --> 01:31:46,000
Because we were like always, you know, walking up and down and going on a hike.

1983
01:31:46,000 --> 01:31:48,000
And just talking about those things.

1984
01:31:48,000 --> 01:31:50,000
And it was always, we always had novel insights.

1985
01:31:50,000 --> 01:31:54,000
And like the conversation always will take you somewhere where you never thought it would go.

1986
01:31:54,000 --> 01:31:57,000
And we were like, okay, we should open this up and just record some of them.

1987
01:31:57,000 --> 01:31:58,000
And so on.

1988
01:31:58,000 --> 01:32:00,000
And just do this from time to time.

1989
01:32:00,000 --> 01:32:01,000
Yeah.

1990
01:32:01,000 --> 01:32:04,000
And because that's kind of the to bring it back to the very beginning.

1991
01:32:04,000 --> 01:32:06,000
Where it's like, okay, at least we know our flaws.

1992
01:32:06,000 --> 01:32:08,000
And we are really bad at explaining sovereign engineering.

1993
01:32:08,000 --> 01:32:12,000
We're really bad at kind of, you know, letting people know that this exists.

1994
01:32:12,000 --> 01:32:14,000
We're really bad at five tweets.

1995
01:32:14,000 --> 01:32:18,000
I'm not explaining our thinking around things, including sovereign engineering.

1996
01:32:18,000 --> 01:32:19,000
Yeah.

1997
01:32:19,000 --> 01:32:20,000
Yeah.

1998
01:32:20,000 --> 01:32:23,000
Because we have these conversations and we explore a bunch of different ideas.

1999
01:32:23,000 --> 01:32:25,000
And then we just kind of keep it to the box.

2000
01:32:25,000 --> 01:32:27,000
Yeah, I'll never sit down and write about it.

2001
01:32:27,000 --> 01:32:28,000
And write about it.

2002
01:32:28,000 --> 01:32:29,000
Yeah, yeah, yeah.

2003
01:32:29,000 --> 01:32:31,000
It's like that's everything, you know.

2004
01:32:31,000 --> 01:32:33,000
I'm no point to think.

2005
01:32:33,000 --> 01:32:37,000
I'm a little bit of a fighting figure.

2006
01:32:37,000 --> 01:32:38,000
We're both to blame.

2007
01:32:38,000 --> 01:32:41,000
And anyway, like maybe hopefully that's useful.

2008
01:32:41,000 --> 01:32:43,000
I'll run the whole transcript through an LLM.

2009
01:32:43,000 --> 01:32:44,000
And that will generate you.

2010
01:32:44,000 --> 01:32:45,000
And say no, we'll come out.

2011
01:32:45,000 --> 01:32:47,000
The shining city in cyberspace.

2012
01:32:47,000 --> 01:32:48,000
That's would be so funny.

2013
01:32:48,000 --> 01:32:51,000
Like I'll let I'll rebuild the website.

2014
01:32:51,000 --> 01:32:54,000
And it will just be a shining city in cyberspace.

2015
01:32:54,000 --> 01:32:59,000
Anyway, it's if you if you listen to every single no solutions conversation,

2016
01:32:59,000 --> 01:33:00,000
you should get an idea.

2017
01:33:00,000 --> 01:33:04,000
Yeah, you'll have a very good idea of the inside engineering is about.

2018
01:33:04,000 --> 01:33:05,000
And what to expect.

2019
01:33:05,000 --> 01:33:06,000
And so on.

2020
01:33:06,000 --> 01:33:07,000
So there shouldn't be any surprises.

2021
01:33:07,000 --> 01:33:13,000
But that's also like, you know, who's going to listen to 15 hours of like waves and street sounds.

2022
01:33:13,000 --> 01:33:14,000
I'm talking.

2023
01:33:14,000 --> 01:33:17,000
So you really have to want it.

2024
01:33:17,000 --> 01:33:22,000
But in our defense, we also we really want to have the people here that really want to be here.

2025
01:33:22,000 --> 01:33:26,000
So again, if you're interested in in Nostar, if you want to try something new,

2026
01:33:26,000 --> 01:33:31,000
if you want to meet the best people in the space and just, you know, experiment with.

2027
01:33:31,000 --> 01:33:33,000
Ekash, Nostar lighting stuff.

2028
01:33:33,000 --> 01:33:34,000
What have you?

2029
01:33:34,000 --> 01:33:36,000
It's it's a fantastic place to be.

2030
01:33:36,000 --> 01:33:38,000
Madera is beautiful.

2031
01:33:38,000 --> 01:33:39,000
Food is great.

2032
01:33:39,000 --> 01:33:42,000
Weather is like 21 degrees centigrade all year round.

2033
01:33:42,000 --> 01:33:45,000
It's like fantastic in terms of climate, all those kind of things.

2034
01:33:46,000 --> 01:33:47,000
It's kids friendly.

2035
01:33:47,000 --> 01:33:49,000
People brought their families as well.

2036
01:33:49,000 --> 01:33:52,000
You know, we're here with our families as well.

2037
01:33:52,000 --> 01:33:55,000
And yeah, that's kind of it.

2038
01:33:55,000 --> 01:34:03,000
And in the, I actually said this as a send off on the very first podcast that you and me recorded together in Francis' place.

2039
01:34:03,000 --> 01:34:04,000
Remember?

2040
01:34:04,000 --> 01:34:05,000
Okay.

2041
01:34:05,000 --> 01:34:06,000
I do remember.

2042
01:34:06,000 --> 01:34:07,000
Yeah, yeah, yeah.

2043
01:34:07,000 --> 01:34:08,000
I do remember.

2044
01:34:08,000 --> 01:34:09,000
Yeah, yeah.

2045
01:34:09,000 --> 01:34:12,000
What I had in mind at that point and I kind of still think that that's in Costa Rica.

2046
01:34:12,000 --> 01:34:13,000
That was in Costa Rica.

2047
01:34:13,000 --> 01:34:14,000
That was in Costa Rica.

2048
01:34:14,000 --> 01:34:15,000
Yeah, that's how a while ago.

2049
01:34:15,000 --> 01:34:23,000
Is that there is a lot of value and kind of fits perfect with Yolomode.

2050
01:34:23,000 --> 01:34:29,000
There is a lot of value in just making a wild decision that can completely change.

2051
01:34:29,000 --> 01:34:31,000
It has the opportunity of changing your life.

2052
01:34:31,000 --> 01:34:33,000
And it's a fucking one decision away.

2053
01:34:33,000 --> 01:34:35,000
It is literally just making the decision.

2054
01:34:35,000 --> 01:34:38,000
Then there is a to the least after that decision.

2055
01:34:38,000 --> 01:34:41,000
But coming on participating in something like summer engineering.

2056
01:34:41,000 --> 01:34:44,000
How many people have told you and me?

2057
01:34:44,000 --> 01:34:45,000
It changed their life.

2058
01:34:45,000 --> 01:34:46,000
Oh, absolutely.

2059
01:34:46,000 --> 01:34:48,000
Many people quit their job and went full time.

2060
01:34:48,000 --> 01:34:49,000
Full retirement.

2061
01:34:49,000 --> 01:34:50,000
Full on retirement.

2062
01:34:50,000 --> 01:34:52,000
Yeah, yeah.

2063
01:34:52,000 --> 01:34:55,000
And you know, they're still alive and can fit themselves.

2064
01:34:55,000 --> 01:34:56,000
I mean, kind of.

2065
01:34:56,000 --> 01:34:57,000
Yeah.

2066
01:34:57,000 --> 01:34:59,000
I mean, Bully for value.

2067
01:34:59,000 --> 01:35:00,000
Yeah.

2068
01:35:00,000 --> 01:35:01,000
They fit themselves.

2069
01:35:01,000 --> 01:35:05,000
You know, they set each other set with us.

2070
01:35:05,000 --> 01:35:10,000
But it's literally you listen to a podcast and you say you know what?

2071
01:35:10,000 --> 01:35:11,000
Fuck it.

2072
01:35:11,000 --> 01:35:12,000
I'm going to do it.

2073
01:35:12,000 --> 01:35:13,000
Yeah.

2074
01:35:13,000 --> 01:35:14,000
Yeah.

2075
01:35:14,000 --> 01:35:16,000
That's what Justin did when he came.

2076
01:35:16,000 --> 01:35:17,000
And I'm going to.

2077
01:35:17,000 --> 01:35:22,000
I'm literally going to quote I'm literally going to quote him on the new website.

2078
01:35:22,000 --> 01:35:24,000
It was like, I'm surprised.

2079
01:35:24,000 --> 01:35:25,000
I'm still here.

2080
01:35:25,000 --> 01:35:26,000
Justin Moon.

2081
01:35:26,000 --> 01:35:27,000
Week six.

2082
01:35:27,000 --> 01:35:32,000
He changed his flight like three times or so because he wanted to stay for one week.

2083
01:35:32,000 --> 01:35:33,000
And then it became two weeks.

2084
01:35:33,000 --> 01:35:35,000
And he stayed for the whole six weeks.

2085
01:35:35,000 --> 01:35:36,000
And he's coming back.

2086
01:35:36,000 --> 01:35:38,000
So we must do something right.

2087
01:35:38,000 --> 01:35:41,000
And that's also one thing maybe worth mentioning.

2088
01:35:41,000 --> 01:35:43,000
Like the first three cohorts.

2089
01:35:43,000 --> 01:35:47,000
We burned ourselves to a crisp because we did everything at all times.

2090
01:35:47,000 --> 01:35:48,000
Yes.

2091
01:35:48,000 --> 01:35:51,000
And now we got better at delegating like again, back to slavery.

2092
01:35:51,000 --> 01:35:53,000
That's going to be the theme of everything.

2093
01:35:53,000 --> 01:35:55,000
Very slavery is going to be the theme of the next cohort.

2094
01:35:55,000 --> 01:35:56,000
You know.

2095
01:35:56,000 --> 01:35:57,000
It's either Mark Yolo mode.

2096
01:35:57,000 --> 01:35:58,000
It emerges.

2097
01:35:58,000 --> 01:36:03,000
And so every week, like we'll have six weeks.

2098
01:36:03,000 --> 01:36:07,000
We also have one week where we won't have fixed programming.

2099
01:36:07,000 --> 01:36:09,000
So a little bit of a break in between.

2100
01:36:09,000 --> 01:36:12,000
And every week, it's going to have a different cap.

2101
01:36:12,000 --> 01:36:16,000
So those are like veterans.

2102
01:36:16,000 --> 01:36:19,000
Alumni that went through the program some of the multiple times.

2103
01:36:19,000 --> 01:36:24,000
And they are going to, you know, make sure that the whole organizational structure works.

2104
01:36:24,000 --> 01:36:25,000
And everyone knows what to do.

2105
01:36:25,000 --> 01:36:26,000
And so on.

2106
01:36:26,000 --> 01:36:29,000
Because otherwise, I just, you know, like I burn out and go insane.

2107
01:36:29,000 --> 01:36:31,000
But that was a great choice.

2108
01:36:31,000 --> 01:36:32,000
It worked perfectly the last time.

2109
01:36:32,000 --> 01:36:33,000
And we're going to do that again.

2110
01:36:33,000 --> 01:36:35,000
I think it worked even better.

2111
01:36:35,000 --> 01:36:36,000
Yeah.

2112
01:36:36,000 --> 01:36:37,000
No, it was really, really good.

2113
01:36:37,000 --> 01:36:41,000
Because everyone will bring their own perspective and have their own spin on things.

2114
01:36:41,000 --> 01:36:42,000
On energy.

2115
01:36:42,000 --> 01:36:43,000
Yeah, exactly.

2116
01:36:43,000 --> 01:36:44,000
And they really go in full.

2117
01:36:44,000 --> 01:36:45,000
Emphasize different things.

2118
01:36:45,000 --> 01:36:46,000
Yeah.

2119
01:36:46,000 --> 01:36:50,000
Because, you know, like I'm, you know, if I'm out there and standing in front of people,

2120
01:36:50,000 --> 01:36:53,000
I will just always repeat the same line, which is like shining stadium cyberspace.

2121
01:36:53,000 --> 01:36:55,000
And quote Taylor all the time.

2122
01:36:55,000 --> 01:36:57,000
And that's, that's also not good.

2123
01:36:57,000 --> 01:36:59,000
That's like.

2124
01:36:59,000 --> 01:37:05,000
So it's better if like Justin or Cali or Aryan or whoever is out there.

2125
01:37:05,000 --> 01:37:09,000
And just talking about something else in the MSTR.

2126
01:37:09,000 --> 01:37:11,000
Well, yeah, that's about it.

2127
01:37:11,000 --> 01:37:13,000
I think I hope that was useful.

2128
01:37:13,000 --> 01:37:14,000
Yeah, I have one.

2129
01:37:14,000 --> 01:37:15,000
It was useful for me.

2130
01:37:15,000 --> 01:37:16,000
Yeah.

2131
01:37:16,000 --> 01:37:19,000
I have one more thing that just came to me.

2132
01:37:19,000 --> 01:37:24,000
Is that I think something like this is become like some engineer.

2133
01:37:24,000 --> 01:37:26,000
It's becoming increasingly important.

2134
01:37:26,000 --> 01:37:32,000
Because we are completely in the age of you can build anything and everything.

2135
01:37:32,000 --> 01:37:36,000
And what to build becomes all the more important.

2136
01:37:36,000 --> 01:37:39,000
Why, why build anything at all?

2137
01:37:39,000 --> 01:37:41,000
If you can't build absolutely everything, why build?

2138
01:37:41,000 --> 01:37:42,000
No, no, no.

2139
01:37:42,000 --> 01:37:43,000
Well, yeah.

2140
01:37:43,000 --> 01:37:44,000
No, no, no, everything.

2141
01:37:44,000 --> 01:37:45,000
Yeah, you're right.

2142
01:37:45,000 --> 01:37:46,000
What, you can build anything.

2143
01:37:46,000 --> 01:37:49,000
No, I'm, I'm obviously being for Jesus, but that's where it's trending.

2144
01:37:49,000 --> 01:37:51,000
And that's kind of the question everyone should ask themselves.

2145
01:37:51,000 --> 01:37:52,000
Right.

2146
01:37:52,000 --> 01:37:53,000
Yeah, yeah.

2147
01:37:53,000 --> 01:37:55,000
So that kind of question is what we explore.

2148
01:37:55,000 --> 01:37:56,000
Yeah.

2149
01:37:56,000 --> 01:37:57,000
That's a very engineering.

2150
01:37:57,000 --> 01:37:59,000
Why building, why build this thing?

2151
01:37:59,000 --> 01:38:01,000
Why put, because I think.

2152
01:38:01,000 --> 01:38:03,000
How should the future look like in some sense?

2153
01:38:03,000 --> 01:38:04,000
In Europe.

2154
01:38:04,000 --> 01:38:05,000
How do you want the future to look like?

2155
01:38:05,000 --> 01:38:06,000
Exactly, exactly.

2156
01:38:06,000 --> 01:38:07,000
And work for that.

2157
01:38:07,000 --> 01:38:08,000
Exactly.

2158
01:38:08,000 --> 01:38:09,000
And build or stand.

2159
01:38:09,000 --> 01:38:10,000
Yeah, and build accordingly.

2160
01:38:10,000 --> 01:38:16,000
Because I think it's all going to come down to, to, to taste.

2161
01:38:16,000 --> 01:38:17,000
And caring.

2162
01:38:17,000 --> 01:38:18,000
What do you care about?

2163
01:38:18,000 --> 01:38:19,000
Absolutely.

2164
01:38:19,000 --> 01:38:20,000
Yeah.

2165
01:38:20,000 --> 01:38:21,000
Tastes and melis.

2166
01:38:21,000 --> 01:38:22,000
Yeah.

2167
01:38:22,000 --> 01:38:23,000
Exactly.

2168
01:38:23,000 --> 01:38:24,000
Carrying is actually a great point.

2169
01:38:24,000 --> 01:38:26,000
That's a, by the way, to quote Verakie again.

2170
01:38:26,000 --> 01:38:28,000
That's the distinction between humans and Allah lambs.

2171
01:38:28,000 --> 01:38:30,000
Allah lambs don't care about the information they are processing.

2172
01:38:30,000 --> 01:38:31,000
We do.

2173
01:38:31,000 --> 01:38:32,000
Yes.

2174
01:38:32,000 --> 01:38:34,000
And it is because we're embodied and blah, blah, blah, blah, blah.

2175
01:38:34,000 --> 01:38:36,000
But I talked about the body mental last five times.

2176
01:38:36,000 --> 01:38:37,000
We talked, I think.

2177
01:38:37,000 --> 01:38:40,000
I'm focusing on the shining city in cyberspace today.

2178
01:38:40,000 --> 01:38:41,000
So there are no bodies.

2179
01:38:41,000 --> 01:38:42,000
There are just slaves.

2180
01:38:42,000 --> 01:38:44,000
The non-embodied slaves.

2181
01:38:44,000 --> 01:38:47,000
They're going to, it's going to be glorious.

2182
01:38:47,000 --> 01:38:49,000
But, yeah.

2183
01:38:49,000 --> 01:38:50,000
Any last words?

2184
01:38:50,000 --> 01:38:51,000
No.

2185
01:38:51,000 --> 01:38:55,000
I gave like six times my, my last words.

2186
01:38:55,000 --> 01:38:58,000
I think I have, I'm out of ideas and I'm.

2187
01:38:58,000 --> 01:38:59,000
Fantastic.

2188
01:38:59,000 --> 01:39:02,000
Now let's drink something, let's eat something.

2189
01:39:02,000 --> 01:39:03,000
That was good.

2190
01:39:03,000 --> 01:39:04,000
That was good.

2191
01:39:04,000 --> 01:39:05,000
All right.

2192
01:39:05,000 --> 01:39:08,000
One more snap.

2193
01:39:08,000 --> 01:39:09,000
We can do it.