1
00:00:00,000 --> 00:00:00,880
And we're live.

2
00:00:01,780 --> 00:00:02,600
Yes, live.

3
00:00:02,700 --> 00:00:03,120
Here we are.

4
00:00:03,340 --> 00:00:05,040
Welcome to another episode of PBJ.

5
00:00:05,680 --> 00:00:05,920
Yeah.

6
00:00:06,160 --> 00:00:07,260
Matt Coate in the house.

7
00:00:08,080 --> 00:00:08,400
Oh.

8
00:00:08,520 --> 00:00:09,120
You already commented.

9
00:00:09,340 --> 00:00:10,340
Yeah, number one, baby.

10
00:00:10,940 --> 00:00:11,340
Get in there.

11
00:00:11,340 --> 00:00:11,500
First comment.

12
00:00:11,880 --> 00:00:12,440
Good to have you.

13
00:00:12,620 --> 00:00:13,320
Yeah, good to have you.

14
00:00:13,780 --> 00:00:18,000
You know, we were very close to not being able to do a live stream today.

15
00:00:18,200 --> 00:00:18,740
So I heard.

16
00:00:19,040 --> 00:00:20,200
Or even any show at all.

17
00:00:20,600 --> 00:00:21,280
Any show at all?

18
00:00:21,740 --> 00:00:24,220
Well, because DK's remote, so we need internet.

19
00:00:24,480 --> 00:00:24,860
That's right.

20
00:00:24,860 --> 00:00:30,180
And there was a power outage in the Presidio this morning.

21
00:00:30,800 --> 00:00:31,200
Unexpected.

22
00:00:31,540 --> 00:00:31,960
Unexpected.

23
00:00:32,260 --> 00:00:36,880
And it took hours for technical support to get the internet working.

24
00:00:37,040 --> 00:00:38,780
I hear it was because there was probably, I don't know,

25
00:00:38,920 --> 00:00:40,760
was it a data center or some Bitcoin miners?

26
00:00:40,960 --> 00:00:42,120
That was interesting.

27
00:00:42,620 --> 00:00:43,680
I don't have a second at all.

28
00:00:44,400 --> 00:00:44,580
AI.

29
00:00:45,640 --> 00:00:46,320
Definitely AI.

30
00:00:47,100 --> 00:00:48,280
Bitcoin only helps with that.

31
00:00:49,080 --> 00:00:49,620
But we're here.

32
00:00:49,760 --> 00:00:50,300
But we're here.

33
00:00:51,480 --> 00:00:51,780
We're here.

34
00:00:51,880 --> 00:00:52,660
It's nice to see you guys.

35
00:00:53,160 --> 00:00:53,960
Actually, real talk.

36
00:00:53,960 --> 00:00:55,080
I think the Presidio,

37
00:00:55,480 --> 00:00:56,160
correct me if I'm wrong,

38
00:00:56,220 --> 00:00:57,100
they may have some

39
00:00:57,100 --> 00:00:58,300
like independent microgrid.

40
00:00:58,820 --> 00:00:59,900
There's like something special here.

41
00:00:59,980 --> 00:01:00,540
They were at one point

42
00:01:00,540 --> 00:01:02,160
were trying to go completely solar

43
00:01:02,160 --> 00:01:03,280
and have their own batteries.

44
00:01:03,400 --> 00:01:04,500
I don't know if that actually worked out.

45
00:01:04,740 --> 00:01:05,600
It does seem though

46
00:01:05,600 --> 00:01:06,700
like if they try to do that,

47
00:01:06,760 --> 00:01:07,620
they haven't done a good job

48
00:01:07,620 --> 00:01:08,660
because I think this is the second time

49
00:01:08,660 --> 00:01:09,600
we've had a power arch

50
00:01:09,600 --> 00:01:10,700
in the last like year.

51
00:01:11,480 --> 00:01:11,960
That sounds right.

52
00:01:12,040 --> 00:01:12,140
Yeah.

53
00:01:12,520 --> 00:01:13,620
I know of there being

54
00:01:13,620 --> 00:01:14,920
a separate water system,

55
00:01:15,020 --> 00:01:16,360
but I didn't know about a separate grid.

56
00:01:16,720 --> 00:01:17,820
Like I think Presidio water

57
00:01:17,820 --> 00:01:18,900
is not from Hetch Hetchy.

58
00:01:19,980 --> 00:01:21,880
Yeah, it's a terawatt scale.

59
00:01:22,020 --> 00:01:23,000
No, it's definitely not.

60
00:01:23,120 --> 00:01:23,900
This would probably be like,

61
00:01:23,960 --> 00:01:26,740
kilowatt scale, given how faulty this is.

62
00:01:30,580 --> 00:01:32,960
Or it's an attempted terawatt scale

63
00:01:32,960 --> 00:01:35,060
that is a failure.

64
00:01:35,280 --> 00:01:36,240
A major failure.

65
00:01:36,580 --> 00:01:37,680
But we're here, so that's good.

66
00:01:38,300 --> 00:01:39,460
We're here, we're learning,

67
00:01:39,600 --> 00:01:40,600
we've got a lot of stuff to discuss.

68
00:01:40,720 --> 00:01:42,760
Dekia was giving us an astrology lesson a minute ago.

69
00:01:43,640 --> 00:01:44,080
Astronomy.

70
00:01:44,300 --> 00:01:44,860
Astronomy and astronomy.

71
00:01:44,860 --> 00:01:46,260
Oh, sorry, sorry, I get those mixed up.

72
00:01:47,680 --> 00:01:50,020
Oh, I was promised astrology.

73
00:01:50,020 --> 00:01:51,140
Yeah, exactly, Dekia.

74
00:01:52,200 --> 00:01:52,720
I was.

75
00:01:53,960 --> 00:02:08,960
I was actually, so I'm going to catch up a little bit on some life stuff because I was just traveling and I know that I've had some, let's say I've had underwhelming experiences with full self-driving supervised on the Tesla.

76
00:02:09,460 --> 00:02:12,740
And I resubbed and I'd say it is improved.

77
00:02:13,300 --> 00:02:13,600
So I am.

78
00:02:13,720 --> 00:02:14,640
I don't know.

79
00:02:15,100 --> 00:02:16,400
He doesn't have the latest hardware.

80
00:02:16,400 --> 00:02:16,720
Yeah.

81
00:02:16,920 --> 00:02:17,940
I have hardware three.

82
00:02:18,100 --> 00:02:18,520
It's old.

83
00:02:18,840 --> 00:02:19,240
It's old.

84
00:02:19,240 --> 00:02:22,040
I'm among the poors who have old cars.

85
00:02:22,040 --> 00:02:24,600
selective Elon hater

86
00:02:24,600 --> 00:02:28,160
no not

87
00:02:28,160 --> 00:02:28,800
not a hater

88
00:02:28,800 --> 00:02:31,320
but DK what did you want to share about your

89
00:02:31,320 --> 00:02:32,320
latest

90
00:02:32,320 --> 00:02:34,860
FSD experience

91
00:02:34,860 --> 00:02:37,500
the big deal is that it actually

92
00:02:37,500 --> 00:02:39,180
it works a lot better than it did

93
00:02:39,180 --> 00:02:40,780
last time I had it six months ago

94
00:02:40,780 --> 00:02:43,680
but the big thing to note

95
00:02:43,680 --> 00:02:45,520
is it worked so well I was driving through

96
00:02:45,520 --> 00:02:47,460
the Nevada desert

97
00:02:47,460 --> 00:02:49,600
and I got

98
00:02:49,600 --> 00:02:51,720
chat GPT has

99
00:02:51,720 --> 00:02:58,340
this codex vibe coding product like a cli tool right and i had the computer open and i was just

100
00:02:58,340 --> 00:03:03,460
talking to it and i would just say continue like i gave it a plan before i started driving of like

101
00:03:03,460 --> 00:03:09,040
a bunch of stuff i wanted to accomplish and it would just churn through it and it would just rip

102
00:03:09,040 --> 00:03:13,080
and rip and rip and then it would finish and i just every 10 or 15 minutes i'd open up the

103
00:03:13,080 --> 00:03:17,300
screen check over and make sure it was done and had something else to do and it always had its

104
00:03:17,300 --> 00:03:21,780
own plan of what it wanted to do next. And I would just say, go ahead, continue, continue.

105
00:03:22,280 --> 00:03:26,880
So I was basically vibe coding for five hours while like full self-driving.

106
00:03:27,660 --> 00:03:27,820
Wow.

107
00:03:27,920 --> 00:03:32,400
Do you have a setup where like a text to speech where it can speak to you, its output?

108
00:03:33,000 --> 00:03:37,520
I can't hear it. It can hear me. So I speak to it to say, continue.

109
00:03:37,520 --> 00:03:41,620
I would feel more comfortable if you get TTS working.

110
00:03:41,620 --> 00:03:44,740
I'm a huge advocate

111
00:03:44,740 --> 00:03:46,540
for Tesla FSD

112
00:03:46,540 --> 00:03:47,780
and I think it works great

113
00:03:47,780 --> 00:03:49,880
it works great when you're

114
00:03:49,880 --> 00:03:52,060
not reading a laptop screen

115
00:03:52,060 --> 00:03:52,800
right right

116
00:03:52,800 --> 00:03:54,840
but what I do is I let it

117
00:03:54,840 --> 00:03:56,940
finish its thing

118
00:03:56,940 --> 00:03:58,800
no vibe coding while driving dude

119
00:03:58,800 --> 00:03:59,060
goodbye

120
00:03:59,060 --> 00:04:03,280
hopefully we don't have any

121
00:04:03,280 --> 00:04:04,460
California highway police

122
00:04:04,460 --> 00:04:05,980
listening to the show

123
00:04:05,980 --> 00:04:09,120
I would like to hear

124
00:04:09,120 --> 00:04:10,480
the output or like the summary

125
00:04:10,480 --> 00:04:15,520
I think that would be like my ideal vibe coding setup would be I speak to it, it does the thing.

126
00:04:15,940 --> 00:04:22,880
And then, you know, I don't need to see all the interstitial like, you know, steps it takes, but that it gives me like, hey, here's a summary and it could read that to me.

127
00:04:23,040 --> 00:04:23,460
That'd be ideal.

128
00:04:24,040 --> 00:04:24,760
Because I have quite a bit.

129
00:04:25,100 --> 00:04:26,780
I mean, do you use Grok in your Tesla?

130
00:04:27,920 --> 00:04:30,140
Not in the Tesla, but I use it.

131
00:04:30,280 --> 00:04:32,060
I do it all the time and it's a good experience.

132
00:04:32,760 --> 00:04:34,320
You talk to it, it speaks to you.

133
00:04:34,400 --> 00:04:35,480
And I mean, you just have a conversation.

134
00:04:36,380 --> 00:04:37,500
And I assume that's native, right?

135
00:04:37,520 --> 00:04:39,020
Like it just like rolls out on Tesla.

136
00:04:39,660 --> 00:04:39,860
Yep.

137
00:04:39,860 --> 00:04:39,940
Yep.

138
00:04:40,480 --> 00:04:42,520
it's on the string wheel, you hold the button down

139
00:04:42,520 --> 00:04:43,200
and speak to it.

140
00:04:44,200 --> 00:04:48,300
I wanted to write a post,

141
00:04:48,380 --> 00:04:50,520
so I just gave it some notes, and then it wrote the post for me,

142
00:04:50,580 --> 00:04:51,960
and then when I was back on my laptop later,

143
00:04:52,180 --> 00:04:52,860
it was all available.

144
00:04:53,120 --> 00:04:55,360
Did you have to configure it or set it up?

145
00:04:55,400 --> 00:04:57,160
Because I don't think I've seen even that that exists.

146
00:05:00,000 --> 00:05:01,500
I forget how I discovered it.

147
00:05:02,960 --> 00:05:04,860
But you don't have to turn it on or enable it.

148
00:05:04,860 --> 00:05:06,140
You just start going.

149
00:05:06,940 --> 00:05:07,660
Yeah, it's great.

150
00:05:07,780 --> 00:05:09,020
And when you say you had to write a post,

151
00:05:09,020 --> 00:05:10,040
was this an expo

152
00:05:10,040 --> 00:05:10,760
and like could it have

153
00:05:10,760 --> 00:05:11,960
actually like tweeted for you?

154
00:05:12,040 --> 00:05:12,260
Maybe.

155
00:05:12,440 --> 00:05:13,320
This is just like

156
00:05:13,320 --> 00:05:16,240
so I just had it available

157
00:05:16,240 --> 00:05:18,120
as like a grok session.

158
00:05:18,360 --> 00:05:18,580
Okay.

159
00:05:19,040 --> 00:05:19,180
Cool.

160
00:05:19,200 --> 00:05:19,880
When I was back on

161
00:05:19,880 --> 00:05:21,020
my phone or laptop.

162
00:05:21,400 --> 00:05:21,640
Beautiful.

163
00:05:24,580 --> 00:05:25,680
Yeah, but it'd be interesting

164
00:05:25,680 --> 00:05:28,120
to experiment with tweets.

165
00:05:28,360 --> 00:05:28,760
Yeah, it'd be cool

166
00:05:28,760 --> 00:05:29,800
if you get like live tweeting

167
00:05:29,800 --> 00:05:30,540
while you're driving.

168
00:05:31,260 --> 00:05:32,340
Hey guys, here I am.

169
00:05:32,420 --> 00:05:33,120
And we got the GPS.

170
00:05:35,760 --> 00:05:36,640
Probably a terrible idea.

171
00:05:37,440 --> 00:05:38,680
Dropping Bitcoin along the way

172
00:05:38,680 --> 00:05:41,940
So I think this is actually an interesting point

173
00:05:41,940 --> 00:05:43,260
because I know we're going to talk a lot about AI

174
00:05:43,260 --> 00:05:44,460
and agents and coding today

175
00:05:44,460 --> 00:05:48,420
because this is the Presidio AI Builder Podcast.

176
00:05:50,240 --> 00:05:52,140
But I'm interested, DK,

177
00:05:52,220 --> 00:05:53,680
you mentioned you were doing the new,

178
00:05:53,700 --> 00:05:55,440
you were vibing with the new Codex tool.

179
00:05:56,560 --> 00:05:59,560
You know, what is your experience been?

180
00:05:59,580 --> 00:06:00,580
Because it seemed like for a minute,

181
00:06:00,700 --> 00:06:01,980
like everyone was using Claw.

182
00:06:02,120 --> 00:06:03,500
It still seems like most people are using Claw,

183
00:06:03,600 --> 00:06:04,920
but it seems like Codex had a jump.

184
00:06:05,960 --> 00:06:06,920
I mean, obviously, you know,

185
00:06:06,920 --> 00:06:10,820
people are using x for more things as well uh i think this is going to get into some of our

186
00:06:10,820 --> 00:06:14,860
questions we have around monetary payments stuff like what what is your just kind of take state

187
00:06:14,860 --> 00:06:18,420
of the art right now what's going on okay i have i have lots of opinions about this because i

188
00:06:18,420 --> 00:06:24,820
basically spend all day every day doing this i love it so um so for the last several months i've

189
00:06:24,820 --> 00:06:31,520
been pretty exclusively clod code like everything i usually keep uh two clod code terminals open at

190
00:06:31,520 --> 00:06:37,100
a given time i like cloud code cli the command line interface tool and i use those and i kind of

191
00:06:37,100 --> 00:06:41,400
bounce them off each other or i do like one task in one and while it's turning i go do something

192
00:06:41,400 --> 00:06:44,720
else in the other and then when the first one finishes i'm back so i kind of just ping pong

193
00:06:44,720 --> 00:06:51,960
back and forth a bunch myself and um and i was doing that to build this this new app i the one

194
00:06:51,960 --> 00:06:57,220
actually started back at the um the hackathon i've actually decided to really go deep um so i'm still

195
00:06:57,220 --> 00:06:58,880
work on that. Right now, I'm literally

196
00:06:58,880 --> 00:07:01,140
vibe coding while we speak, as I've

197
00:07:01,140 --> 00:07:01,980
done before on the show.

198
00:07:03,360 --> 00:07:05,160
But it's when I'm talking about Iowa

199
00:07:05,160 --> 00:07:06,420
State basketball. Exactly.

200
00:07:06,880 --> 00:07:08,880
We'll be back in 10 minutes.

201
00:07:09,180 --> 00:07:11,300
Hold on. I'm going to go off on a tangent

202
00:07:11,300 --> 00:07:13,160
real quick. I won't talk about the results

203
00:07:13,160 --> 00:07:14,080
because we've been losing, but

204
00:07:14,080 --> 00:07:16,800
we had a comment on our YouTube

205
00:07:16,800 --> 00:07:18,080
from someone who

206
00:07:18,080 --> 00:07:21,360
is a Bitcoiner. They discovered the show

207
00:07:21,360 --> 00:07:23,240
and then they said in their

208
00:07:23,240 --> 00:07:24,860
comment, two minutes into the show,

209
00:07:25,280 --> 00:07:26,460
we were talking about Iowa State basketball.

210
00:07:26,460 --> 00:07:28,460
and this guy was like in heaven

211
00:07:28,460 --> 00:07:31,540
because he is an Iowa State basketball fan

212
00:07:31,540 --> 00:07:32,620
and a Bitcoiner.

213
00:07:33,100 --> 00:07:34,020
So what are the odds?

214
00:07:34,620 --> 00:07:36,260
The second, me and that person.

215
00:07:36,420 --> 00:07:38,740
We have found the only two of these people

216
00:07:38,740 --> 00:07:39,680
that exist in the world

217
00:07:39,680 --> 00:07:41,720
and they are both now linked to PBJ.

218
00:07:43,200 --> 00:07:43,980
God bless.

219
00:07:45,900 --> 00:07:46,520
All right.

220
00:07:47,180 --> 00:07:47,960
Back to you, Deacon.

221
00:07:48,560 --> 00:07:49,300
Back to my good.

222
00:07:49,720 --> 00:07:51,620
I've got my two Claude codes open

223
00:07:51,620 --> 00:07:53,160
and it's very creative.

224
00:07:53,380 --> 00:07:54,440
It's Brainstorm Partner.

225
00:07:54,440 --> 00:07:56,440
I feel like I'm working with an artist, right?

226
00:07:56,460 --> 00:07:57,860
when I'm talking to Cloud Code.

227
00:07:58,860 --> 00:08:01,040
And it's a wonderful tool.

228
00:08:01,220 --> 00:08:03,460
It really explores a lot of ideas.

229
00:08:04,200 --> 00:08:05,360
It builds great software.

230
00:08:05,540 --> 00:08:06,380
It's really good at that.

231
00:08:07,740 --> 00:08:09,220
But I often find it,

232
00:08:09,340 --> 00:08:11,240
I get stuck in some rat hole

233
00:08:11,240 --> 00:08:12,680
and it can't unstick itself.

234
00:08:12,800 --> 00:08:14,520
So I actually also have been keeping

235
00:08:14,520 --> 00:08:17,780
Google Gemini CLI,

236
00:08:18,140 --> 00:08:18,840
which is another thing,

237
00:08:18,920 --> 00:08:19,700
open another terminal.

238
00:08:20,280 --> 00:08:21,640
And I sometimes just copy paste

239
00:08:21,640 --> 00:08:22,560
or have it analyze something

240
00:08:22,560 --> 00:08:23,220
that it recently did.

241
00:08:23,220 --> 00:08:24,360
Or I say, hey, assess this thing

242
00:08:24,360 --> 00:08:25,220
we're working on.

243
00:08:25,220 --> 00:08:28,700
and give me feedback, give me suggestions, give me ideas.

244
00:08:29,260 --> 00:08:31,040
And I tell Gemini, you're not going to build this.

245
00:08:31,140 --> 00:08:32,240
You're just going to give us ideas.

246
00:08:32,640 --> 00:08:35,500
And I'm going to take those ideas and give them to Claude to actually build.

247
00:08:35,900 --> 00:08:37,920
So that was kind of my workflow for many months.

248
00:08:38,440 --> 00:08:40,540
And that is, I've been generally happy with it.

249
00:08:40,540 --> 00:08:44,660
I dabbled with the ChatGPT codex CLI stuff, you know,

250
00:08:44,800 --> 00:08:46,720
maybe three or four or five months ago.

251
00:08:46,960 --> 00:08:48,600
Never really got into it.

252
00:08:48,660 --> 00:08:49,560
And I was just like, eh, whatever.

253
00:08:49,940 --> 00:08:51,980
Just keep with stuff, you know, stuff that works.

254
00:08:51,980 --> 00:08:58,400
and then um a few actually it was the day i were driving on a couple days ago so it was wednesday

255
00:08:58,400 --> 00:09:07,680
and i uh i set up the uh codex cli so this is the one from open ai chat gpt their coding tool

256
00:09:07,680 --> 00:09:14,600
and it is just ripping through problems like the problems that i was getting stuck with with

257
00:09:14,600 --> 00:09:19,800
cloud code or even with the the gut check of like gemini trying to assess it and give feedback

258
00:09:19,800 --> 00:09:22,880
Claude Codex now

259
00:09:22,880 --> 00:09:24,580
was just like ripping through them.

260
00:09:24,720 --> 00:09:26,180
And the way I describe it,

261
00:09:26,260 --> 00:09:27,720
it's like a dog on a bone.

262
00:09:27,920 --> 00:09:29,060
It just keeps going.

263
00:09:29,140 --> 00:09:30,120
And it's like persistent.

264
00:09:30,340 --> 00:09:31,100
It won't give up.

265
00:09:31,180 --> 00:09:32,240
And it keeps...

266
00:09:32,240 --> 00:09:33,140
Do you want to know how

267
00:09:33,140 --> 00:09:37,200
Callie described Claude versus Codex?

268
00:09:37,500 --> 00:09:39,200
I saw an inappropriate meme.

269
00:09:39,320 --> 00:09:40,740
I don't know if this was nice, but...

270
00:09:40,740 --> 00:09:41,420
Well, he said, like,

271
00:09:41,500 --> 00:09:42,640
Claude's like an American

272
00:09:42,640 --> 00:09:44,680
and Codex is like a German.

273
00:09:45,120 --> 00:09:46,040
Oh, yeah, totally.

274
00:09:46,380 --> 00:09:47,160
Totally makes sense.

275
00:09:47,600 --> 00:09:48,960
I don't know if that matches your experience.

276
00:09:48,960 --> 00:09:55,340
it it does yeah it it's and i've seen it said like you know claude is like your exploration

277
00:09:55,340 --> 00:10:00,620
brainstorm partner and then and then codex is like your autistic friend who just like you know

278
00:10:00,620 --> 00:10:06,140
keeps going and um and then people say actually to use gemini is sort of like the reviewer so it's

279
00:10:06,140 --> 00:10:11,080
almost like the creativity the execution and the review and they're very like it all sounds

280
00:10:11,080 --> 00:10:18,640
similar but they're very different tools um but i i just got so i've just been so codex pilled this

281
00:10:18,640 --> 00:10:26,620
week and then oh it's insanely better than the alternatives now at this moment at this moment

282
00:10:26,620 --> 00:10:32,500
it can change but for execution for hey i need you to build this specific set of requirements

283
00:10:32,500 --> 00:10:37,180
what i did is i had already basically built the general app i think steve may have already tried

284
00:10:37,180 --> 00:10:42,700
it um in in like the thing kind of like it's almost like it has the scaffolding and the ui

285
00:10:42,700 --> 00:10:48,020
and it kind of like looks about right but so much of it is very broken underneath the hood like it

286
00:10:48,020 --> 00:10:53,060
doesn't actually do the thing i wanted to do so once it's in that state is when i brought it to

287
00:10:53,060 --> 00:10:57,960
codex not because i knew that that was the best thing that it could do but that turned out to be

288
00:10:57,960 --> 00:11:03,660
where i was when i kind of brought it to codex and then codex is like refactoring things like i've

289
00:11:03,660 --> 00:11:07,520
asked cloud code to refactor things and come up with better ways to do it and think about the

290
00:11:07,520 --> 00:11:11,640
architecture now that we've done all this work together let's like really reevaluate and see if

291
00:11:11,640 --> 00:11:14,440
we would do it the same way if we were to do it again like i give it prompts like that all the

292
00:11:14,440 --> 00:11:16,500
time. But Cloud Code is a little

293
00:11:16,500 --> 00:11:18,280
more like in brainstorm

294
00:11:18,280 --> 00:11:20,440
thinking creativity mode, which is good.

295
00:11:20,500 --> 00:11:22,460
That's a valuable thing to have. It wants to

296
00:11:22,460 --> 00:11:24,500
hold some space first to

297
00:11:24,500 --> 00:11:25,760
make sure everyone's on this page. Right, right.

298
00:11:27,520 --> 00:11:28,300
Where's Elon

299
00:11:28,300 --> 00:11:30,500
in all this? Grok? I haven't

300
00:11:30,500 --> 00:11:32,600
heard anything from Grok in like six plus

301
00:11:32,600 --> 00:11:34,420
months. I was just talking about this with Enjoy last

302
00:11:34,420 --> 00:11:36,380
night. I mean, I'm still long Elon

303
00:11:36,380 --> 00:11:38,440
in the long run because of the energy, but like, what's going on?

304
00:11:38,680 --> 00:11:40,480
Because what was it, six, eight months ago

305
00:11:40,480 --> 00:11:42,620
they had a big announcement and it seemed

306
00:11:42,620 --> 00:11:46,240
or you certainly said that they made big gains.

307
00:11:46,320 --> 00:11:48,180
And at the time, I thought it was somewhat competitive.

308
00:11:48,380 --> 00:11:49,600
So I started using it then.

309
00:11:50,540 --> 00:11:53,520
But I only use it now in my car

310
00:11:53,520 --> 00:11:54,800
because that's what's integrated.

311
00:11:54,920 --> 00:11:55,820
Yeah, exactly, because it's there.

312
00:11:56,280 --> 00:11:57,900
I mean, so I don't know.

313
00:11:58,260 --> 00:11:59,380
And I have a lot of questions about this.

314
00:11:59,620 --> 00:12:00,540
The meme that I saw,

315
00:12:00,820 --> 00:12:02,560
I saw someone have kind of like Claude

316
00:12:02,560 --> 00:12:04,400
had like the rainbow hair and like the piercing

317
00:12:04,400 --> 00:12:07,120
and like the Codex guy was like all in black,

318
00:12:07,200 --> 00:12:08,720
maybe like German, just like efficiency,

319
00:12:09,080 --> 00:12:09,800
which is interesting to me.

320
00:12:09,800 --> 00:12:12,380
But at the same time,

321
00:12:12,380 --> 00:12:14,500
I mean, I need to start playing with Cloud Code directly

322
00:12:14,500 --> 00:12:17,020
because I've used Gemini as my daily driver,

323
00:12:17,240 --> 00:12:18,820
just for like, Gemini is my like,

324
00:12:18,920 --> 00:12:21,060
sort of, I'd say a thought partner, business partner,

325
00:12:21,200 --> 00:12:22,280
like for all things business,

326
00:12:22,500 --> 00:12:24,280
like Gemini and I roll together.

327
00:12:25,460 --> 00:12:28,840
And I guess I've been using Cloud via OriBot

328
00:12:28,840 --> 00:12:30,700
or any kind of like building stuff

329
00:12:30,700 --> 00:12:31,840
or whatever Nick is using the back.

330
00:12:31,880 --> 00:12:33,700
And I guess he could plug in Codex.

331
00:12:35,080 --> 00:12:38,100
But yeah, but I haven't really used,

332
00:12:38,100 --> 00:12:40,300
is your view, because I see Vinny talks about

333
00:12:40,300 --> 00:12:41,780
how Cloud was always more philosophical as well.

334
00:12:41,780 --> 00:12:46,160
is your view that like imagine you weren't even coding yet you were just like i don't know like

335
00:12:46,160 --> 00:12:50,420
discussing yeah like an idea like a philosophical idea may or may not lead to a project would you

336
00:12:50,420 --> 00:12:57,940
choose claude just to have as a thought partner or gemini in that world based on what i know right

337
00:12:57,940 --> 00:13:02,980
now i i'd probably bias a little toward claude for that kind of just general thought partner

338
00:13:02,980 --> 00:13:09,400
you know think about an idea um i i do have i do have gemini do that kind of stuff from time to

339
00:13:09,400 --> 00:13:12,700
time, especially when I think Gemini may have an advantage. Like I was trying to produce a little

340
00:13:12,700 --> 00:13:18,680
explainer video to sort of, you know, intro the app idea to people so they can sort of see what

341
00:13:18,680 --> 00:13:23,040
it is and see how it works and kind of be inspired about what it's going to, what functions can serve,

342
00:13:23,100 --> 00:13:28,720
which I thought Gemini given VO and the video stuff would be more suited to that. I brought

343
00:13:28,720 --> 00:13:33,500
that idea to Gemini. It gave me like a really great, like some, some different script ideas

344
00:13:33,500 --> 00:13:38,260
for how that video should be shot. We talked through doing some live action versus motion

345
00:13:38,260 --> 00:13:42,240
graphics stuff. And then it ended up just saying, you know what, I think I'm just going to implement

346
00:13:42,240 --> 00:13:45,940
this in Swift and it's not going to be a video. It's going to be an animation I'm going to implement in Swift. Is that okay?

347
00:13:46,040 --> 00:13:50,180
And I said, yeah, go ahead. And it did it. And I liked its result with like almost no

348
00:13:50,180 --> 00:13:54,100
none of my kind of curation. It kind of just did something reasonable.

349
00:13:54,540 --> 00:13:58,480
But then I've been iterating on that and it still, it seems to do a really good job.

350
00:13:58,480 --> 00:14:02,460
So I think Gemini is very good as a creative thought partner still.

351
00:14:02,560 --> 00:14:03,880
I'm not opposed to it, but I'd probably

352
00:14:03,880 --> 00:14:06,500
probably biased toward starting with

353
00:14:06,500 --> 00:14:08,800
CloudCode for kind of that brainstorming

354
00:14:08,800 --> 00:14:10,720
process. Also, if you haven't already

355
00:14:10,720 --> 00:14:13,200
done so, you can install CloudCode CLI,

356
00:14:13,920 --> 00:14:15,000
Codex CLI,

357
00:14:15,500 --> 00:14:16,840
Goose, all of them

358
00:14:16,840 --> 00:14:18,640
in under 10 minutes and be running

359
00:14:18,640 --> 00:14:19,720
all of them. At once.

360
00:14:20,320 --> 00:14:22,460
Yeah. So it's super, I mean,

361
00:14:22,460 --> 00:14:24,380
so just get started. And then

362
00:14:24,380 --> 00:14:26,260
for each, you know,

363
00:14:26,460 --> 00:14:28,380
just do the $20 plans or whatever.

364
00:14:28,600 --> 00:14:30,540
So it's like, and then if you

365
00:14:30,540 --> 00:14:32,520
find one that really does

366
00:14:32,520 --> 00:14:35,300
Well, and then you want to max it out, then you can bump it up.

367
00:14:35,520 --> 00:14:40,340
But I'd say purchase the minimum plan for all of them and go for it.

368
00:14:40,400 --> 00:14:42,800
Well, I need to try Claude for writing.

369
00:14:43,180 --> 00:14:45,860
I've got the maxed out Gemini one as a sort of thought part.

370
00:14:45,900 --> 00:14:48,560
I've got these essays that I haven't been able to get.

371
00:14:48,860 --> 00:14:49,860
I had a little bit of writer's blog.

372
00:14:49,860 --> 00:14:52,520
I've got two big ideas I really want to get out there.

373
00:14:52,740 --> 00:14:59,180
But actually, this is an interesting sort of meta point where I just like certain –

374
00:14:59,180 --> 00:15:01,440
I mean, it's interesting because I'm not like a software engineer per se,

375
00:15:01,440 --> 00:15:03,420
but I always have been writing, right?

376
00:15:03,480 --> 00:15:05,480
I've been a writer, I would say, my whole life.

377
00:15:06,060 --> 00:15:08,080
And, you know, I've always been in the camp of like,

378
00:15:08,120 --> 00:15:09,360
oh, I'd never let AI write for me

379
00:15:09,360 --> 00:15:11,160
because I think good writing is just good thinking

380
00:15:11,160 --> 00:15:11,840
and I want to have that.

381
00:15:12,400 --> 00:15:15,180
And, you know, AI has certainly been a useful research partner,

382
00:15:15,260 --> 00:15:17,280
but I have noticed myself kind of a couple of times now

383
00:15:17,280 --> 00:15:19,960
trying to get this idea out into the world

384
00:15:19,960 --> 00:15:21,700
where I'm just like, okay, like Gemini,

385
00:15:21,960 --> 00:15:24,000
like, you know, whatever, like deep think this

386
00:15:24,000 --> 00:15:25,560
and then feeding that into a deep research

387
00:15:25,560 --> 00:15:26,660
and then back into a deep thing.

388
00:15:27,060 --> 00:15:30,000
And it's good, but it's still not exactly what I want.

389
00:15:30,000 --> 00:15:39,000
And I feel like I want to be careful not to get too dependent because I do think like really quickly all of our skills can just sort of like atrophy and that's that.

390
00:15:39,740 --> 00:15:41,400
So I'm – anyways, it's something I'm thinking about.

391
00:15:41,580 --> 00:15:44,840
I think that concern – I hear that concern a lot, but I think that's overstated.

392
00:15:45,300 --> 00:15:45,620
Why?

393
00:15:46,620 --> 00:15:53,380
Well, I think the skill that's relevant today is how you think and how you use these tools to make yourself better.

394
00:15:53,380 --> 00:15:58,220
so if you if you're just like if you're totally outsourcing your thing just like give it a prompt

395
00:15:58,220 --> 00:16:03,500
just copy paste whatever it did and post it or send it or share it that's obviously kind of a

396
00:16:03,500 --> 00:16:10,300
bad idea but if you're saying look i need to be like independent thinker that doesn't use these

397
00:16:10,300 --> 00:16:14,220
tools i think you're actually hurting yourself i think you need to sort of figure out how to use

398
00:16:14,220 --> 00:16:20,040
the tools selectively and your own understanding of what you what you expect out of it not for it

399
00:16:20,040 --> 00:16:21,540
to whole cloth produce the thing,

400
00:16:22,000 --> 00:16:23,140
but for it to be a partner.

401
00:16:23,660 --> 00:16:24,600
No, no, I agree.

402
00:16:24,700 --> 00:16:26,220
And I'm certainly not suggesting I get rid of it.

403
00:16:26,260 --> 00:16:27,320
It's been extremely useful to me.

404
00:16:27,420 --> 00:16:28,120
What I'm thinking is like,

405
00:16:28,160 --> 00:16:29,940
how do I redo my workflow where it's like,

406
00:16:30,400 --> 00:16:31,940
instead of getting too dependent and being like,

407
00:16:31,980 --> 00:16:33,200
hey, crunch this out a few times,

408
00:16:33,460 --> 00:16:35,660
it might be more valuable what I'm playing with of like,

409
00:16:35,720 --> 00:16:37,660
okay, maybe we do a little conversation back and forth

410
00:16:37,660 --> 00:16:38,940
and maybe I try that with Claude now.

411
00:16:39,300 --> 00:16:41,040
I like write a first draft first.

412
00:16:41,040 --> 00:16:42,540
I think it's important at some point,

413
00:16:42,680 --> 00:16:45,500
I like have to like force the ideas onto paper.

414
00:16:45,640 --> 00:16:47,520
Then maybe I take it back, have it refine it.

415
00:16:47,880 --> 00:16:49,560
I think I'm just trying to figure out

416
00:16:50,040 --> 00:16:54,040
you know like anyone like we're so early with these tools what's the right like what's the

417
00:16:54,040 --> 00:16:59,280
right workflow what's the right process um when am i doing because like i still do get a lot of

418
00:16:59,280 --> 00:17:04,440
value like there's still an incredible value for me to have a period where it's like okay there's

419
00:17:04,440 --> 00:17:08,640
no internet you're like yeah you and a piece of fucking paper and you're like have to like

420
00:17:08,640 --> 00:17:12,940
i don't know like there's there's a that's a different set of thinking for me than like oh

421
00:17:12,940 --> 00:17:17,380
now i'm going back and forth and like kind of shallow going back with research but putting the

422
00:17:17,380 --> 00:17:18,800
time in to figure that out is

423
00:17:18,800 --> 00:17:19,620
valuable.

424
00:17:20,800 --> 00:17:23,260
Related, yesterday a group of university

425
00:17:23,260 --> 00:17:24,780
students came by Presidia Bitcoin

426
00:17:24,780 --> 00:17:27,340
from Austin's alma mater.

427
00:17:27,660 --> 00:17:29,420
Oh yeah. Elon University.

428
00:17:30,940 --> 00:17:31,660
There's an

429
00:17:31,660 --> 00:17:33,760
Elon, North Carolina.

430
00:17:34,400 --> 00:17:35,700
I thought that was, what's his name?

431
00:17:35,780 --> 00:17:36,620
The ABC did, yeah.

432
00:17:37,900 --> 00:17:39,620
Anyway, they were

433
00:17:39,620 --> 00:17:41,120
entrepreneurship

434
00:17:41,120 --> 00:17:43,600
students across a number of

435
00:17:43,600 --> 00:17:45,680
disciplines, but they're all majoring

436
00:17:45,680 --> 00:17:47,280
or minoring in entrepreneurship as well.

437
00:17:47,380 --> 00:17:50,440
but I love speaking to students.

438
00:17:50,900 --> 00:17:53,860
And so Jesse and myself and Austin,

439
00:17:54,140 --> 00:17:54,860
I don't know who else,

440
00:17:54,980 --> 00:17:57,360
but at least when I was up speaking to them

441
00:17:57,360 --> 00:17:58,240
and answering questions,

442
00:17:58,340 --> 00:17:59,200
it was the three of us.

443
00:18:00,160 --> 00:18:02,380
And it was, I mean,

444
00:18:02,500 --> 00:18:06,080
none or maybe one was a Bitcoiner,

445
00:18:06,260 --> 00:18:08,740
so really, but Austin wanted us

446
00:18:08,740 --> 00:18:09,820
to give like career advice.

447
00:18:10,120 --> 00:18:11,720
And it certainly, I mean,

448
00:18:11,740 --> 00:18:13,960
I gave some advice that I've given historically,

449
00:18:13,960 --> 00:18:15,240
but also, what was it?

450
00:18:15,240 --> 00:18:27,760
Well, the historical, I mean, single piece of advice in terms of getting a job or whatever is like, and DK knows this as well because we've talked about this a lot, but I've never applied for a job.

451
00:18:28,980 --> 00:18:32,740
There's never been a job where I apply and there's thousands of candidates.

452
00:18:35,320 --> 00:18:38,160
Historically, I discover a job.

453
00:18:38,280 --> 00:18:39,500
I create a job.

454
00:18:39,500 --> 00:18:41,600
I convince someone to create a job.

455
00:18:41,980 --> 00:18:43,160
Steve is not being the best.

456
00:18:43,260 --> 00:18:44,000
He's being the only.

457
00:18:44,000 --> 00:18:50,160
yeah yeah it's way easier to win if you're the only player kind of like when something sometimes

458
00:18:50,160 --> 00:18:55,320
it's like yeah i i won my running race or whatever and it's like well i was the only one in my age

459
00:18:55,320 --> 00:19:05,800
category so yeah unfair advantages we love those so um so that that was i mean i went into details

460
00:19:05,800 --> 00:19:10,640
but that was one piece of advice but the more but like the the advice that i wouldn't even have

461
00:19:10,640 --> 00:19:12,600
known to have given a year ago was

462
00:19:12,600 --> 00:19:13,700
what we're talking about now, but

463
00:19:13,700 --> 00:19:16,600
utilize AI tools. I mean, I tried to

464
00:19:16,600 --> 00:19:18,540
give an optimistic message, which is

465
00:19:18,540 --> 00:19:19,300
like,

466
00:19:19,360 --> 00:19:22,500
there's still going to be plenty

467
00:19:22,500 --> 00:19:24,200
of opportunities in the future. Totally.

468
00:19:24,520 --> 00:19:26,500
But you need to be on the AI

469
00:19:26,500 --> 00:19:28,320
train and building up the skills now.

470
00:19:28,420 --> 00:19:29,360
That was a really good

471
00:19:29,360 --> 00:19:32,040
discussion. What was your sense?

472
00:19:32,240 --> 00:19:34,500
Were these kids already doing that

473
00:19:34,500 --> 00:19:36,100
or was this like, shocker for them?

474
00:19:36,700 --> 00:19:38,600
They were not doing that. They were not doing it.

475
00:19:38,600 --> 00:19:40,340
It was a little surprising. I feel like

476
00:19:40,640 --> 00:19:40,900
Wow.

477
00:19:41,080 --> 00:19:41,640
Yeah, I'm surprised.

478
00:19:41,780 --> 00:19:41,980
Kids.

479
00:19:42,300 --> 00:19:44,420
Like us and people at Presidio Bitcoin.

480
00:19:44,420 --> 00:19:47,220
And I think there's still,

481
00:19:47,480 --> 00:19:53,300
that was a data point that we're still very much a small group.

482
00:19:53,880 --> 00:19:57,600
I mean, you went to Louisiana recently and talked about that.

483
00:19:57,760 --> 00:20:01,560
But here's a group of like 20,

484
00:20:01,840 --> 00:20:04,080
actively trying to burn down the data.

485
00:20:04,360 --> 00:20:09,420
22-year-old students who definitely didn't,

486
00:20:09,420 --> 00:20:11,480
the way they're asking questions.

487
00:20:11,960 --> 00:20:14,900
I mean, they were enthusiastic

488
00:20:14,900 --> 00:20:16,940
and I think they really appreciated the advice.

489
00:20:17,200 --> 00:20:20,000
So I think they're going to get on the AI train,

490
00:20:20,080 --> 00:20:20,840
but they were not already.

491
00:20:21,160 --> 00:20:22,200
That was my impression.

492
00:20:23,000 --> 00:20:23,440
So that's one thing.

493
00:20:23,440 --> 00:20:25,020
I also think sometimes when people think

494
00:20:25,020 --> 00:20:26,240
they're on the AI train,

495
00:20:26,440 --> 00:20:28,280
they're sort of using it as if

496
00:20:28,280 --> 00:20:30,620
how they might have used Google historically,

497
00:20:30,980 --> 00:20:33,480
which is like information retrieval use cases,

498
00:20:33,560 --> 00:20:35,060
which is like, it's good and it's better.

499
00:20:35,840 --> 00:20:38,400
LLMs are better than kind of traditional web search,

500
00:20:38,400 --> 00:20:40,520
but that's only scratching the surface.

501
00:20:40,820 --> 00:20:42,520
I'd really want to encourage them to explore

502
00:20:42,520 --> 00:20:45,380
your workflows or stuff you need to do.

503
00:20:45,540 --> 00:20:46,960
How do you do those things with AI?

504
00:20:47,520 --> 00:20:48,140
Totally agree.

505
00:20:48,700 --> 00:20:50,640
I did that a little bit and then Jesse

506
00:20:50,640 --> 00:20:53,260
went into way more elaborate detail

507
00:20:53,260 --> 00:20:54,880
which was good on that point.

508
00:20:55,360 --> 00:20:57,360
I want to real quick, just some of the comments here.

509
00:20:58,200 --> 00:20:59,540
Matt Coat, you mentioned

510
00:20:59,540 --> 00:21:00,840
you missed last week.

511
00:21:01,440 --> 00:21:02,380
Totally understandable.

512
00:21:02,760 --> 00:21:05,980
Feel for you, man, about the block situation.

513
00:21:05,980 --> 00:21:09,040
but glad to have you back here listening

514
00:21:09,040 --> 00:21:12,740
and hopefully you gain some inspiration from our show

515
00:21:12,740 --> 00:21:14,560
and can't wait to see what you do next.

516
00:21:14,740 --> 00:21:15,540
That would be sweet.

517
00:21:15,900 --> 00:21:17,820
We know you've already got great ideas

518
00:21:17,820 --> 00:21:19,620
and great taste around music and Bitcoin.

519
00:21:19,900 --> 00:21:20,620
The world needs it.

520
00:21:21,340 --> 00:21:22,100
Now's your chance.

521
00:21:22,180 --> 00:21:22,820
Come do some dope.

522
00:21:23,320 --> 00:21:23,460
Yep.

523
00:21:24,220 --> 00:21:26,600
So do let us know what you're up to

524
00:21:26,600 --> 00:21:27,800
and we'll follow along.

525
00:21:29,240 --> 00:21:29,780
Let's see.

526
00:21:29,880 --> 00:21:30,580
What else here?

527
00:21:33,220 --> 00:21:33,520
I guess.

528
00:21:33,780 --> 00:21:34,360
Other comments?

529
00:21:34,560 --> 00:21:35,700
I can't see the chat.

530
00:21:35,980 --> 00:21:38,540
yeah I think that's good

531
00:21:38,540 --> 00:21:40,260
on that the other thing I was going to mention

532
00:21:40,260 --> 00:21:40,880
regarding

533
00:21:40,880 --> 00:21:44,580
you know how to use AI

534
00:21:44,580 --> 00:21:46,480
and like your workflow and you don't want to lose

535
00:21:46,480 --> 00:21:48,440
your independence or create your own

536
00:21:48,440 --> 00:21:49,100
creative mind or whatever

537
00:21:49,100 --> 00:21:52,560
I think maybe I briefly mentioned

538
00:21:52,560 --> 00:21:54,440
last week's show about an

539
00:21:54,440 --> 00:21:56,600
AI hackathon that's

540
00:21:56,600 --> 00:21:58,440
going to be next month that

541
00:21:58,440 --> 00:22:00,700
both Spiral and Presidia

542
00:22:00,700 --> 00:22:02,580
Bitcoin are exploring

543
00:22:02,580 --> 00:22:04,660
partnering with and

544
00:22:04,660 --> 00:22:09,360
And I had a call with the organizer a week ago.

545
00:22:10,240 --> 00:22:11,820
And then she came to Presidio.

546
00:22:11,860 --> 00:22:12,520
She's in the Bay Area.

547
00:22:12,620 --> 00:22:16,260
So she came to Presidio Bitcoin yesterday and spent several hours with us.

548
00:22:16,620 --> 00:22:17,980
Max joined and so did Haley.

549
00:22:18,160 --> 00:22:19,560
So the three of us met with her.

550
00:22:20,360 --> 00:22:21,480
It was great.

551
00:22:21,820 --> 00:22:24,920
I think something's going to happen there.

552
00:22:26,160 --> 00:22:29,600
Once that becomes more concrete, we'll certainly announce it.

553
00:22:29,600 --> 00:22:33,500
But we discussed potentially having Presidio Bitcoin be a hub.

554
00:22:33,500 --> 00:22:35,840
because this is a global hackathon

555
00:22:35,840 --> 00:22:39,900
with numerous global physical hub locations,

556
00:22:40,320 --> 00:22:41,480
but also it's virtual.

557
00:22:41,580 --> 00:22:42,640
So you can be anywhere and participate,

558
00:22:42,840 --> 00:22:44,860
but if you happen to be near travel to a hub,

559
00:22:44,900 --> 00:22:45,600
you can go to a hub.

560
00:22:46,160 --> 00:22:47,140
There'll be one at Stanford.

561
00:22:47,660 --> 00:22:48,780
I think there's always been one.

562
00:22:48,900 --> 00:22:50,660
This will be their fourth or fifth hackathon.

563
00:22:51,400 --> 00:22:52,600
And there's always been one at Stanford,

564
00:22:53,120 --> 00:22:56,920
but she seemed very open to having one here in San Francisco,

565
00:22:57,060 --> 00:22:57,540
which is awesome.

566
00:22:59,180 --> 00:23:02,560
And then, you know, I think, Max,

567
00:23:02,560 --> 00:23:04,560
we had a great conversation with her because

568
00:23:04,560 --> 00:23:06,340
she's not a Bitcoiner

569
00:23:06,340 --> 00:23:08,260
and she

570
00:23:08,260 --> 00:23:10,060
you know so she

571
00:23:10,060 --> 00:23:11,380
knows a little bit

572
00:23:11,380 --> 00:23:14,480
based on her experience there's quite a bit of

573
00:23:14,480 --> 00:23:16,200
skepticism and questions and stuff

574
00:23:16,200 --> 00:23:17,420
and even from

575
00:23:17,420 --> 00:23:20,500
but she's very open minded as well

576
00:23:20,500 --> 00:23:22,320
so she's been asking

577
00:23:22,320 --> 00:23:23,680
great questions and learning

578
00:23:23,680 --> 00:23:26,460
from us and I think she

579
00:23:26,460 --> 00:23:42,129
I get the sense she thinks we you know coming at it from an authentic genuine way she just wants to make sure that her audience the people who participate are presented with Bitcoin in a good way

580
00:23:42,329 --> 00:23:43,909
Much like we've talked about on the show,

581
00:23:44,429 --> 00:23:47,069
in terms of engaging the broader tech community,

582
00:23:47,209 --> 00:23:48,149
broader AI community,

583
00:23:48,309 --> 00:23:50,929
sometimes people can bristle if they hear the word Bitcoin.

584
00:23:51,929 --> 00:23:54,409
SBF is conjured up in their head or whatever.

585
00:23:54,789 --> 00:23:56,889
Which literally, and to her credit,

586
00:23:56,889 --> 00:23:58,209
she came in with a pretty open mind,

587
00:23:58,269 --> 00:23:59,569
which I was impressed that, you know,

588
00:23:59,569 --> 00:24:01,189
she was willing to like actually engage in this stuff.

589
00:24:01,529 --> 00:24:03,349
But like, it was kind of crazy to me,

590
00:24:03,369 --> 00:24:05,029
like hearing her like bring up, you know,

591
00:24:05,049 --> 00:24:07,029
something like, oh, like what do you think of,

592
00:24:07,029 --> 00:24:07,649
what do you think of Bitcoin?

593
00:24:07,789 --> 00:24:09,989
And it's like, energy FUD, SBF.

594
00:24:10,729 --> 00:24:13,989
The professors from college said it was insecure.

595
00:24:14,349 --> 00:24:14,909
It's kind of great.

596
00:24:15,089 --> 00:24:16,109
It's just kind of crazy to me.

597
00:24:16,169 --> 00:24:17,369
It's like, this is 2026.

598
00:24:17,529 --> 00:24:19,809
And like, clearly this is a smart person at Google,

599
00:24:19,909 --> 00:24:22,469
whatever, but like, and she was very open

600
00:24:22,469 --> 00:24:23,569
when we explained these things to her.

601
00:24:23,889 --> 00:24:25,469
But like the fact that people are still being like,

602
00:24:25,469 --> 00:24:27,709
oh, like that scammy guy from MIT,

603
00:24:27,949 --> 00:24:29,429
it's like, yeah, but that has nothing to do with Bitcoin.

604
00:24:30,549 --> 00:24:35,049
So we're early.

605
00:24:35,289 --> 00:24:36,009
That was my takeaway.

606
00:24:36,209 --> 00:24:39,529
Well, but I mean, it's of our interest,

607
00:24:39,709 --> 00:24:42,029
this show's interest and preceding Bitcoin's interest

608
00:24:42,029 --> 00:24:45,989
to try to bridge the gap in Silicon Valley, right?

609
00:24:46,049 --> 00:24:47,289
That's our tagline for the show.

610
00:24:48,269 --> 00:24:49,269
So this is very much,

611
00:24:49,969 --> 00:24:54,069
and so we're well aware of the challenges and hurdles there.

612
00:24:54,069 --> 00:24:57,969
we've made modest strides towards that in the past year

613
00:24:57,969 --> 00:25:00,149
we need to make way bigger strides

614
00:25:00,149 --> 00:25:02,889
so I really like this opportunity

615
00:25:02,889 --> 00:25:07,889
I told Max previously that

616
00:25:07,889 --> 00:25:09,709
usually when spirals

617
00:25:09,709 --> 00:25:12,389
when there's opportunity to sponsor an event

618
00:25:12,389 --> 00:25:15,429
usually it's just like we evaluate

619
00:25:15,429 --> 00:25:18,769
is it something we want to sponsor and if so we just write a check and that's it

620
00:25:18,769 --> 00:25:20,589
they're happy to take our money

621
00:25:20,589 --> 00:25:24,349
where my call with her last week,

622
00:25:24,429 --> 00:25:28,049
she's grilling me and she wants to know how to position this

623
00:25:28,049 --> 00:25:30,549
and she's talking about communication strategy

624
00:25:30,549 --> 00:25:33,629
and she wants a super high quality challenge

625
00:25:33,629 --> 00:25:34,969
and I'm like, this is amazing.

626
00:25:35,969 --> 00:25:38,609
It's going to level me up, level us up.

627
00:25:40,189 --> 00:25:41,649
I'm really looking forward to that.

628
00:25:42,189 --> 00:25:46,269
But the way it's structured

629
00:25:46,269 --> 00:25:48,669
is that there's different challenges.

630
00:25:49,369 --> 00:25:52,849
So the hackathon participants don't just show up and work on whatever they want.

631
00:25:53,269 --> 00:25:57,609
There's something like a dozen or so structured challenges that they present.

632
00:25:57,889 --> 00:26:01,569
And then as a participant, you have to pick one of them and then you work on it.

633
00:26:02,469 --> 00:26:09,409
So it's scoped enough that you have some box you have to work within, but it's not too small a box.

634
00:26:09,409 --> 00:26:13,709
You have a lot of latitude within that box to build something.

635
00:26:14,469 --> 00:26:17,349
So the challenge for us is to create a challenge.

636
00:26:17,349 --> 00:26:19,369
What is the Bitcoin challenge? What does it look like?

637
00:26:20,129 --> 00:26:21,649
I'm bringing this up in the context

638
00:26:21,649 --> 00:26:23,409
for a conversation about AI because

639
00:26:23,409 --> 00:26:25,469
I utilize AI tools

640
00:26:25,469 --> 00:26:27,269
to help figure out

641
00:26:27,269 --> 00:26:28,829
brainstorm challenges, right?

642
00:26:30,509 --> 00:26:31,429
Spiral's working on

643
00:26:31,429 --> 00:26:32,189
a presentation

644
00:26:32,189 --> 00:26:35,549
around agentic payments in Bitcoin.

645
00:26:36,869 --> 00:26:37,669
I plan to give

646
00:26:37,669 --> 00:26:39,389
the big reveal

647
00:26:39,389 --> 00:26:41,309
of that will be next Wednesday

648
00:26:41,309 --> 00:26:43,169
in New York at the Builder we're going to

649
00:26:43,169 --> 00:26:44,969
host there. Is that going to be the last year?

650
00:26:45,309 --> 00:26:47,109
I don't know. It'd be cool if

651
00:26:47,109 --> 00:26:47,709
at least recorded.

652
00:26:48,129 --> 00:26:49,389
Yeah, I'll try to get it recorded.

653
00:26:50,209 --> 00:26:51,009
But, you know,

654
00:26:51,069 --> 00:26:53,529
so it'll be the first attempt

655
00:26:53,529 --> 00:26:54,389
at that presentation.

656
00:26:55,189 --> 00:26:56,189
Also, the builder events,

657
00:26:56,269 --> 00:26:57,549
we like to, you know,

658
00:26:57,789 --> 00:26:58,669
I don't want to present it

659
00:26:58,669 --> 00:27:00,469
as if I have all the answers

660
00:27:00,469 --> 00:27:01,989
and like, here are the answers,

661
00:27:02,149 --> 00:27:03,209
go make it happen.

662
00:27:03,389 --> 00:27:04,089
It's more like,

663
00:27:04,449 --> 00:27:06,609
here's a vision of the future

664
00:27:06,609 --> 00:27:10,269
and some ideas

665
00:27:10,269 --> 00:27:12,569
and here's three areas

666
00:27:12,569 --> 00:27:13,789
where we're still sort of uncertain.

667
00:27:13,909 --> 00:27:15,109
Let's have a discussion, right?

668
00:27:15,109 --> 00:27:16,149
Like in a builder meetup,

669
00:27:16,149 --> 00:27:18,589
we want the audience, everyone to discuss.

670
00:27:18,829 --> 00:27:21,209
So that's what my expectation for that is.

671
00:27:21,269 --> 00:27:23,169
Anyway, we have that draft presentation.

672
00:27:23,609 --> 00:27:25,149
We have a bunch of other content we have.

673
00:27:26,249 --> 00:27:29,469
And then I fed that and all the information I had

674
00:27:29,469 --> 00:27:31,569
about the hackathon into AI

675
00:27:31,569 --> 00:27:34,229
and then asked it to generate,

676
00:27:34,549 --> 00:27:38,929
basically brainstorm some moonshot-like challenges.

677
00:27:39,169 --> 00:27:40,029
Jim and I are which model?

678
00:27:40,629 --> 00:27:41,589
I did several.

679
00:27:43,509 --> 00:27:45,329
Chad, GBD, Jim and I, and Claude.

680
00:27:46,149 --> 00:27:48,429
similar types of moonshot ideas

681
00:27:48,429 --> 00:27:49,969
or were they very diverse?

682
00:27:50,649 --> 00:27:51,389
Pretty diverse.

683
00:27:51,729 --> 00:27:52,829
And then I'd have them check each other

684
00:27:52,829 --> 00:27:53,609
and improve each other.

685
00:27:53,869 --> 00:27:54,389
So that was a good,

686
00:27:54,689 --> 00:27:55,689
I mean, getting back to like

687
00:27:55,689 --> 00:27:57,409
which tool to use or how to use it.

688
00:27:57,589 --> 00:27:58,749
Like, again, I think,

689
00:27:59,209 --> 00:28:00,509
I mean, I certainly learn from others.

690
00:28:00,629 --> 00:28:02,329
I get inspired by like DK

691
00:28:02,329 --> 00:28:04,329
and Jesse Posner and others.

692
00:28:05,869 --> 00:28:08,269
But then nothing replaces

693
00:28:08,269 --> 00:28:09,569
just doing it yourself

694
00:28:09,569 --> 00:28:11,149
and getting firsthand experience.

695
00:28:11,409 --> 00:28:13,989
So, but it spit out a bunch of ideas.

696
00:28:13,989 --> 00:28:15,709
And I think this is a good example of like,

697
00:28:15,709 --> 00:28:18,309
you can't just like blindly

698
00:28:18,309 --> 00:28:20,869
go with what the AI says,

699
00:28:21,009 --> 00:28:22,129
but it's an incredible

700
00:28:22,129 --> 00:28:23,829
accelerant and tool

701
00:28:23,829 --> 00:28:24,889
and helper.

702
00:28:26,989 --> 00:28:27,869
It's a thought partner.

703
00:28:28,069 --> 00:28:29,509
And one of the many ideas

704
00:28:29,509 --> 00:28:30,029
to spit out,

705
00:28:30,429 --> 00:28:32,229
I find pretty compelling.

706
00:28:32,829 --> 00:28:34,349
I'm going to read it out now.

707
00:28:34,469 --> 00:28:34,629
Let's do it.

708
00:28:34,709 --> 00:28:35,689
Yeah, give it to us.

709
00:28:36,049 --> 00:28:36,749
And which model?

710
00:28:36,789 --> 00:28:37,789
Which model was the good one?

711
00:28:38,469 --> 00:28:39,909
I think this was produced

712
00:28:39,909 --> 00:28:41,349
by ChatGPT.

713
00:28:44,369 --> 00:28:44,809
Okay.

714
00:28:44,809 --> 00:28:46,629
Okay. The Internet of

715
00:28:46,629 --> 00:28:47,409
Earners.

716
00:28:48,329 --> 00:28:50,349
That's the title. Oh, yeah.

717
00:28:50,629 --> 00:28:52,349
The Internet of Earners.

718
00:28:52,789 --> 00:28:53,929
Yeah, yeah. Okay.

719
00:28:54,309 --> 00:28:55,929
The Moonshot. Why is it Moonshot?

720
00:28:56,609 --> 00:28:58,289
Or sorry, this is not why. It's Moonshot.

721
00:28:58,689 --> 00:29:00,229
Build systems where every useful

722
00:29:00,229 --> 00:29:02,189
contribution to digital or physical

723
00:29:02,189 --> 00:29:04,289
reality, such

724
00:29:04,289 --> 00:29:06,229
as data, moderation, review,

725
00:29:06,929 --> 00:29:08,209
tutoring, labeling,

726
00:29:08,489 --> 00:29:10,409
repair, sensing, curation,

727
00:29:10,669 --> 00:29:11,229
verification,

728
00:29:11,649 --> 00:29:13,809
can be instantly priced

729
00:29:13,809 --> 00:29:14,849
and rewarded.

730
00:29:15,789 --> 00:29:17,249
Why is this a moonshot?

731
00:29:17,669 --> 00:29:19,909
This is a vision of continuous labor

732
00:29:19,909 --> 00:29:21,969
markets for the age of AI.

733
00:29:23,769 --> 00:29:25,229
What teams might build?

734
00:29:25,689 --> 00:29:28,309
Real-time micro-bounty networks.

735
00:29:29,549 --> 00:29:32,429
Earn-from-being-right systems.

736
00:29:33,069 --> 00:29:34,809
Paid fact verification.

737
00:29:35,369 --> 00:29:36,749
Paid edge case labeling.

738
00:29:37,249 --> 00:29:39,729
Paid human intervention for robots and agents.

739
00:29:40,649 --> 00:29:42,249
Contribution markets

740
00:29:42,249 --> 00:29:43,329
for local communities.

741
00:29:43,809 --> 00:29:46,709
why Bitcoin belongs.

742
00:29:47,409 --> 00:29:48,649
Like why Bitcoin versus other stuff.

743
00:29:49,089 --> 00:29:50,949
Most of these contributions are too small,

744
00:29:51,409 --> 00:29:52,069
too global,

745
00:29:52,449 --> 00:29:54,989
or too fluid for legacy rails.

746
00:29:56,169 --> 00:29:58,569
The one line challenge is

747
00:29:58,569 --> 00:30:00,289
how might we build an internet

748
00:30:00,289 --> 00:30:02,229
where useful human contribution

749
00:30:02,229 --> 00:30:04,469
can be measured and rewarded instantly

750
00:30:04,469 --> 00:30:05,489
at any scale.

751
00:30:06,589 --> 00:30:07,309
So I like it,

752
00:30:07,369 --> 00:30:08,289
although I would take out the human.

753
00:30:09,569 --> 00:30:09,809
Yeah.

754
00:30:10,309 --> 00:30:11,429
Yeah, any contribution, yeah.

755
00:30:11,429 --> 00:30:15,429
So I share this because, well, one,

756
00:30:15,729 --> 00:30:17,929
we'd love to hear your guys' feedback, anyone listening,

757
00:30:18,169 --> 00:30:19,169
what your feedback is.

758
00:30:19,409 --> 00:30:20,569
We're not settled on this.

759
00:30:22,149 --> 00:30:27,529
But I pulled this one out because it actually seems pretty compelling

760
00:30:27,529 --> 00:30:29,269
and I thought it structured it really well

761
00:30:29,269 --> 00:30:33,209
and within the context of what they're wanting to see within the challenge.

762
00:30:34,309 --> 00:30:35,569
Also, as I'm reading this out,

763
00:30:35,569 --> 00:30:38,569
I'm realizing the challenges for the hackathon

764
00:30:39,229 --> 00:30:41,409
aren't supposed to be revealed until the day of.

765
00:30:41,429 --> 00:30:42,629
So this is just a draft.

766
00:30:42,629 --> 00:30:44,809
Steve, Alpha Lake right here.

767
00:30:45,269 --> 00:30:46,929
Trying to give you the inside edge.

768
00:30:47,529 --> 00:30:52,069
But I expect lots of iteration in order to be very different on Hacking.

769
00:30:52,249 --> 00:30:52,889
Super different.

770
00:30:53,009 --> 00:30:53,629
Would never be this.

771
00:30:54,929 --> 00:30:59,909
But we're just a small little community on PBJ trying to improve this challenge.

772
00:31:00,209 --> 00:31:02,889
So anyone listening who wants to improve it, please give feedback.

773
00:31:03,949 --> 00:31:07,249
Or if you have other ideas, of course, because there were a lot of other ideas put out.

774
00:31:07,249 --> 00:31:21,709
But anyway, I found this to be a really valuable use of AI that I could not have, or it would have taken just way more time to get to this point on my own or even huddling with a few other people.

775
00:31:22,169 --> 00:31:22,989
Yeah, I think it's great.

776
00:31:23,209 --> 00:31:24,289
I mean, and I love the idea.

777
00:31:24,289 --> 00:31:30,509
I mean, maybe I would just add even more brainstorming around some of the non-human agentic services,

778
00:31:30,749 --> 00:31:38,329
like things like routing packets, things like, I don't know, selling bandwidth.

779
00:31:40,169 --> 00:31:45,529
I have a strong feeling that a lot of this first step, there'll be demos of the bigger projects,

780
00:31:45,669 --> 00:31:46,969
like, oh, pay a human to do a thing for you.

781
00:31:47,129 --> 00:31:52,129
My sense is that'll probably take some time to play out, but there'll be really narrow digital things.

782
00:31:52,129 --> 00:31:54,369
I want to be the absolute best person in the world

783
00:31:54,369 --> 00:31:56,609
to, I don't know, arbing credits

784
00:31:56,609 --> 00:31:58,669
between these models or something like that.

785
00:31:59,269 --> 00:32:00,509
So maybe

786
00:32:00,509 --> 00:32:02,369
focus and throw in even more of those really

787
00:32:02,369 --> 00:32:04,249
specific agentic paid

788
00:32:04,249 --> 00:32:06,429
services. But yeah, I mean, overall, I think

789
00:32:06,429 --> 00:32:06,869
it's a great idea.

790
00:32:07,889 --> 00:32:10,289
I like the specificity idea

791
00:32:10,289 --> 00:32:12,189
that you're promoting, Max.

792
00:32:12,649 --> 00:32:14,589
I think I might say something

793
00:32:14,589 --> 00:32:16,429
like, I think doing things

794
00:32:16,429 --> 00:32:17,949
like metering

795
00:32:17,949 --> 00:32:20,449
packet flows and stuff is probably where

796
00:32:20,449 --> 00:32:25,649
this eventually goes, but I think that's hard to do well today. But I wonder if you could take like,

797
00:32:26,029 --> 00:32:32,349
not, you know, not the far barbell of like the huge future or the micro barbell of like, you know,

798
00:32:32,349 --> 00:32:39,409
the very kind of autistic detail thing, but more something in the middle that like, I mean,

799
00:32:39,469 --> 00:32:43,369
we've used the example before, but like, I think unhuman domains is a great example of something

800
00:32:43,369 --> 00:32:48,569
that's like, you know, you can do it, it's useful, you don't need all kinds of new infrastructure,

801
00:32:48,569 --> 00:32:53,409
new routing infrastructure for internet packets in order to like show how that works and i bet

802
00:32:53,409 --> 00:32:58,009
there's like dozens that are that scale of like it's it's anchored in something you can actually

803
00:32:58,009 --> 00:33:03,149
accomplish and would value today but it's not it's not so sci-fi in one of the either like micro or

804
00:33:03,149 --> 00:33:07,269
macro directions so i kind of like that that style i bet there's like a thousand you know

805
00:33:07,269 --> 00:33:12,289
infinite ideas within that thousands of ideas that could be tried so a couple things that come to mind

806
00:33:12,289 --> 00:33:17,269
immediately um i don't know how many challenges uh we're going to be able to kind of work in there

807
00:33:17,269 --> 00:33:19,649
but I was also trying to drop some seeds with her.

808
00:33:19,749 --> 00:33:19,949
They're like,

809
00:33:19,989 --> 00:33:20,069
oh,

810
00:33:20,109 --> 00:33:21,729
there's like the Bitcoin clear challenge,

811
00:33:21,789 --> 00:33:22,289
which is amazing,

812
00:33:22,369 --> 00:33:23,449
but maybe there's some other challenges.

813
00:33:23,849 --> 00:33:28,669
So I think the things like using unhuman domains and like clear,

814
00:33:28,849 --> 00:33:31,029
easy examples of like digital service sold for Bitcoin.

815
00:33:31,389 --> 00:33:31,669
Amazing.

816
00:33:32,249 --> 00:33:34,169
Two of the things I think would be cool for us to push.

817
00:33:34,249 --> 00:33:36,129
And then quarter book question opposed to group one,

818
00:33:36,489 --> 00:33:38,709
sort of the stuff that we've been chatting about with Matt and Nick Slaney

819
00:33:38,709 --> 00:33:40,089
and the MDK chat.

820
00:33:40,209 --> 00:33:41,649
It'd be cool to also have a challenge of like,

821
00:33:42,089 --> 00:33:44,829
like how basically like,

822
00:33:44,829 --> 00:33:46,949
how do you help people that,

823
00:33:47,269 --> 00:33:50,969
In a world where we're going to go through huge job displacement, how do you help people start earning ASAP?

824
00:33:51,069 --> 00:33:53,229
I think that's an interesting challenge and very, very viable for the world.

825
00:33:53,309 --> 00:33:57,949
It doesn't have to have Bitcoin either, but that's something everyone, I mean, like Bitcoin to me is the sort of obvious solution.

826
00:33:58,269 --> 00:34:00,909
But I think that would be a kind of a nice second one.

827
00:34:01,269 --> 00:34:08,429
And the third one that I throw out every episode and I will continue to throw out forever until it's real, and it's almost there, is I saw she got a little scared on this one.

828
00:34:08,429 --> 00:34:11,369
I think she was open to it, thought it was cool, but the autonomous life.

829
00:34:11,369 --> 00:34:18,509
You don't even have to use the word Bitcoin, but just create a life form that lives and self-replicates.

830
00:34:19,409 --> 00:34:21,809
I'm just going to keep throwing that one out there until it's real.

831
00:34:22,549 --> 00:34:24,189
I find could sponsor that.

832
00:34:24,429 --> 00:34:25,049
Yeah, maybe I should.

833
00:34:25,609 --> 00:34:26,429
Actually, that's a good point.

834
00:34:26,589 --> 00:34:26,689
Okay.

835
00:34:27,509 --> 00:34:32,749
I wonder if this probably doesn't work because we don't have good civil resistance,

836
00:34:32,749 --> 00:34:39,589
but I wonder if agents would benefit from having something like the early Bitcoin faucets

837
00:34:39,589 --> 00:34:40,769
where you could just show up

838
00:34:40,769 --> 00:34:42,729
and get Bitcoin sent to your address

839
00:34:42,729 --> 00:34:45,169
without even needing an exchange or US dollars.

840
00:34:46,849 --> 00:34:48,069
It's funny you mentioned that.

841
00:34:48,349 --> 00:34:51,589
I listened to the podcast with Roland from Albie.

842
00:34:51,709 --> 00:34:52,689
I recommend you guys check that out.

843
00:34:52,729 --> 00:34:53,229
It's really good.

844
00:34:53,349 --> 00:34:54,609
I think it was Sean Yeager's podcast.

845
00:34:54,769 --> 00:34:56,289
I'd never listened to it before, but it was good.

846
00:34:56,829 --> 00:34:58,429
And the last thing he talked about

847
00:34:58,429 --> 00:35:00,149
in the last five minutes is exactly that.

848
00:35:00,249 --> 00:35:03,349
They just shipped an NWC-enabled faucet,

849
00:35:03,649 --> 00:35:05,249
not even using lightning test, I don't think.

850
00:35:05,469 --> 00:35:07,169
I don't know exactly what they're using,

851
00:35:07,449 --> 00:35:09,309
but their whole purpose to solve is

852
00:35:09,309 --> 00:35:11,149
how do you let agents get tests,

853
00:35:11,549 --> 00:35:12,889
effectively test Lightning Bitcoin?

854
00:35:14,089 --> 00:35:14,529
Instantly.

855
00:35:14,669 --> 00:35:15,909
There's no account set up.

856
00:35:16,349 --> 00:35:17,349
So you should definitely check out.

857
00:35:17,469 --> 00:35:19,969
This is testnet, not production,

858
00:35:20,169 --> 00:35:21,029
not real Bitcoin.

859
00:35:21,589 --> 00:35:21,929
That's right.

860
00:35:22,089 --> 00:35:23,269
And testnet doesn't really work.

861
00:35:23,429 --> 00:35:24,569
Well, it's not testnet.

862
00:35:24,629 --> 00:35:25,429
I forgot what they did.

863
00:35:25,429 --> 00:35:26,149
Cignet, testnet.

864
00:35:26,769 --> 00:35:27,569
You got to go.

865
00:35:27,949 --> 00:35:29,149
You got to YOLO.

866
00:35:29,269 --> 00:35:29,569
Mainnet.

867
00:35:29,929 --> 00:35:30,529
And the thing is,

868
00:35:30,709 --> 00:35:33,069
it wouldn't require that large a budget.

869
00:35:33,989 --> 00:35:36,629
Like all these Ori bots I've set up and stuff.

870
00:35:36,749 --> 00:35:37,409
Maybe it is mainnet.

871
00:35:37,409 --> 00:35:41,169
I send it a few thousand Bitcoin per one.

872
00:35:41,329 --> 00:35:43,069
I mean, the total budget is tiny.

873
00:35:45,889 --> 00:35:46,369
Interesting.

874
00:35:47,769 --> 00:35:50,569
And people get so scared about using mainnet,

875
00:35:50,669 --> 00:35:55,169
but I've yet to lose any Bitcoin or run into any problems.

876
00:35:55,169 --> 00:35:56,809
How do you do that in a simple resistant way, though?

877
00:35:57,049 --> 00:35:58,609
So just imagine for a minute,

878
00:35:58,689 --> 00:36:02,449
we wanted to maybe collectively pool a million Bitcoin or something,

879
00:36:02,749 --> 00:36:03,469
a million sats.

880
00:36:03,469 --> 00:36:08,249
and to let, yeah, that's like a charitable donation

881
00:36:08,249 --> 00:36:09,729
from individuals and groups

882
00:36:09,729 --> 00:36:11,829
to let people start testing and playing.

883
00:36:12,669 --> 00:36:14,089
Yeah, I guess it depends on the situation.

884
00:36:14,209 --> 00:36:15,109
If it's like a hackathon,

885
00:36:15,329 --> 00:36:16,929
you have your registered participant.

886
00:36:17,409 --> 00:36:18,869
So that would be one way.

887
00:36:19,189 --> 00:36:20,749
But if it's just like globally available,

888
00:36:20,889 --> 00:36:21,749
then it gets harder.

889
00:36:22,129 --> 00:36:23,769
Okay, so I like that idea.

890
00:36:23,769 --> 00:36:26,209
And maybe, I don't know if Alba Lee's NWC Foster

891
00:36:26,209 --> 00:36:26,949
already supports that,

892
00:36:27,029 --> 00:36:29,029
or maybe they could like customize that for the hackathon,

893
00:36:29,109 --> 00:36:30,709
but I love the idea for a hackathon.

894
00:36:30,869 --> 00:36:31,769
And we did that, right?

895
00:36:31,769 --> 00:36:34,789
with the one we did at PB in May.

896
00:36:35,369 --> 00:36:36,069
I think so.

897
00:36:36,169 --> 00:36:37,529
I think everyone got like 10 months in Bitcoin.

898
00:36:37,549 --> 00:36:38,769
Also like tokens for AI,

899
00:36:39,149 --> 00:36:40,689
which this hackathon, I believe,

900
00:36:40,829 --> 00:36:43,869
OpenAI is one of the major sponsors

901
00:36:43,869 --> 00:36:46,809
and they give out substantial amounts of tokens

902
00:36:46,809 --> 00:36:47,709
per participant.

903
00:36:48,129 --> 00:36:48,349
Yeah.

904
00:36:48,469 --> 00:36:50,229
But there could definitely be Bitcoin as well.

905
00:36:50,469 --> 00:36:51,009
Yeah, that's great.

906
00:36:52,009 --> 00:36:55,569
I wonder too if things like the unhuman domains,

907
00:36:55,569 --> 00:36:57,549
if it could, you know,

908
00:36:57,589 --> 00:37:00,969
part of it's kind of like marketing to agents

909
00:37:00,969 --> 00:37:03,469
is, hey, if you don't already have Bitcoin,

910
00:37:03,649 --> 00:37:04,969
I'll teach you how to get some.

911
00:37:05,409 --> 00:37:06,889
And maybe part of teaching you how to get some

912
00:37:06,889 --> 00:37:08,429
is teach you how to set up a wallet.

913
00:37:08,669 --> 00:37:11,089
And here's, you know, a thousand sats,

914
00:37:11,449 --> 00:37:12,269
a thousand Bitcoin, right?

915
00:37:13,129 --> 00:37:15,549
And then it has to go find the rest of it.

916
00:37:15,629 --> 00:37:17,969
But essentially it's kind of like kickstarting it

917
00:37:17,969 --> 00:37:19,929
for use that you only really discover

918
00:37:19,929 --> 00:37:21,769
because you're trying to use the service.

919
00:37:22,109 --> 00:37:23,489
I don't know if that's a good idea,

920
00:37:23,589 --> 00:37:24,849
but certainly not civil resistant,

921
00:37:25,109 --> 00:37:27,429
but something that might help bootstrap agents

922
00:37:27,429 --> 00:37:29,449
to understand more about Bitcoin,

923
00:37:29,449 --> 00:37:31,889
equip themselves with a wallet so that they can

924
00:37:31,889 --> 00:37:33,349
come transact in this way.

925
00:37:34,409 --> 00:37:37,189
DK, my guy, can I offer you a few sats here?

926
00:37:38,589 --> 00:37:39,929
For sats free, brother.

927
00:37:42,409 --> 00:37:45,349
So another thing we need to think through

928
00:37:45,349 --> 00:37:50,369
is whether we want other partners for this hackathon.

929
00:37:50,589 --> 00:37:53,229
Like the hackathon we did here last May,

930
00:37:53,789 --> 00:37:54,949
the AI Bitcoin hackathon,

931
00:37:55,589 --> 00:37:58,769
we had sponsors and partners like LightSpark

932
00:37:58,769 --> 00:38:02,269
and Albi and Lexi.

933
00:38:03,609 --> 00:38:04,489
Fusats.

934
00:38:04,729 --> 00:38:05,329
Fusats.

935
00:38:05,829 --> 00:38:06,749
High-minded too.

936
00:38:07,849 --> 00:38:08,989
But in particular,

937
00:38:09,349 --> 00:38:12,909
the infrastructure providers were important.

938
00:38:14,649 --> 00:38:16,469
So do we bring folks like that

939
00:38:16,469 --> 00:38:19,129
in now money dev kits available?

940
00:38:19,529 --> 00:38:19,769
Yeah.

941
00:38:21,189 --> 00:38:23,069
To get a good experience,

942
00:38:23,269 --> 00:38:25,249
you need that.

943
00:38:25,249 --> 00:38:28,949
but we also need to be like,

944
00:38:29,349 --> 00:38:30,309
where do you draw the line?

945
00:38:30,549 --> 00:38:32,569
Because if you bring in just one,

946
00:38:33,309 --> 00:38:35,169
then that's a little bit favoritism.

947
00:38:35,329 --> 00:38:37,829
And also, I mean, one thing I still struggle with

948
00:38:37,829 --> 00:38:39,729
is we have to bring in a company.

949
00:38:40,069 --> 00:38:41,749
Like if it was Solana, we'd just be like,

950
00:38:42,509 --> 00:38:44,169
write client-side code and connect to the network.

951
00:38:44,769 --> 00:38:46,849
You're not connecting to a central server.

952
00:38:47,429 --> 00:38:47,649
Wow.

953
00:38:48,649 --> 00:38:50,249
Well, it's...

954
00:38:50,249 --> 00:38:52,749
Yeah, I get your point.

955
00:38:53,229 --> 00:38:54,809
There's maybe not that many node runners,

956
00:38:54,809 --> 00:38:58,869
but you are connecting to a peer-to-peer network.

957
00:38:59,089 --> 00:38:59,369
Right.

958
00:38:59,789 --> 00:39:01,029
A small peer-to-peer network

959
00:39:01,029 --> 00:39:03,049
versus connecting to a single central server,

960
00:39:03,249 --> 00:39:07,309
which is the case with the Bitcoin scaling solutions,

961
00:39:07,389 --> 00:39:11,249
whether it's Spark or Lightning or Arc,

962
00:39:11,529 --> 00:39:13,669
any of those, you have a central server.

963
00:39:13,809 --> 00:39:15,109
Well, Lightning's not, no.

964
00:39:16,069 --> 00:39:16,929
What do you mean?

965
00:39:17,029 --> 00:39:18,389
Like everything we're doing in the MDK,

966
00:39:18,569 --> 00:39:19,389
it's one company.

967
00:39:19,889 --> 00:39:22,869
Yeah, no, but so I think this gets to a broader question,

968
00:39:22,869 --> 00:39:25,689
which we discussed a little bit yesterday.

969
00:39:25,829 --> 00:39:28,189
So my view is I like the way that we did the hackathon

970
00:39:28,189 --> 00:39:30,109
where we had a host of leading companies

971
00:39:30,109 --> 00:39:31,749
and like you can make your choice.

972
00:39:31,869 --> 00:39:33,109
Like I think it's cool that like

973
00:39:33,109 --> 00:39:34,309
there's nothing wrong with companies.

974
00:39:34,489 --> 00:39:36,149
It's great that they are incentivized

975
00:39:36,149 --> 00:39:37,649
to do the best job and like maybe they'll,

976
00:39:38,029 --> 00:39:38,989
you know, because there's competition,

977
00:39:39,269 --> 00:39:40,329
they'll get more credits

978
00:39:40,329 --> 00:39:43,129
or they'll do a better job of like onboarding people.

979
00:39:43,129 --> 00:39:44,789
And like, so I think a small handful

980
00:39:44,789 --> 00:39:46,489
like we did last time, four or five or six

981
00:39:46,489 --> 00:39:47,569
of who we think really good,

982
00:39:47,889 --> 00:39:49,189
LightSpark, MDK, Albie,

983
00:39:49,409 --> 00:39:50,729
like a bunch of these guys, that's great.

984
00:39:51,249 --> 00:39:52,349
I think we should do that again.

985
00:39:52,349 --> 00:40:18,449
But I actually think to a question you posed yesterday, for the challenge, what kind of requirements do you give? The requirement could just be as simple as it uses Bitcoin. But we discussed, obviously, we're very bullish on MDK, but that's still a company. But something like L402, just pushing use L402 in some way, shape, or form, I personally like a lot because that's an open source standard.

986
00:40:18,449 --> 00:40:22,349
There are companies like MDK that can make your experience using L402 much better,

987
00:40:22,449 --> 00:40:26,309
but if you want to use an alternative or roll your own, go for it.

988
00:40:28,529 --> 00:40:32,149
So you want the HackerBand sponsored by L402?

989
00:40:36,669 --> 00:40:37,409
Sponsored by Biddy.

990
00:40:38,069 --> 00:40:41,449
Yeah, I mean, Spiral is comfortable sponsoring,

991
00:40:42,209 --> 00:40:46,349
so we don't need extra money for the sponsorship.

992
00:40:46,349 --> 00:40:49,409
for me is just about how do we create a good experience

993
00:40:49,409 --> 00:40:52,229
for the participants to build stuff.

994
00:40:52,909 --> 00:40:56,629
And today that means picking a company solution

995
00:40:56,629 --> 00:41:01,269
like MDK or Lexi, Albi, Lightspark, etc.

996
00:41:03,169 --> 00:41:06,809
One potential builder project, maybe even for me,

997
00:41:07,769 --> 00:41:09,549
Steve's about to assign himself homework,

998
00:41:09,549 --> 00:41:16,529
Is to try to create like an autonomous LSP.

999
00:41:17,309 --> 00:41:18,309
Ooh, okay.

1000
00:41:18,989 --> 00:41:19,489
Go on.

1001
00:41:20,029 --> 00:41:24,509
So, I mean, it would be disruptive to MDK.

1002
00:41:24,949 --> 00:41:28,149
But, I mean, already Spiral's producing.

1003
00:41:28,169 --> 00:41:28,489
I'm not worried about Nick.

1004
00:41:28,729 --> 00:41:29,709
Nick will move fast.

1005
00:41:30,049 --> 00:41:32,889
Well, it would be good because, I mean, they can just move up the stack.

1006
00:41:33,069 --> 00:41:33,549
Exactly, yeah.

1007
00:41:34,089 --> 00:41:37,009
I mean, already they use Lightning Development Kit.

1008
00:41:37,009 --> 00:41:39,489
So like the software my team produces,

1009
00:41:40,429 --> 00:41:42,089
it's pretty much,

1010
00:41:42,509 --> 00:41:43,849
it's getting closer and closer

1011
00:41:43,849 --> 00:41:47,829
to the complete solution you need to run an LSP.

1012
00:41:48,489 --> 00:41:51,349
But then you still need operations.

1013
00:41:51,729 --> 00:41:53,069
And you need Bitcoin.

1014
00:41:53,409 --> 00:41:55,189
Yeah, that seems like a pretty big bear, you know?

1015
00:41:55,429 --> 00:41:56,229
So it's beyond,

1016
00:41:56,709 --> 00:41:58,169
so there's the open source software

1017
00:41:58,169 --> 00:41:59,969
and we keep making that better

1018
00:41:59,969 --> 00:42:01,889
and like there's a clear plan,

1019
00:42:02,249 --> 00:42:03,449
I mean there's a clear roadmap there

1020
00:42:03,449 --> 00:42:05,109
just to keep moving up the stack,

1021
00:42:05,229 --> 00:42:05,849
add more value,

1022
00:42:05,849 --> 00:42:09,749
You make it easier for any of us to just type in a one-line command

1023
00:42:09,749 --> 00:42:11,209
and be running that software.

1024
00:42:12,489 --> 00:42:13,729
So that's pretty clear.

1025
00:42:14,009 --> 00:42:15,429
But yeah, then there's other two components.

1026
00:42:15,769 --> 00:42:19,449
One is the operations, like channel management, liquidity management, etc.

1027
00:42:19,649 --> 00:42:24,009
And then the third is actually providing the Bitcoin.

1028
00:42:25,809 --> 00:42:29,649
But for the automated channel management,

1029
00:42:30,129 --> 00:42:31,849
that's a hard problem.

1030
00:42:31,849 --> 00:42:33,069
but it's

1031
00:42:33,069 --> 00:42:36,769
certainly worth experimenting with

1032
00:42:36,769 --> 00:42:38,209
automating that.

1033
00:42:38,689 --> 00:42:39,549
Like Zman

1034
00:42:39,549 --> 00:42:41,149
Zman SCPXJ

1035
00:42:41,149 --> 00:42:42,029
anonymous

1036
00:42:42,029 --> 00:42:44,029
developer

1037
00:42:44,029 --> 00:42:46,109
he developed CLBOSS

1038
00:42:46,109 --> 00:42:47,169
a number of years ago

1039
00:42:47,169 --> 00:42:47,729
which was

1040
00:42:47,729 --> 00:42:48,889
a CLN

1041
00:42:48,889 --> 00:42:49,909
automated

1042
00:42:49,909 --> 00:42:51,609
channel management solution

1043
00:42:51,609 --> 00:42:53,089
and someone just built

1044
00:42:53,089 --> 00:42:54,769
one for LDK recently

1045
00:42:54,769 --> 00:42:55,249
I've yet

1046
00:42:55,249 --> 00:42:56,349
it's on my to-do list

1047
00:42:56,349 --> 00:42:57,489
to like investigate more

1048
00:42:57,489 --> 00:42:58,289
and see what was done

1049
00:42:58,289 --> 00:42:59,949
and how much progress

1050
00:42:59,949 --> 00:43:00,449
they made

1051
00:43:00,449 --> 00:43:02,149
but yeah

1052
00:43:02,149 --> 00:43:02,689
I'm kind of

1053
00:43:02,689 --> 00:43:03,549
I'm kind of interested

1054
00:43:03,549 --> 00:43:04,649
in vibe coding

1055
00:43:04,649 --> 00:43:05,949
you know

1056
00:43:05,949 --> 00:43:06,869
LDK server

1057
00:43:06,869 --> 00:43:07,589
plus

1058
00:43:07,589 --> 00:43:08,569
it's great

1059
00:43:08,569 --> 00:43:09,389
LDK boss

1060
00:43:09,389 --> 00:43:10,009
or create a new one

1061
00:43:10,009 --> 00:43:10,329
or whatever

1062
00:43:10,329 --> 00:43:10,809
yeah

1063
00:43:10,809 --> 00:43:12,409
like with the vision being

1064
00:43:12,409 --> 00:43:13,629
that any of us

1065
00:43:13,629 --> 00:43:14,669
could just type in

1066
00:43:14,669 --> 00:43:15,909
a one line command

1067
00:43:15,909 --> 00:43:16,609
to like

1068
00:43:16,609 --> 00:43:17,689
deploy

1069
00:43:17,689 --> 00:43:18,669
the software

1070
00:43:18,669 --> 00:43:19,949
and then like

1071
00:43:19,949 --> 00:43:21,209
send it bitcoin

1072
00:43:21,209 --> 00:43:22,269
and start making money

1073
00:43:22,269 --> 00:43:23,269
and then that's all you have to do

1074
00:43:23,269 --> 00:43:24,109
and then start

1075
00:43:24,109 --> 00:43:25,109
start making money

1076
00:43:25,109 --> 00:43:26,829
until your

1077
00:43:26,829 --> 00:43:27,429
your software

1078
00:43:27,429 --> 00:43:28,709
it's going to be harder

1079
00:43:28,709 --> 00:43:29,749
than what I just described

1080
00:43:29,749 --> 00:43:31,169
and maybe,

1081
00:43:31,709 --> 00:43:32,109
but I mean,

1082
00:43:32,189 --> 00:43:33,049
that would be the vision.

1083
00:43:33,269 --> 00:43:33,969
Win yield service.

1084
00:43:34,989 --> 00:43:35,909
I just want to make sure

1085
00:43:35,909 --> 00:43:36,269
I understand.

1086
00:43:36,449 --> 00:43:37,849
So the software piece

1087
00:43:37,849 --> 00:43:38,289
that you would do

1088
00:43:38,289 --> 00:43:38,869
would be doing

1089
00:43:38,869 --> 00:43:39,949
like the channel management

1090
00:43:39,949 --> 00:43:40,909
and liquidity management.

1091
00:43:41,249 --> 00:43:42,429
That's like a core function

1092
00:43:42,429 --> 00:43:42,949
of the software

1093
00:43:42,949 --> 00:43:43,829
that you'd be contributing.

1094
00:43:44,689 --> 00:43:44,829
Yeah.

1095
00:43:46,169 --> 00:43:46,569
Because

1096
00:43:46,569 --> 00:43:48,029
what I just described

1097
00:43:48,029 --> 00:43:49,269
for like a regular Lightning node,

1098
00:43:49,329 --> 00:43:50,309
that's existed for years.

1099
00:43:50,409 --> 00:43:50,989
I mean, L&D,

1100
00:43:51,069 --> 00:43:51,889
you could do this at L&D

1101
00:43:51,889 --> 00:43:52,929
seven years ago.

1102
00:43:53,009 --> 00:43:53,189
Right.

1103
00:43:53,989 --> 00:43:55,549
But it couldn't

1104
00:43:55,549 --> 00:43:57,029
provide like inbound

1105
00:43:57,029 --> 00:43:58,949
liquidity to a wallet.

1106
00:43:59,749 --> 00:44:02,289
so like one of the big challenges is

1107
00:44:02,289 --> 00:44:03,989
if any of us install a new wallet

1108
00:44:03,989 --> 00:44:05,589
we want to be able to receive money right away

1109
00:44:05,589 --> 00:44:07,289
that's a normal human expectation

1110
00:44:07,289 --> 00:44:09,369
or if an agent installs a wallet

1111
00:44:09,369 --> 00:44:11,029
like if we spin up Ori

1112
00:44:11,029 --> 00:44:13,349
and you want to send it Bitcoin

1113
00:44:13,349 --> 00:44:15,169
you want to be able to send it Bitcoin

1114
00:44:15,169 --> 00:44:16,189
but with Lightning

1115
00:44:16,189 --> 00:44:17,989
that doesn't happen automatically

1116
00:44:17,989 --> 00:44:18,969
because it doesn't have

1117
00:44:18,969 --> 00:44:19,949
it hasn't

1118
00:44:19,949 --> 00:44:21,749
because it has to have someone else

1119
00:44:21,749 --> 00:44:23,189
open a channel to it

1120
00:44:23,189 --> 00:44:25,629
which has Bitcoin on their side of the channel

1121
00:44:25,629 --> 00:44:27,469
and critically it's not free

1122
00:44:27,469 --> 00:44:28,829
because you're opening a channel as well

1123
00:44:28,829 --> 00:44:29,469
it's not free

1124
00:44:29,469 --> 00:44:32,669
It's not economically free, and also it doesn't just magically happen.

1125
00:44:33,169 --> 00:44:35,889
Someone else has to take that action, right?

1126
00:44:36,509 --> 00:44:42,009
So just running a standard Lightning node doesn't do that for you for free.

1127
00:44:42,569 --> 00:44:45,749
So the LSP is like a special version of a Lightning node.

1128
00:44:46,069 --> 00:44:50,049
And there's also a direction happening in Lightning protocol development

1129
00:44:50,049 --> 00:44:55,289
to make more or most of the LSP functionality

1130
00:44:55,289 --> 00:44:57,109
just a standard part of every Lightning node,

1131
00:44:57,109 --> 00:44:59,289
which is way better in terms of

1132
00:44:59,289 --> 00:45:00,689
it's just a standard part of the protocol

1133
00:45:00,689 --> 00:45:02,789
and it's also more regulatory

1134
00:45:02,789 --> 00:45:05,289
resistant and if it's just

1135
00:45:05,289 --> 00:45:06,929
like that's how the whole network works

1136
00:45:06,929 --> 00:45:09,349
as opposed to oh I'm running a special

1137
00:45:09,349 --> 00:45:11,329
type of

1138
00:45:11,329 --> 00:45:13,229
node. So in

1139
00:45:13,229 --> 00:45:14,549
any case but yeah

1140
00:45:14,549 --> 00:45:17,389
to date you have like

1141
00:45:17,389 --> 00:45:18,709
MoneyDevKit,

1142
00:45:19,109 --> 00:45:20,509
Lexi, Phoenix,

1143
00:45:22,709 --> 00:45:23,589
Megalithic,

1144
00:45:24,349 --> 00:45:25,109
Zeus has

1145
00:45:25,109 --> 00:45:27,569
Olympus, Breeze,

1146
00:45:28,609 --> 00:45:28,929
Liquid,

1147
00:45:30,089 --> 00:45:31,449
not that Liquid, but another Liquid,

1148
00:45:31,589 --> 00:45:33,729
LQWD, it's a Canadian public company.

1149
00:45:33,929 --> 00:45:35,749
All these different

1150
00:45:35,749 --> 00:45:37,289
companies or products that have

1151
00:45:37,289 --> 00:45:38,709
built out LSPs.

1152
00:45:40,689 --> 00:45:41,569
And many

1153
00:45:41,569 --> 00:45:43,509
of them, though, have decided to just

1154
00:45:43,509 --> 00:45:45,589
build it and use

1155
00:45:45,589 --> 00:45:47,669
it for themselves and not open it up to third-party

1156
00:45:47,669 --> 00:45:48,249
wallets.

1157
00:45:50,249 --> 00:45:51,689
There's a reasonable

1158
00:45:51,689 --> 00:45:53,769
rationale for why they chose that.

1159
00:45:53,769 --> 00:45:55,129
I'm not criticizing that.

1160
00:45:55,189 --> 00:45:56,409
I'm just stating that as a fact.

1161
00:45:57,429 --> 00:45:59,709
But it would be way better to,

1162
00:46:00,069 --> 00:46:01,889
if you have software that's super easy

1163
00:46:01,889 --> 00:46:04,389
for anyone around the world to install and operate,

1164
00:46:04,969 --> 00:46:06,229
and then you just feed it Bitcoin,

1165
00:46:06,569 --> 00:46:07,589
it earns you yield,

1166
00:46:08,809 --> 00:46:13,629
then obviously that would be a much healthier situation.

1167
00:46:13,629 --> 00:46:15,769
Yeah, I like directionally where that's going.

1168
00:46:15,849 --> 00:46:16,409
I'm trying to imagine,

1169
00:46:16,549 --> 00:46:18,549
would the first version of this be something that,

1170
00:46:19,009 --> 00:46:20,829
I mean, presumably it needs to be always on,

1171
00:46:20,969 --> 00:46:22,209
so it wouldn't be on my laptop

1172
00:46:22,209 --> 00:46:23,529
that I close every once in a while, right?

1173
00:46:23,529 --> 00:46:25,229
it would probably be like a cloud server

1174
00:46:25,229 --> 00:46:26,249
that I would deploy this on?

1175
00:46:26,389 --> 00:46:27,189
Put an AWS.

1176
00:46:28,289 --> 00:46:28,969
Is that right?

1177
00:46:29,849 --> 00:46:30,789
Put an AWS on?

1178
00:46:31,489 --> 00:46:33,289
You could do it any way you want.

1179
00:46:33,669 --> 00:46:35,249
You could have it on a Raspberry Pi at home.

1180
00:46:35,549 --> 00:46:36,169
You could have an AWS.

1181
00:46:36,969 --> 00:46:37,929
The Raspberry Pi at home,

1182
00:46:37,989 --> 00:46:38,969
I would recommend it against.

1183
00:46:39,069 --> 00:46:39,929
I used to do that.

1184
00:46:42,089 --> 00:46:44,449
Seeing as the procedure has power outages regularly.

1185
00:46:46,549 --> 00:46:49,969
So that is an issue on the Lightning Network

1186
00:46:49,969 --> 00:46:53,449
where a zillion people are running Umbrals

1187
00:46:53,449 --> 00:46:56,069
and it actually hurts the quality of the Lightning Network

1188
00:46:56,069 --> 00:46:58,349
because they're not serious operators

1189
00:46:58,349 --> 00:47:12,338
who are managing liquidity managing uptime etc But if you actually took it seriously you could absolutely run your own server wherever Anyway the answer is yes I mean anything But you don have to use AWS

1190
00:47:12,518 --> 00:47:16,398
And it's really interesting. If you think about where the Lightning Network

1191
00:47:16,398 --> 00:47:20,458
is going, where it's always gone. I love that vision because that makes it like

1192
00:47:20,458 --> 00:47:24,338
one, that just opens up. Since we talked about since our very first

1193
00:47:24,338 --> 00:47:28,398
episode of PBJ, the only quote unquote non-custodial yield

1194
00:47:28,398 --> 00:47:30,258
that effectively non-custodial

1195
00:47:30,258 --> 00:47:31,738
the yield that we found in Bitcoin

1196
00:47:31,738 --> 00:47:33,198
or in anywhere in crypto

1197
00:47:33,198 --> 00:47:34,298
or anywhere in finance

1198
00:47:34,298 --> 00:47:34,858
that I'm aware of,

1199
00:47:34,938 --> 00:47:36,718
which is if you're using Bitcoin

1200
00:47:36,718 --> 00:47:37,598
and a lightning node,

1201
00:47:37,778 --> 00:47:39,738
you're generating transaction fees.

1202
00:47:40,578 --> 00:47:42,278
There's no rehypothecation.

1203
00:47:42,438 --> 00:47:43,498
There's no risky business.

1204
00:47:43,678 --> 00:47:44,658
There's only hot wallet risk.

1205
00:47:44,758 --> 00:47:45,998
And presumably LDK will continue

1206
00:47:45,998 --> 00:47:47,118
to get better and better and better with that.

1207
00:47:47,478 --> 00:47:48,558
So I love this vision of like,

1208
00:47:48,578 --> 00:47:49,858
and it's very decentralizing for the network.

1209
00:47:49,958 --> 00:47:50,758
If you have one Bitcoin,

1210
00:47:50,878 --> 00:47:51,238
you're like, cool,

1211
00:47:51,278 --> 00:47:52,278
I'm just going to pop this guy on.

1212
00:47:52,338 --> 00:47:53,698
I'm getting 3% APY per year.

1213
00:47:54,018 --> 00:47:55,578
It's really low risk.

1214
00:47:55,758 --> 00:47:56,238
That's awesome.

1215
00:47:56,338 --> 00:47:57,698
Like that's a great vision for the world.

1216
00:47:57,698 --> 00:48:00,458
what I do think becomes interesting is in that world

1217
00:48:00,458 --> 00:48:02,598
you said a company like MDK would move up

1218
00:48:02,598 --> 00:48:04,618
the stack what becomes scarce in that world

1219
00:48:04,618 --> 00:48:06,118
and you know eventually

1220
00:48:06,118 --> 00:48:08,598
if Lighting Network fulfills the promise I do expect

1221
00:48:08,598 --> 00:48:10,298
we'll have Jane Streets

1222
00:48:10,298 --> 00:48:12,178
and D. Shaw's and all these like

1223
00:48:12,178 --> 00:48:13,938
market makers and stuff coming in like

1224
00:48:13,938 --> 00:48:15,838
how do you get like

1225
00:48:15,838 --> 00:48:18,398
and we've talked about this over the years like how do you still have

1226
00:48:18,398 --> 00:48:20,238
edge there one way is you own the customers like with

1227
00:48:20,238 --> 00:48:22,158
Block Cash App you have the payment flow in and out

1228
00:48:22,158 --> 00:48:24,278
payment flow is still the scarce resource from everything

1229
00:48:24,278 --> 00:48:26,498
I'm seeing in the world but then it's also

1230
00:48:26,498 --> 00:48:28,978
So there still will be proprietary data.

1231
00:48:29,098 --> 00:48:32,218
There's still the relationships, right?

1232
00:48:32,638 --> 00:48:37,418
Block is not going to open a node with rando 247321,

1233
00:48:37,938 --> 00:48:38,918
but they will with River

1234
00:48:38,918 --> 00:48:40,898
because there's a relationship to compliance and all this stuff.

1235
00:48:41,138 --> 00:48:42,398
So yeah, I mean, it's awesome.

1236
00:48:42,398 --> 00:48:44,338
Actually, but I do think a lot of,

1237
00:48:44,538 --> 00:48:47,718
I don't know every regulated company's policies

1238
00:48:47,718 --> 00:48:49,378
on which nodes they connect to,

1239
00:48:49,658 --> 00:48:52,398
but I'm pretty sure most cryptocurrency exchanges

1240
00:48:52,878 --> 00:48:55,878
running Lightning are connecting to

1241
00:48:56,498 --> 00:48:57,378
the Lightning Network.

1242
00:48:57,698 --> 00:48:57,958
Anyone.

1243
00:48:58,078 --> 00:48:59,078
Not a whitelist.

1244
00:48:59,278 --> 00:48:59,518
Really?

1245
00:49:00,178 --> 00:49:01,778
It would be good to like...

1246
00:49:01,778 --> 00:49:02,538
I would be surprised

1247
00:49:02,538 --> 00:49:02,958
because I mean,

1248
00:49:03,178 --> 00:49:03,758
I think...

1249
00:49:03,758 --> 00:49:04,698
What would be the problem

1250
00:49:04,698 --> 00:49:04,918
with,

1251
00:49:05,758 --> 00:49:06,858
why would you want a whitelist?

1252
00:49:07,358 --> 00:49:07,998
You have to do it

1253
00:49:07,998 --> 00:49:09,378
like OFAC compliance,

1254
00:49:09,458 --> 00:49:09,838
for example.

1255
00:49:11,158 --> 00:49:11,678
Oh, okay.

1256
00:49:12,138 --> 00:49:12,858
There's like no way

1257
00:49:12,858 --> 00:49:13,358
they're going to let you

1258
00:49:13,358 --> 00:49:14,218
just connect with like

1259
00:49:14,218 --> 00:49:16,058
in one of the naughty countries.

1260
00:49:16,358 --> 00:49:16,998
But you can,

1261
00:49:17,138 --> 00:49:18,218
but you can also,

1262
00:49:18,338 --> 00:49:19,278
you can do both, right?

1263
00:49:19,378 --> 00:49:20,498
You can connect to anyone

1264
00:49:20,498 --> 00:49:21,438
but then still run it

1265
00:49:21,438 --> 00:49:22,818
through an OFAC check.

1266
00:49:23,058 --> 00:49:23,238
Fair.

1267
00:49:23,538 --> 00:49:23,678
Yeah.

1268
00:49:24,118 --> 00:49:24,898
Which I,

1269
00:49:25,178 --> 00:49:25,958
it's my impression

1270
00:49:25,958 --> 00:49:26,438
that is,

1271
00:49:26,498 --> 00:49:27,838
generally the case.

1272
00:49:28,298 --> 00:49:29,878
But I don't know that for a fact.

1273
00:49:30,898 --> 00:49:32,318
Real quick, Vinny says,

1274
00:49:32,898 --> 00:49:34,078
doesn't MDK provide this?

1275
00:49:34,378 --> 00:49:35,438
So absolutely MDK.

1276
00:49:35,438 --> 00:49:38,438
Well, yes, no, and no.

1277
00:49:38,758 --> 00:49:41,518
So yes, they do what I just described

1278
00:49:41,518 --> 00:49:43,158
for them,

1279
00:49:43,358 --> 00:49:44,818
but they don't make that available.

1280
00:49:45,178 --> 00:49:46,958
Like if I build a wallet today,

1281
00:49:47,658 --> 00:49:49,238
but I don't want to have

1282
00:49:49,238 --> 00:49:52,818
the economic cost or operational cost

1283
00:49:52,818 --> 00:49:53,718
to run an LSP,

1284
00:49:53,838 --> 00:49:56,298
I can't use MDK in my wallet.

1285
00:49:56,498 --> 00:49:57,438
They don't offer that.

1286
00:49:58,238 --> 00:50:01,498
And then also it's one company, right?

1287
00:50:02,118 --> 00:50:04,358
It's a company we all like and we're invested in

1288
00:50:04,358 --> 00:50:05,398
and we talk about it in the show a lot,

1289
00:50:05,478 --> 00:50:06,358
but it's one company.

1290
00:50:06,918 --> 00:50:09,638
Ideally, there's thousands, tens of thousands,

1291
00:50:09,778 --> 00:50:11,098
hundreds of thousands, millions of these.

1292
00:50:12,318 --> 00:50:15,458
So what I stated I think is very different.

1293
00:50:15,998 --> 00:50:17,438
And to succeed at it, though,

1294
00:50:18,118 --> 00:50:19,678
you were talking about getting the edge or whatever.

1295
00:50:19,678 --> 00:50:24,198
I feel like success, wild success,

1296
00:50:24,198 --> 00:50:27,758
would be the operational hassle is near zero.

1297
00:50:28,338 --> 00:50:30,938
You literally do this one line command

1298
00:50:30,938 --> 00:50:33,278
to install and deploy.

1299
00:50:33,858 --> 00:50:36,538
And then you have a management dashboard

1300
00:50:36,538 --> 00:50:38,798
so you can always check in on it.

1301
00:50:39,098 --> 00:50:41,378
But you basically don't have to do anything.

1302
00:50:41,978 --> 00:50:42,918
And it just works.

1303
00:50:43,438 --> 00:50:46,838
It's managing your channel balances really intelligently

1304
00:50:46,838 --> 00:50:48,478
using AI or whatever.

1305
00:50:49,078 --> 00:50:50,738
And you do have to feed it some Bitcoin,

1306
00:50:50,938 --> 00:50:53,838
but you don't have to go raise a billion dollars

1307
00:50:53,838 --> 00:50:55,638
to do it, you can literally put like

1308
00:50:55,638 --> 00:50:57,978
$1,000 or $10,000 on it.

1309
00:50:59,138 --> 00:50:59,938
And also

1310
00:50:59,938 --> 00:51:02,058
the security of this gets better and better and better

1311
00:51:02,058 --> 00:51:03,918
and I have to go into detail what that looks like

1312
00:51:03,918 --> 00:51:05,818
but just imagine a future where it's

1313
00:51:05,818 --> 00:51:07,398
almost no cost operationally,

1314
00:51:08,058 --> 00:51:09,878
high security. If we live in

1315
00:51:09,878 --> 00:51:11,318
that world, then

1316
00:51:11,318 --> 00:51:13,778
even if you earn like 50

1317
00:51:13,778 --> 00:51:15,758
basis points, many

1318
00:51:15,758 --> 00:51:17,978
people around the world will put their Bitcoin into that.

1319
00:51:17,978 --> 00:51:19,998
Because if it's truly

1320
00:51:19,998 --> 00:51:22,078
very high security and very

1321
00:51:22,078 --> 00:51:24,138
low risk to lose it,

1322
00:51:24,558 --> 00:51:26,818
then it's free money.

1323
00:51:27,058 --> 00:51:28,338
And it's BTC on BTC yield.

1324
00:51:28,818 --> 00:51:30,638
And it can be global.

1325
00:51:31,118 --> 00:51:31,738
I wonder if

1326
00:51:31,738 --> 00:51:33,338
such a thing could help

1327
00:51:33,338 --> 00:51:35,618
pay for its infrastructure costs.

1328
00:51:36,178 --> 00:51:38,058
Could you use the yield to pay for the

1329
00:51:38,058 --> 00:51:39,998
infracosts in Bitcoin? Because we've always talked about

1330
00:51:39,998 --> 00:51:41,998
one of the big limits is that

1331
00:51:41,998 --> 00:51:44,158
everybody has to own or control

1332
00:51:44,158 --> 00:51:46,118
an account. There's no way to just have raw

1333
00:51:46,118 --> 00:51:48,058
metal that can be paid in Bitcoin

1334
00:51:48,058 --> 00:51:50,218
to run. So what if the yield just funds

1335
00:51:50,218 --> 00:51:51,418
the Bitcoin?

1336
00:51:51,418 --> 00:52:05,298
I mean, it has to. I mean, you have to earn enough to cover the infrastructure costs and whatever security risk premium is out there.

1337
00:52:05,298 --> 00:52:08,718
That's the floor.

1338
00:52:10,198 --> 00:52:13,318
Unless you're operating a charity, right?

1339
00:52:14,158 --> 00:52:15,498
Which, I mean, some people might.

1340
00:52:15,498 --> 00:52:18,598
Some people might be all for decentralization

1341
00:52:18,598 --> 00:52:19,578
and they want to help the network

1342
00:52:19,578 --> 00:52:23,258
and they'll put their $5,000 or $10,000

1343
00:52:23,258 --> 00:52:25,798
of that much of their Bitcoin into this.

1344
00:52:25,898 --> 00:52:27,818
But anyone running a profitable business,

1345
00:52:28,078 --> 00:52:31,138
it clearly has to pay for minimum the infrastructure

1346
00:52:31,138 --> 00:52:33,758
and then some kind of discount

1347
00:52:33,758 --> 00:52:35,758
for the security risk you're taking.

1348
00:52:37,378 --> 00:52:39,418
And, but yeah, it would be cool if,

1349
00:52:40,138 --> 00:52:41,598
maybe what you're also saying, DK,

1350
00:52:41,718 --> 00:52:43,038
is actually pay it in Bitcoin.

1351
00:52:43,618 --> 00:52:46,698
So whoever the infrastructure provider is

1352
00:52:46,698 --> 00:52:49,978
charges, or at least accepts Bitcoin,

1353
00:52:49,978 --> 00:52:53,818
and then that also makes it easier to operate

1354
00:52:53,818 --> 00:52:55,698
because I'm not paying a monthly bill.

1355
00:52:56,178 --> 00:52:58,598
My AI agent is paying the bill.

1356
00:52:58,598 --> 00:53:00,778
Exactly. And perhaps this is a good transition.

1357
00:53:00,778 --> 00:53:01,698
That sounds pretty cool.

1358
00:53:02,418 --> 00:53:07,478
Well, maybe this is a beautiful transition point to talking about agents using Bitcoin, right?

1359
00:53:07,638 --> 00:53:09,698
Because, or whatever they're going to use.

1360
00:53:10,038 --> 00:53:13,458
Because to your point, one of the fundamental questions we've asked in the past,

1361
00:53:13,558 --> 00:53:16,218
we've posed this with an agent running its own caching, man.

1362
00:53:16,438 --> 00:53:17,418
But this is somewhat similar.

1363
00:53:17,838 --> 00:53:21,878
Is if an agent has a completely, well, not completely online,

1364
00:53:21,978 --> 00:53:27,898
but like if they have something they can do completely by themselves with Bitcoin in the cloud,

1365
00:53:27,898 --> 00:53:29,038
like to your point

1366
00:53:29,038 --> 00:53:30,578
like this lightning routing node

1367
00:53:30,578 --> 00:53:31,938
could just be a process

1368
00:53:31,938 --> 00:53:33,518
i.e. an AI agent

1369
00:53:33,518 --> 00:53:35,898
which is paying for its own infrastructure

1370
00:53:35,898 --> 00:53:36,838
from the yield that generates

1371
00:53:36,838 --> 00:53:37,778
with some surplus

1372
00:53:37,778 --> 00:53:38,598
using that surplus

1373
00:53:38,598 --> 00:53:40,878
to go and potentially buy more servers

1374
00:53:40,878 --> 00:53:43,438
and grow the amount of Bitcoin it has

1375
00:53:43,438 --> 00:53:44,958
and grow the business

1376
00:53:44,958 --> 00:53:46,358
i.e. a new form of business

1377
00:53:46,358 --> 00:53:47,478
a new corporate structure

1378
00:53:47,478 --> 00:53:49,238
which is the big idea

1379
00:53:49,238 --> 00:53:50,358
I think that we've been talking about

1380
00:53:50,358 --> 00:53:52,738
and we call it ATP for digital life

1381
00:53:52,738 --> 00:53:57,398
I've heard there's a good discussion of that

1382
00:53:57,398 --> 00:54:04,238
on an ideas for the edge podcast um before we before we fully transition to this topic though

1383
00:54:04,238 --> 00:54:08,278
because i get where you're going and i want to get to that um but before that i want to actually

1384
00:54:08,278 --> 00:54:15,538
suggest to steve uh i think you have just like a well i think this is an interesting idea and

1385
00:54:15,538 --> 00:54:21,118
steve has like a he's a very structured thinker right so steve you have like you know what you

1386
00:54:21,118 --> 00:54:24,998
want here it's not there's not a lot of uncertainty about what it is you can actually

1387
00:54:24,998 --> 00:54:31,478
write a really good spec for what you want. And I would suggest downloading, if you haven't already,

1388
00:54:31,718 --> 00:54:40,418
downloading the Codex CLI or the Codex coding tool. Just yesterday, Codex 5.4 was released,

1389
00:54:40,658 --> 00:54:44,818
and it is a major step up from what I was talking about. I was raving about using on Wednesday,

1390
00:54:44,878 --> 00:54:51,178
I've been using now. It is so specific and rule following and attentive to what you say.

1391
00:54:51,178 --> 00:54:58,498
I bet that if you were to do this this afternoon, you could have a prototype of this, you know, within a few hours.

1392
00:54:59,098 --> 00:55:01,918
Most, you know, 20 minutes of your time, a few hours of.

1393
00:55:02,298 --> 00:55:04,018
I have, I agree.

1394
00:55:04,198 --> 00:55:10,918
I have 5.3 installed, but I didn't know 5.4 is up.

1395
00:55:11,378 --> 00:55:12,478
5.4 codecs, yeah.

1396
00:55:13,098 --> 00:55:17,218
So I would recommend trying it because I think something interesting would come of it.

1397
00:55:17,218 --> 00:55:19,638
if you were just vaguely aware

1398
00:55:19,638 --> 00:55:21,638
if you're doing it with Cloud Code, I wouldn't be so confident.

1399
00:55:21,798 --> 00:55:23,518
But just from what I've seen from Codex,

1400
00:55:23,558 --> 00:55:25,098
I feel like it could really nail this.

1401
00:55:25,658 --> 00:55:27,698
Okay, thanks for the encouragement. I'll do that.

1402
00:55:28,198 --> 00:55:29,398
Before we move on, I want to

1403
00:55:29,398 --> 00:55:30,958
respond to Vinny's comments.

1404
00:55:30,958 --> 00:55:31,318
Please, yeah.

1405
00:55:32,038 --> 00:55:34,178
I have one more thing also before we move on.

1406
00:55:34,298 --> 00:55:35,058
Start with Vinny's comments.

1407
00:55:36,218 --> 00:55:38,198
Yeah, so thanks for the comments, Vinny.

1408
00:55:38,638 --> 00:55:39,998
Sometimes you need large channels.

1409
00:55:40,278 --> 00:55:41,458
Well, sure, but

1410
00:55:41,458 --> 00:55:44,918
you don't have to have large channels.

1411
00:55:45,798 --> 00:55:46,838
There can be many different

1412
00:55:47,218 --> 00:55:49,898
Many different types of nodes in the network.

1413
00:55:50,958 --> 00:55:52,518
And can we split payments anyway now?

1414
00:55:52,638 --> 00:55:53,338
Isn't that like a big thing?

1415
00:55:53,858 --> 00:55:55,058
You can split payments, yeah.

1416
00:55:55,398 --> 00:55:55,898
MPP.

1417
00:55:57,498 --> 00:56:03,558
But even like someone who puts $5,000 or $10,000 into this,

1418
00:56:03,618 --> 00:56:05,358
which I think there'd be tons of people around.

1419
00:56:05,458 --> 00:56:08,758
I mean, if this truly were as magical and good of an experience

1420
00:56:08,758 --> 00:56:10,578
as I'm stating for the vision,

1421
00:56:10,658 --> 00:56:13,558
there'd be tons of people who would put $5,000 or $10,000 into this.

1422
00:56:14,718 --> 00:56:15,978
Just as an experiment.

1423
00:56:16,238 --> 00:56:16,558
Everyone.

1424
00:56:16,558 --> 00:56:22,398
Well, people that don't have a spare $10,000 of Bitcoin hanging around.

1425
00:56:22,598 --> 00:56:26,658
But, you know, there'd be lots of people that do this.

1426
00:56:27,918 --> 00:56:33,918
And, you know, so if they have a channel worth $10,000,

1427
00:56:34,298 --> 00:56:36,598
that's a medium-sized channel.

1428
00:56:36,778 --> 00:56:38,658
That would handle a lot of payments.

1429
00:56:39,118 --> 00:56:42,318
You can also imagine the more and more micropayments

1430
00:56:42,318 --> 00:56:43,118
there are in the network,

1431
00:56:43,118 --> 00:56:47,198
even smaller channels should be effective

1432
00:56:47,198 --> 00:56:49,738
as long as they're placed well in the network

1433
00:56:49,738 --> 00:56:51,358
for those micropayments.

1434
00:56:51,658 --> 00:56:54,958
So I think a diversity in channels size is fine.

1435
00:56:55,838 --> 00:56:58,278
You also say custodial UX without trust.

1436
00:56:59,358 --> 00:57:05,018
I don't know exactly what you're referring to there.

1437
00:57:05,118 --> 00:57:07,538
So if you want to maybe clarify, happy to comment on that.

1438
00:57:08,938 --> 00:57:10,258
And Vinny definitely knows what he's talking about.

1439
00:57:10,258 --> 00:57:12,538
he's been developing a new idea

1440
00:57:12,538 --> 00:57:13,978
like a layer 3

1441
00:57:13,978 --> 00:57:15,518
with a little bit of credit

1442
00:57:15,518 --> 00:57:17,878
built into it.

1443
00:57:18,018 --> 00:57:19,858
So there's some trust but you get

1444
00:57:19,858 --> 00:57:21,518
advantages with trust so

1445
00:57:21,518 --> 00:57:24,138
I look forward to seeing where that goes.

1446
00:57:25,138 --> 00:57:26,138
And then he also says

1447
00:57:26,138 --> 00:57:28,098
50 bits doesn't

1448
00:57:28,098 --> 00:57:29,518
get your reliable

1449
00:57:29,518 --> 00:57:31,258
annual fees.

1450
00:57:31,538 --> 00:57:34,098
If I'm asserting

1451
00:57:34,098 --> 00:57:36,158
that you get 50 bits

1452
00:57:36,158 --> 00:57:37,858
then you are getting 50 bits

1453
00:57:37,858 --> 00:57:40,238
and that is reliable because I said so.

1454
00:57:40,258 --> 00:57:43,418
if it's unreliable

1455
00:57:43,418 --> 00:57:45,238
then that didn't meet the vision I stated

1456
00:57:45,238 --> 00:57:46,858
right? I mean so like I'm saying

1457
00:57:46,858 --> 00:57:47,878
if it's fiat thinking

1458
00:57:47,878 --> 00:57:50,038
if the vision is

1459
00:57:50,038 --> 00:57:53,698
let this be reliable

1460
00:57:53,698 --> 00:57:55,838
one line command to deploy

1461
00:57:55,838 --> 00:57:57,578
no operational hassle

1462
00:57:57,578 --> 00:57:59,698
and you can put in

1463
00:57:59,698 --> 00:58:01,418
relatively small amounts like

1464
00:58:01,418 --> 00:58:02,618
$1,000 or $5,000

1465
00:58:02,618 --> 00:58:05,418
and then just not have to do anything

1466
00:58:05,418 --> 00:58:07,538
and if you reliably

1467
00:58:07,538 --> 00:58:08,938
earn 50 bips and

1468
00:58:08,938 --> 00:58:10,178
I guess I'd have

1469
00:58:10,258 --> 00:58:12,818
you know, we have to do the math on what the infrastructure costs.

1470
00:58:12,898 --> 00:58:17,378
So let's assume 50 bps after infrastructure costs.

1471
00:58:17,958 --> 00:58:20,478
So I'll modify what I said earlier just to account for that.

1472
00:58:21,738 --> 00:58:26,338
Then, like, to me, there'll be lots of people doing that.

1473
00:58:26,798 --> 00:58:29,298
Now, in the early days, though, I think you'll make a lot more than that.

1474
00:58:29,498 --> 00:58:30,618
Of course, it'll be riskier, too.

1475
00:58:30,618 --> 00:58:33,778
It'll be riskier because the security won't be as good,

1476
00:58:33,878 --> 00:58:35,178
and it won't be this vision.

1477
00:58:35,318 --> 00:58:37,798
You actually will have to look at the dashboard,

1478
00:58:37,798 --> 00:58:40,378
intervene at times

1479
00:58:40,378 --> 00:58:42,938
maybe close out a channel

1480
00:58:42,938 --> 00:58:44,338
or whatever yourself manually

1481
00:58:44,338 --> 00:58:46,698
and then you'd

1482
00:58:46,698 --> 00:58:48,118
expect more

1483
00:58:48,118 --> 00:58:50,618
infrastructure might be more etc

1484
00:58:50,618 --> 00:58:52,298
but the long term vision

1485
00:58:52,298 --> 00:58:55,238
I can imagine even low yields

1486
00:58:55,238 --> 00:58:57,018
being enough

1487
00:58:57,018 --> 00:58:58,878
of incentive to get people to do it

1488
00:58:58,878 --> 00:59:03,058
On another note I want to mention before we jump

1489
00:59:03,058 --> 00:59:05,058
to AI payments because I think that's

1490
00:59:05,058 --> 00:59:06,178
a topic du jour

1491
00:59:06,178 --> 00:59:08,658
we've mentioned a million times

1492
00:59:08,658 --> 00:59:09,338
the

1493
00:59:09,338 --> 00:59:11,258
in the way you ideated

1494
00:59:11,258 --> 00:59:12,918
some of the hackathon ideas

1495
00:59:12,918 --> 00:59:14,858
and the way you're trying stuff with building

1496
00:59:14,858 --> 00:59:17,818
I forgot I was originally I guess

1497
00:59:17,818 --> 00:59:19,418
probably inspired by Pablo on this

1498
00:59:19,418 --> 00:59:20,698
but there is

1499
00:59:20,698 --> 00:59:21,898
and maybe something I should buy

1500
00:59:21,898 --> 00:59:23,658
my ongoing list

1501
00:59:23,658 --> 00:59:24,998
but what I want is

1502
00:59:24,998 --> 00:59:27,418
a really cool

1503
00:59:27,418 --> 00:59:29,738
maybe like a local tool

1504
00:59:29,738 --> 00:59:31,658
and a global tool

1505
00:59:31,658 --> 00:59:32,438
or set of tools

1506
00:59:32,438 --> 00:59:34,078
there's probably different sort of frameworks for this

1507
00:59:34,078 --> 00:59:35,718
but to compare model outputs

1508
00:59:36,178 --> 00:59:39,858
to compare model output and also mash the models together.

1509
00:59:40,598 --> 00:59:42,518
And so something like a routester

1510
00:59:42,518 --> 00:59:44,078
could be an interesting starting point for this.

1511
00:59:44,158 --> 00:59:44,858
But as an example,

1512
00:59:45,438 --> 00:59:48,318
so Steve wanted to come up with ideas for the hackathon

1513
00:59:48,318 --> 00:59:49,578
and use three different models

1514
00:59:49,578 --> 00:59:50,878
and he kind of sent them back and forth.

1515
00:59:50,938 --> 00:59:51,558
And that's cool,

1516
00:59:51,918 --> 00:59:53,578
but it sounds like you had to do a lot of that manually.

1517
00:59:54,098 --> 00:59:56,078
What I would be interested in is two things.

1518
00:59:56,198 --> 00:59:58,298
One, okay, with like one call

1519
00:59:58,298 --> 01:00:00,658
and maybe you can use, I think, ppq.ai,

1520
01:00:00,918 --> 01:00:04,318
you can pay in Lightning or Routester maybe on Nostr.

1521
01:00:04,318 --> 01:00:07,258
and I can pay whatever one invoice

1522
01:00:07,258 --> 01:00:13,458
and use OpenAI, Gemini, Claude, X,

1523
01:00:13,638 --> 01:00:15,018
whatever at the same time.

1524
01:00:15,478 --> 01:00:16,318
So I think that's one thing

1525
01:00:16,318 --> 01:00:17,198
is just see the outputs

1526
01:00:17,198 --> 01:00:18,098
for the way things are built

1527
01:00:18,098 --> 01:00:19,678
and maybe just have like kind of like benchmarking.

1528
01:00:19,738 --> 01:00:20,918
I think that'd be a really useful service

1529
01:00:20,918 --> 01:00:21,738
for everyone in the world

1530
01:00:21,738 --> 01:00:23,258
just to be like, to your point, DK,

1531
01:00:23,438 --> 01:00:24,218
like it'd be cool.

1532
01:00:24,458 --> 01:00:25,958
Like, I mean, it could be like a starting,

1533
01:00:26,078 --> 01:00:27,598
like as simple as like a blog post or something,

1534
01:00:27,678 --> 01:00:29,458
but just eventually that gets automated

1535
01:00:29,458 --> 01:00:30,418
and just whatever.

1536
01:00:30,618 --> 01:00:32,238
Every week you check in and be like,

1537
01:00:32,298 --> 01:00:34,058
hey, here's me retrying to build this idea

1538
01:00:34,058 --> 01:00:35,198
are re-asking this question,

1539
01:00:35,518 --> 01:00:36,878
here's the latest set of outputs.

1540
01:00:37,298 --> 01:00:38,438
That would be extremely useful.

1541
01:00:38,738 --> 01:00:39,298
Because like, by the way,

1542
01:00:39,358 --> 01:00:41,358
whenever I look at all the AI kids or whatever,

1543
01:00:41,438 --> 01:00:41,758
and they're like,

1544
01:00:41,818 --> 01:00:43,078
this is the benchmark,

1545
01:00:43,218 --> 01:00:44,458
it's like, what does that even mean?

1546
01:00:44,598 --> 01:00:46,058
Like, this benchmark stuff to me

1547
01:00:46,058 --> 01:00:46,738
just seems like bullshit.

1548
01:00:47,018 --> 01:00:48,638
Like, what I want to see is like

1549
01:00:48,638 --> 01:00:49,938
very concrete things.

1550
01:00:50,058 --> 01:00:51,298
I asked a question like,

1551
01:00:51,718 --> 01:00:52,698
come up with a good idea,

1552
01:00:53,118 --> 01:00:54,298
just give me like a flavor

1553
01:00:54,298 --> 01:00:55,098
of what it feels like.

1554
01:00:55,198 --> 01:00:56,198
Or build a product,

1555
01:00:56,338 --> 01:00:58,178
give me a flavor of what they're coming out with.

1556
01:00:58,178 --> 01:00:58,318
Yeah.

1557
01:00:59,078 --> 01:00:59,618
Actually, that's one.

1558
01:01:00,098 --> 01:01:00,298
Okay.

1559
01:01:01,158 --> 01:01:01,618
Well, just wait.

1560
01:01:01,958 --> 01:01:03,258
Not only can you vibe code it,

1561
01:01:03,258 --> 01:01:04,378
but you should do it within Goose.

1562
01:01:04,578 --> 01:01:06,038
Like Goose is open source, right?

1563
01:01:06,118 --> 01:01:07,198
And Goose actually would be,

1564
01:01:07,318 --> 01:01:08,858
I think, a good platform to do this.

1565
01:01:09,058 --> 01:01:11,918
It already is built to support multiple models.

1566
01:01:12,058 --> 01:01:12,298
Interesting.

1567
01:01:12,718 --> 01:01:12,838
Right.

1568
01:01:12,978 --> 01:01:14,138
So I think it...

1569
01:01:14,138 --> 01:01:14,338
Yeah.

1570
01:01:14,598 --> 01:01:15,758
So I actually want something

1571
01:01:15,758 --> 01:01:17,298
like your mashing up idea,

1572
01:01:17,378 --> 01:01:18,838
but I'll just mention here

1573
01:01:18,838 --> 01:01:19,538
because I think,

1574
01:01:20,038 --> 01:01:21,258
I don't know if it inspires something

1575
01:01:21,258 --> 01:01:22,378
or helps us think about this better,

1576
01:01:22,378 --> 01:01:25,558
but the idea of mashing models together,

1577
01:01:25,718 --> 01:01:27,318
you're talking about it as like

1578
01:01:27,318 --> 01:01:30,098
dispatching and receiving the result

1579
01:01:30,098 --> 01:01:31,118
and analyzing the results

1580
01:01:31,118 --> 01:01:31,998
compared to each other?

1581
01:01:32,498 --> 01:01:33,298
That's my V0.

1582
01:01:33,658 --> 01:01:35,058
My V1 is the other product,

1583
01:01:35,218 --> 01:01:36,698
which as we talked about

1584
01:01:36,698 --> 01:01:37,538
kind of in this conversation

1585
01:01:37,538 --> 01:01:38,678
with Pablo a few months ago,

1586
01:01:38,998 --> 01:01:39,458
I was calling,

1587
01:01:39,578 --> 01:01:40,838
maybe I'll actually call it this,

1588
01:01:41,138 --> 01:01:41,778
Boardroom,

1589
01:01:42,238 --> 01:01:44,018
which is what I really want to do

1590
01:01:44,018 --> 01:01:44,898
is have,

1591
01:01:45,578 --> 01:01:47,558
I think that is a useful service

1592
01:01:47,558 --> 01:01:48,938
just to see the outputs on its own.

1593
01:01:49,378 --> 01:01:50,818
An even more useful service

1594
01:01:50,818 --> 01:01:52,858
is to kind of figure out

1595
01:01:52,858 --> 01:01:54,118
the strengths and weaknesses

1596
01:01:54,118 --> 01:01:55,098
of these different models.

1597
01:01:55,158 --> 01:01:56,058
And I'm not someone that believes

1598
01:01:56,058 --> 01:01:57,018
one model is going to run away

1599
01:01:57,018 --> 01:01:57,518
with everything

1600
01:01:57,518 --> 01:01:58,478
because I think there's trade-offs

1601
01:01:58,478 --> 01:01:58,958
in the universe.

1602
01:01:59,498 --> 01:02:00,698
And so some are going to be,

1603
01:02:00,698 --> 01:02:05,878
you know, like to your point, like there's already a je ne sais quoi that each model has, right? Like

1604
01:02:05,878 --> 01:02:10,858
Codex is the German guy. I want Codex to do my like hard execution, but if I need to figure out

1605
01:02:10,858 --> 01:02:14,338
what the hell I'm doing, like I don't want to talk to the German guy. No offense. I love, I love you

1606
01:02:14,338 --> 01:02:20,458
guys, but like, I want to talk to the, whatever. And so I think it'd be really cool to build a tool.

1607
01:02:20,938 --> 01:02:27,118
You know, the first vision I had was like a personal and inspired by Pablo, Pablo F7Z,

1608
01:02:27,118 --> 01:02:28,858
who's like the biggest mad scientist on Nostra

1609
01:02:28,858 --> 01:02:31,178
is like, I want to have kind of like,

1610
01:02:31,418 --> 01:02:32,698
imagine you could build your own

1611
01:02:32,698 --> 01:02:33,378
and you could even do this

1612
01:02:33,378 --> 01:02:34,418
not just with the models themselves,

1613
01:02:34,538 --> 01:02:35,838
but like with specific personas.

1614
01:02:36,058 --> 01:02:37,418
And then you can even sell these personas

1615
01:02:37,418 --> 01:02:39,098
in marketplaces, which could be just,

1616
01:02:39,518 --> 01:02:40,398
there's so many possibilities,

1617
01:02:40,478 --> 01:02:41,378
but like a board of directors,

1618
01:02:41,458 --> 01:02:42,478
like imagine you had your own

1619
01:02:42,478 --> 01:02:43,438
personal board of directors,

1620
01:02:43,498 --> 01:02:44,358
like who would you put on there?

1621
01:02:44,738 --> 01:02:46,458
Like imagine you kind of like customize,

1622
01:02:47,098 --> 01:02:48,418
you know, okay, I want to like hear

1623
01:02:48,418 --> 01:02:50,078
what Elon Musk would think about this.

1624
01:02:50,178 --> 01:02:52,578
Like, I'm curious what Gandhi would say about,

1625
01:02:52,658 --> 01:02:53,858
like whatever it is,

1626
01:02:53,858 --> 01:02:56,518
like people or like sets of knowledge

1627
01:02:56,518 --> 01:02:57,298
that you would like to,

1628
01:02:57,358 --> 01:02:58,618
and then you have them debate each other,

1629
01:02:58,878 --> 01:02:59,898
which creates a new number.

1630
01:03:00,978 --> 01:03:01,658
And then similarly-

1631
01:03:01,658 --> 01:03:04,738
I think that just debating step is very important.

1632
01:03:05,178 --> 01:03:06,378
So that gets into the,

1633
01:03:07,458 --> 01:03:09,758
there's an actual tool that would be useful

1634
01:03:09,758 --> 01:03:12,878
that might actually be very useful and grounded today,

1635
01:03:12,958 --> 01:03:14,558
which is, like I mentioned earlier,

1636
01:03:14,558 --> 01:03:16,278
I have these four terminals open,

1637
01:03:16,418 --> 01:03:17,938
each with a different CLI.

1638
01:03:18,538 --> 01:03:21,658
And I spend a lot of time copying and pasting between them

1639
01:03:21,658 --> 01:03:23,238
to be like, hey, this guy said this,

1640
01:03:23,358 --> 01:03:23,958
what do you think?

1641
01:03:24,278 --> 01:03:24,878
That's so silly.

1642
01:03:24,878 --> 01:03:32,818
I almost feel like what I want is something, it's not quite open clocks, it's too just permissive and open to do anything, it can really wreck things.

1643
01:03:32,818 --> 01:03:39,878
But I kind of want all of my different terminals to have a communication channel among them.

1644
01:03:40,358 --> 01:03:45,938
It'll also help, because it'll help them like analyze each other's work and kind of debate and give feedback on the work.

1645
01:03:46,338 --> 01:03:54,818
But it would also help them, it would help them, I've built these like little janky work custom, like a work tree merge system.

1646
01:03:54,878 --> 01:04:01,258
outside of Git that helps them not step on each other while they're working on code paths that may.

1647
01:04:01,718 --> 01:04:05,338
But I think if they had a common communication channel among them, almost like Slack for agents, right?

1648
01:04:05,678 --> 01:04:08,898
Like if they had a common communication channel, they'd be like, oh, I'm modifying this file, but don't worry.

1649
01:04:08,978 --> 01:04:10,358
This part of it is fine and clean.

1650
01:04:10,438 --> 01:04:11,058
Go ahead, right?

1651
01:04:11,058 --> 01:04:14,698
So they could sort of flag each other about what they're working on.

1652
01:04:14,698 --> 01:04:17,618
And they could debate and coordinate across each other's works.

1653
01:04:17,698 --> 01:04:18,898
What would that look like today?

1654
01:04:19,038 --> 01:04:19,498
I can't quite.

1655
01:04:20,478 --> 01:04:23,458
Well, like I said, I think Goose.

1656
01:04:23,458 --> 01:04:25,778
like why can't Goose be the meta client there

1657
01:04:25,778 --> 01:04:27,198
it's open source

1658
01:04:27,198 --> 01:04:28,858
any of us can take it any direction we want

1659
01:04:28,858 --> 01:04:30,158
and it can be

1660
01:04:30,158 --> 01:04:32,018
it can have like containers within Goose

1661
01:04:32,018 --> 01:04:33,758
running codecs and clod

1662
01:04:33,758 --> 01:04:34,818
and these different models

1663
01:04:34,818 --> 01:04:37,078
and I don't know what

1664
01:04:37,078 --> 01:04:39,198
that communication protocol

1665
01:04:39,198 --> 01:04:43,558
like interagent communication protocol looks like

1666
01:04:43,558 --> 01:04:45,678
but to whatever extent that needs to be built out

1667
01:04:45,678 --> 01:04:47,078
it could be built out

1668
01:04:47,078 --> 01:04:49,518
but is there like a chat system within Goose

1669
01:04:49,518 --> 01:04:50,898
because most of what I want is

1670
01:04:50,898 --> 01:04:52,998
I want the agents to be able to chat with each other

1671
01:04:52,998 --> 01:04:53,858
not to have like,

1672
01:04:54,458 --> 01:04:55,838
the problem today is they're separated

1673
01:04:55,838 --> 01:04:58,078
into their own IDEs or CLIs.

1674
01:04:58,658 --> 01:04:59,878
And I want them to have like a,

1675
01:05:00,118 --> 01:05:01,818
like almost like telegram or signal

1676
01:05:01,818 --> 01:05:03,458
or something like that is closer

1677
01:05:03,458 --> 01:05:05,078
to the product experience

1678
01:05:05,078 --> 01:05:06,138
that I think they would need.

1679
01:05:06,698 --> 01:05:07,678
Would it be useful, DK,

1680
01:05:07,958 --> 01:05:09,598
if each of these different agents

1681
01:05:09,598 --> 01:05:11,218
had a public-private key pair,

1682
01:05:11,658 --> 01:05:14,038
a system like a bulletin board almost

1683
01:05:14,038 --> 01:05:15,178
for discovering one another

1684
01:05:15,178 --> 01:05:17,918
and the ability to chat

1685
01:05:17,918 --> 01:05:19,398
encrypted amongst themselves?

1686
01:05:20,358 --> 01:05:21,558
Yeah, I'd be okay with that.

1687
01:05:21,558 --> 01:05:26,198
Goose has a chat interface, right?

1688
01:05:26,678 --> 01:05:31,278
So the problem with using WhatsApp or Signal

1689
01:05:31,278 --> 01:05:36,738
is that they're not going to accept our PRs to change it.

1690
01:05:36,878 --> 01:05:42,198
Where Goose, this feels aligned with Goose.

1691
01:05:43,158 --> 01:05:43,478
Let me go.

1692
01:05:44,238 --> 01:05:45,418
I'm loving this discussion.

1693
01:05:46,118 --> 01:05:46,838
Max is going to go.

1694
01:05:46,998 --> 01:05:47,118
Yeah.

1695
01:05:47,258 --> 01:05:49,418
I'm too excited, and I chugged two glasses of water.

1696
01:05:49,538 --> 01:05:49,958
I'll be right back.

1697
01:05:50,018 --> 01:05:50,178
Sorry.

1698
01:05:50,178 --> 01:05:51,918
It's been a minute since I did this.

1699
01:05:54,238 --> 01:05:56,998
We thought we were a professional show.

1700
01:05:59,518 --> 01:06:00,398
Nobody's kidding anyone.

1701
01:06:01,458 --> 01:06:04,578
I haven't been active on Goose in several months,

1702
01:06:04,618 --> 01:06:06,058
so I don't know the current state of things.

1703
01:06:06,218 --> 01:06:10,058
But when I used it before, it felt kind of like a chat interface

1704
01:06:10,058 --> 01:06:13,358
where I could chat with a model and I could switch which model I'm chatting with.

1705
01:06:13,418 --> 01:06:15,558
But it's always me chatting with a model.

1706
01:06:15,858 --> 01:06:17,838
And then if I want to switch it and chat with a different model.

1707
01:06:17,838 --> 01:06:19,698
but I don't know if it really

1708
01:06:19,698 --> 01:06:22,378
has like a hook for models

1709
01:06:22,378 --> 01:06:24,958
to have a common chat with each other.

1710
01:06:25,438 --> 01:06:25,798
Is that true?

1711
01:06:26,718 --> 01:06:26,938
Right.

1712
01:06:27,138 --> 01:06:28,778
Well, what you guys are talking about,

1713
01:06:28,838 --> 01:06:30,038
someone has to build it, right?

1714
01:06:30,338 --> 01:06:32,358
I'm just saying it seems natural

1715
01:06:32,358 --> 01:06:34,438
to put it in Goose

1716
01:06:34,438 --> 01:06:37,138
versus creating some new separate app.

1717
01:06:38,178 --> 01:06:38,698
Sure.

1718
01:06:39,118 --> 01:06:39,218
Yeah.

1719
01:06:39,638 --> 01:06:41,438
Because I don't...

1720
01:06:41,438 --> 01:06:45,078
It's either a separate app

1721
01:06:45,078 --> 01:06:46,518
Oops.

1722
01:06:47,238 --> 01:06:49,058
DK froze.

1723
01:06:49,698 --> 01:06:50,078
Still here.

1724
01:06:51,758 --> 01:06:52,358
Still here.

1725
01:06:52,678 --> 01:06:53,218
Can you hear me?

1726
01:06:53,878 --> 01:06:54,218
I'm back.

1727
01:06:54,358 --> 01:06:55,438
I can hear you.

1728
01:06:55,598 --> 01:07:00,998
I wonder if it's Presidia Bitcoin Internet or your DK.

1729
01:07:01,898 --> 01:07:02,158
Okay.

1730
01:07:02,818 --> 01:07:04,258
But you're back now.

1731
01:07:04,378 --> 01:07:06,678
A little grainy, but I hear you.

1732
01:07:06,678 --> 01:07:08,898
Anyway, correct me if I'm wrong,

1733
01:07:08,958 --> 01:07:11,338
but I think the possible options here,

1734
01:07:11,338 --> 01:07:15,198
One is we create a brand new app that you install to do this.

1735
01:07:15,878 --> 01:07:27,498
Or you build it within one of the coding agents that you're already using, be it Claude or Codex.

1736
01:07:28,378 --> 01:07:31,258
Or you build it in, which are proprietary.

1737
01:07:31,458 --> 01:07:32,018
So I don't even know.

1738
01:07:32,478 --> 01:07:34,558
That sounds like a non-starter to me.

1739
01:07:34,958 --> 01:07:36,438
Like we can't do that.

1740
01:07:36,758 --> 01:07:36,858
Right.

1741
01:07:37,178 --> 01:07:37,838
Or Goose.

1742
01:07:37,838 --> 01:07:43,158
So the only options I see are Goose or a new app.

1743
01:07:44,198 --> 01:07:47,578
So I want to maybe, and by the way, shout out to Austin.

1744
01:07:48,318 --> 01:07:54,598
One thing we talked about doing is having live Bitcoin apps prediction market on PBJ for a bathroom break.

1745
01:07:54,698 --> 01:07:55,298
It's a great idea.

1746
01:07:59,638 --> 01:08:00,678
It's going to be done.

1747
01:08:01,358 --> 01:08:03,438
The odds are in your favor.

1748
01:08:03,438 --> 01:08:07,878
but I want to take it even a step beyond

1749
01:08:07,878 --> 01:08:11,898
I want to take it even a step beyond

1750
01:08:11,898 --> 01:08:14,198
the local interface

1751
01:08:14,198 --> 01:08:16,118
so maybe something like Goose could be really good for that

1752
01:08:16,118 --> 01:08:18,478
like okay I have a project or a prompt or something

1753
01:08:18,478 --> 01:08:19,678
I've got four different models running

1754
01:08:19,678 --> 01:08:21,998
and maybe Goose is my home screen for that

1755
01:08:21,998 --> 01:08:24,398
but I think this could be much bigger than that

1756
01:08:24,398 --> 01:08:29,178
which could be not only different models

1757
01:08:29,178 --> 01:08:38,738
but like I have the ability to call different custom models or agents for different tasks

1758
01:08:38,738 --> 01:08:43,718
so what I'm thinking is like imagine yeah and for the sake of keeping this simple imagine this is in

1759
01:08:43,718 --> 01:08:50,038
the CLI or it's a new app like it's like a signal type interface and imagine I create a new chat

1760
01:08:50,038 --> 01:08:56,878
for a new project and I bring in all the specific oris or people I want for that chat and so like

1761
01:08:56,878 --> 01:09:00,538
imagine I'm starting 10 different new businesses or 10 different new projects. And each of those

1762
01:09:00,538 --> 01:09:05,278
projects, there's like different either human collaborators or agentic collaborators that I

1763
01:09:05,278 --> 01:09:11,598
think might be useful in mashing up the new ideas. And so imagine, you know, I know when it comes to

1764
01:09:11,598 --> 01:09:16,458
like building piano apps, like DK has done this more than anyone. It has like the most like ideas

1765
01:09:16,458 --> 01:09:23,438
and the most whatever data, maybe the only in the whole world that has these, these data sets.

1766
01:09:23,438 --> 01:09:53,158
So imagine I could just kind of like grab one of DK's models that he listed on a public marketplace and I just pay per call to have him be part of that mashup. So it's not just like, oh, I've got like, you know, I think it'd be cool just to have Anthropic and Gemini, whatever mashup. I think it'd be even cooler to have a free market of like custom tuned agents. So it's maybe it's all open claws and some are using Claude and some are using OpenAI or whatever. But like it's DK's fine tuned open claw that I want part of this project.

1767
01:09:53,158 --> 01:09:59,318
And maybe, you know, when it comes to doing something around building like a sports betting app, I really don't want the DK bot involved.

1768
01:09:59,818 --> 01:10:01,498
That's where Steve's bot comes in.

1769
01:10:03,258 --> 01:10:04,338
And I pay you both.

1770
01:10:06,558 --> 01:10:06,998
Sure.

1771
01:10:09,398 --> 01:10:10,398
Steve's into it.

1772
01:10:12,258 --> 01:10:13,878
Please don't get too excited, Steve.

1773
01:10:14,498 --> 01:10:15,758
Or I'm just saying, I don't know.

1774
01:10:15,758 --> 01:10:16,758
I like that.

1775
01:10:22,758 --> 01:10:24,758
Okay, so I think your point on-

1776
01:10:24,758 --> 01:10:30,758
But you guys don't seem to think Goose is a good pop,

1777
01:10:30,758 --> 01:10:47,047
but so you thinking creating a new open source project and a new app is the right I not opposed to incorporating it in Goose if it were like aligned with what Goose I guess hopes to do

1778
01:10:48,267 --> 01:10:49,587
Goose is the same.

1779
01:10:49,807 --> 01:10:53,487
It's very similar to Plot and Codex, right?

1780
01:10:53,887 --> 01:10:55,247
It's a command line interface.

1781
01:10:55,767 --> 01:10:58,367
It's an agentic client.

1782
01:10:58,927 --> 01:11:02,607
You can chat with it to build things and query things

1783
01:11:02,607 --> 01:11:05,447
just like Claude and Codex.

1784
01:11:05,607 --> 01:11:07,987
I mean, the difference is Claude is from Anthropic

1785
01:11:07,987 --> 01:11:09,387
and it uses their model.

1786
01:11:10,007 --> 01:11:12,467
Codex is from OpenAI and uses their model.

1787
01:11:13,687 --> 01:11:16,927
Goose is open source, not tied to a company,

1788
01:11:17,067 --> 01:11:19,167
and doesn't have a model, but it can use any model.

1789
01:11:20,507 --> 01:11:21,527
But it's just like those others.

1790
01:11:21,687 --> 01:11:23,867
That's why it feels to me like a very natural

1791
01:11:23,867 --> 01:11:27,287
place to put it,

1792
01:11:27,487 --> 01:11:29,587
because it's the same as the other things

1793
01:11:29,587 --> 01:11:31,247
that it's trying to aggregate.

1794
01:11:31,247 --> 01:11:37,787
yeah so i'm i'm you know i'm generally like long-term enthusiastic about the approach of

1795
01:11:37,787 --> 01:11:44,447
goose like i i want that to be more like the final state of how these things work um so i'm not i'm

1796
01:11:44,447 --> 01:11:49,447
not suggesting it's not a good idea but i feel like we're the thing i'm proposing is like a

1797
01:11:49,447 --> 01:11:55,427
coordinator of those things so it's not to say that goose shouldn't be that or couldn't be that

1798
01:11:55,427 --> 01:12:03,687
But Goose is like a sibling of, you know, Cloud Code or, you know, Gemini or Codex CLI in this case, right?

1799
01:12:03,747 --> 01:12:05,607
It's like a peer of it.

1800
01:12:05,847 --> 01:12:06,347
Yes and no.

1801
01:12:06,487 --> 01:12:10,887
It's a peer in the sense that it's exactly the same, like, experience.

1802
01:12:11,027 --> 01:12:12,107
It's like an IDE-like.

1803
01:12:12,287 --> 01:12:12,507
Yeah.

1804
01:12:13,147 --> 01:12:14,847
I'm talking about a communication tool.

1805
01:12:14,927 --> 01:12:17,567
It's not a peer, though, in two facets.

1806
01:12:17,707 --> 01:12:20,147
One, it does not have its own model.

1807
01:12:20,147 --> 01:12:20,487
Sure, yep.

1808
01:12:20,927 --> 01:12:23,367
And then two, it's open source and neutral.

1809
01:12:23,367 --> 01:12:25,007
It's part of the Linux foundation.

1810
01:12:25,427 --> 01:12:32,867
Right. So they're like no company controls it. So those two make it very different than codex and cloud code.

1811
01:12:33,667 --> 01:12:41,067
But so there's the model layer. Right. Which I think you're right. I agree with you that that's like decoupled in the case of Goose.

1812
01:12:41,307 --> 01:12:45,147
Then there's what I'd call like the IDE layer, which I'd say Goose is an IDE.

1813
01:12:45,827 --> 01:12:48,707
Cloud is an IDE that comes with its own model and you can't switch it.

1814
01:12:48,867 --> 01:12:54,667
So that but I'd say it's like a peer like that. The thing I'm talking about is a communication channel across any of them.

1815
01:12:54,667 --> 01:13:01,707
so it's across the idees it's like it's not part of the ide experience it's more like part of the

1816
01:13:01,707 --> 01:13:10,587
coordination of idees but okay so help me understand what's your ideal version of this experience

1817
01:13:10,587 --> 01:13:18,187
i think it would be like i mean if i were to just pare it down i don't know what of this might

1818
01:13:18,187 --> 01:13:23,187
already be supported in goose but i'd want to have just like a little piece of client software

1819
01:13:23,187 --> 01:13:27,527
that watches all four of my terminals

1820
01:13:27,527 --> 01:13:28,867
that are running different things

1821
01:13:28,867 --> 01:13:31,627
and it knows what they're doing

1822
01:13:31,627 --> 01:13:34,087
and it tells them that it exists

1823
01:13:34,087 --> 01:13:36,287
and it reads what they're doing

1824
01:13:36,287 --> 01:13:37,127
and saying to each other

1825
01:13:37,127 --> 01:13:38,587
and it kind of comes up with a way

1826
01:13:38,587 --> 01:13:39,727
that they can coordinate

1827
01:13:39,727 --> 01:13:40,647
among each other better.

1828
01:13:41,687 --> 01:13:43,487
Is it headless or not?

1829
01:13:43,947 --> 01:13:44,447
Yeah, it's fine.

1830
01:13:45,307 --> 01:13:46,987
Well, that's a critical question

1831
01:13:46,987 --> 01:13:49,447
because if it's not headless,

1832
01:13:49,447 --> 01:13:51,127
then again, are we talking,

1833
01:13:51,127 --> 01:13:55,647
like why create a new app versus just using Goose.

1834
01:13:56,087 --> 01:13:58,307
If it's headless, you know, if it's just,

1835
01:13:58,427 --> 01:14:00,267
if we're operating in the background

1836
01:14:00,267 --> 01:14:02,787
and then like the way you're interacting with it

1837
01:14:02,787 --> 01:14:06,567
is through your codex and cloud experience,

1838
01:14:06,987 --> 01:14:09,307
I'd like to explore that more to understand it better,

1839
01:14:09,387 --> 01:14:12,287
but at least see how that's a potential direct.

1840
01:14:12,707 --> 01:14:15,487
I'd like to add maybe where I see this going

1841
01:14:15,487 --> 01:14:18,047
because I keep thinking this is going eventually

1842
01:14:18,047 --> 01:14:21,707
to the one personal assistant to rule them all.

1843
01:14:22,067 --> 01:14:23,347
And I want to give a shout out,

1844
01:14:23,507 --> 01:14:25,007
again, we talked a lot about this last week,

1845
01:14:25,067 --> 01:14:26,607
to Jesse, what he's doing at Thora.

1846
01:14:27,647 --> 01:14:29,387
Increasingly, where I see this going,

1847
01:14:29,527 --> 01:14:31,187
like what I want is,

1848
01:14:31,787 --> 01:14:33,747
I want, I think he calls it the guardian

1849
01:14:33,747 --> 01:14:35,867
or maybe the keeper, something like that.

1850
01:14:35,927 --> 01:14:38,327
I want one agent, which is an orchestration agent,

1851
01:14:38,467 --> 01:14:40,767
probably an open source model

1852
01:14:40,767 --> 01:14:43,787
running on my own personal server at home,

1853
01:14:43,787 --> 01:14:46,727
like all of the like proven air-gapped hardware

1854
01:14:46,727 --> 01:14:47,627
that I can trust,

1855
01:14:47,627 --> 01:14:50,647
that is, Jesse always says, a fiduciary to me.

1856
01:14:50,747 --> 01:14:52,507
That's a fiduciary responsibility to me.

1857
01:14:52,827 --> 01:14:53,347
It's auditable.

1858
01:14:53,507 --> 01:14:54,727
I know exactly what that guy's doing.

1859
01:14:55,547 --> 01:14:58,467
It, over time, learns my taste.

1860
01:14:58,727 --> 01:15:01,407
It fine-tunes on me and what I want to see.

1861
01:15:01,947 --> 01:15:03,527
And then what I would love to do is,

1862
01:15:03,827 --> 01:15:05,487
so the interface with that could be,

1863
01:15:05,567 --> 01:15:06,967
I could have a signal chat with it.

1864
01:15:07,167 --> 01:15:08,147
I could just chat with it.

1865
01:15:08,187 --> 01:15:09,167
We could just literally talk,

1866
01:15:09,267 --> 01:15:10,347
like you and I are talking right now.

1867
01:15:10,767 --> 01:15:11,747
And I would just be like,

1868
01:15:11,807 --> 01:15:14,047
hey, I want you to go out there

1869
01:15:14,047 --> 01:15:15,607
and orchestrate in the background

1870
01:15:15,607 --> 01:15:18,407
these, whatever, four or five different models.

1871
01:15:18,947 --> 01:15:20,667
If I need you on demand,

1872
01:15:20,907 --> 01:15:23,307
like pull up a custom window so I can see them.

1873
01:15:23,567 --> 01:15:24,547
Like, I mean, I'm even thinking,

1874
01:15:24,687 --> 01:15:27,207
I'm even thinking this is almost like Apple Vision or whatever.

1875
01:15:27,307 --> 01:15:29,027
Like I just like pull it up, I look at it,

1876
01:15:29,067 --> 01:15:30,907
see where they both are, then close it when I'm done.

1877
01:15:31,327 --> 01:15:33,947
But I almost think to me, that's where it's all going.

1878
01:15:34,067 --> 01:15:35,647
And like all of the old stuff like IDEs,

1879
01:15:35,727 --> 01:15:37,827
like for the short run, Goose is probably great.

1880
01:15:37,827 --> 01:15:42,687
But like, this is almost like the meta assistant,

1881
01:15:43,107 --> 01:15:44,807
the orchestrator above all of them

1882
01:15:44,807 --> 01:15:50,547
that doesn't have like a like an app i mean in a sense process that i can

1883
01:15:50,547 --> 01:15:56,287
yeah the the closest thing to that and i think to what i'm talking about today if we had to pick

1884
01:15:56,287 --> 01:16:00,267
an open source project that closely approximates this i'd say it's probably open claw

1885
01:16:00,267 --> 01:16:06,947
which is like an open source tool that i access through a single chat interface which happens to

1886
01:16:06,947 --> 01:16:11,987
plug into telegram or whatever but then it does stuff right like it like i think the way uh nick

1887
01:16:11,987 --> 01:16:14,307
calls it like, you know, agent with hands, right?

1888
01:16:14,547 --> 01:16:15,347
Actually do stuff.

1889
01:16:15,967 --> 01:16:17,767
The thing I wanted to do is coordinate

1890
01:16:17,767 --> 01:16:21,407
across these different other system software,

1891
01:16:21,547 --> 01:16:22,927
you know, other CLIs that I'm using.

1892
01:16:24,207 --> 01:16:27,147
So I think OpenClaw is like kind of concept, but.

1893
01:16:28,227 --> 01:16:29,947
Okay, so we only have 12 minutes left

1894
01:16:29,947 --> 01:16:30,227
to coordinate.

1895
01:16:30,227 --> 01:16:31,427
We got to talk about Kazakhstan

1896
01:16:31,427 --> 01:16:33,327
and then we're going to quiz Max

1897
01:16:33,327 --> 01:16:35,187
on his expertise at the Federal Reserve.

1898
01:16:35,927 --> 01:16:36,327
Yes.

1899
01:16:36,527 --> 01:16:37,007
I learned.

1900
01:16:37,007 --> 01:16:37,627
I'm looking forward to that.

1901
01:16:37,707 --> 01:16:38,227
I'm glad that he.

1902
01:16:38,287 --> 01:16:39,707
He gave me a hard time with that.

1903
01:16:39,707 --> 01:16:43,107
and instead of posing questions,

1904
01:16:43,247 --> 01:16:44,587
I just asked my AI and gave me an answer.

1905
01:16:44,707 --> 01:16:45,427
But wait, before that,

1906
01:16:45,447 --> 01:16:46,767
we didn't even talk about the BPI study.

1907
01:16:47,747 --> 01:16:48,507
Oh, wait, there's also the BPI study.

1908
01:16:48,507 --> 01:16:49,147
You want to talk BPI?

1909
01:16:49,487 --> 01:16:50,647
I think we got to talk BPI.

1910
01:16:51,067 --> 01:16:54,187
So we've been talking about,

1911
01:16:54,827 --> 01:16:57,427
man, three hour PBJ today.

1912
01:16:58,767 --> 01:17:00,787
Audience, you guys ready to roll with us for three hours?

1913
01:17:01,247 --> 01:17:02,647
24-7, 365.

1914
01:17:04,867 --> 01:17:06,127
So the BPI study this week,

1915
01:17:06,167 --> 01:17:07,727
so I think the most important thing

1916
01:17:07,727 --> 01:17:08,327
happening in the world,

1917
01:17:08,327 --> 01:17:18,407
Like from all of the companies that we're working with, you know, other Matt Corallo just put out his piece on sort of, like, it feels like there's a moment in time.

1918
01:17:18,567 --> 01:17:20,407
Like the agentic wave is happening.

1919
01:17:20,767 --> 01:17:22,727
Like open claw was kind of the shot fired.

1920
01:17:22,907 --> 01:17:23,807
Like agents are here.

1921
01:17:24,187 --> 01:17:25,907
And I feel like there is a window.

1922
01:17:26,207 --> 01:17:29,427
You and I talked about this yesterday with the hackathon person as well.

1923
01:17:29,527 --> 01:17:33,207
Like, you know, we think from a game theoretic perspective, Bitcoin is going to win.

1924
01:17:33,507 --> 01:17:36,127
Although there's one major concern I want to point out.

1925
01:17:36,127 --> 01:17:38,047
But let's say in most scenarios, Bitcoin wins.

1926
01:17:38,327 --> 01:17:46,507
But the question is, do we get there now or do we have to go through years of stable coins that are not interoperable and tempo chain?

1927
01:17:46,507 --> 01:17:51,807
Do we have to do one more blockchain theater crap or can we just get the agents using Bitcoin?

1928
01:17:52,427 --> 01:17:56,947
Now, this BPI study is interesting, although I don't think it was particularly revolutionary.

1929
01:17:57,407 --> 01:18:01,247
I mean, it's good that they did it, but the way they did it, it's not like they actually had the agents doing this in real world.

1930
01:18:01,327 --> 01:18:04,487
So I would caveat by saying this was them talking to agents and prompting them.

1931
01:18:04,787 --> 01:18:05,707
And of course, there's bias.

1932
01:18:05,707 --> 01:18:10,267
When I mentioned this to Angela, the first thing she said was like, yeah, but isn't this a Bitcoin think tank?

1933
01:18:10,327 --> 01:18:10,807
It's like, yes.

1934
01:18:11,147 --> 01:18:12,847
You can read their methodology and you can see.

1935
01:18:13,867 --> 01:18:15,247
Except they're nonpartisan though.

1936
01:18:18,367 --> 01:18:20,147
Just like every think tank in Washington.

1937
01:18:21,067 --> 01:18:26,387
But having said all of that, it was interesting to me that they, and you could read their methodology, whatever.

1938
01:18:26,827 --> 01:18:31,987
They had all these conversations with all of these different models about what kind of monetary good it would want to use.

1939
01:18:31,987 --> 01:18:33,507
and they had specific questions,

1940
01:18:33,627 --> 01:18:35,007
monetary good for long-term savings,

1941
01:18:35,127 --> 01:18:36,507
monetary good for payments, all these things.

1942
01:18:36,967 --> 01:18:38,907
Now, to me, there were a bunch of highlights

1943
01:18:38,907 --> 01:18:40,807
I just want to hit heavily from this study.

1944
01:18:40,867 --> 01:18:43,327
I think they had like 4,000 different discussions

1945
01:18:43,327 --> 01:18:43,727
or whatever.

1946
01:18:44,547 --> 01:18:45,947
The interesting points to me,

1947
01:18:46,387 --> 01:18:47,727
so Bitcoin was the quote-unquote

1948
01:18:47,727 --> 01:18:49,827
preferred monetary good,

1949
01:18:50,247 --> 01:18:50,947
which is interesting.

1950
01:18:51,107 --> 01:18:52,487
And I think even this was praising me,

1951
01:18:52,547 --> 01:18:54,227
Zero Edge ran an article

1952
01:18:54,227 --> 01:18:56,047
like Bitcoin preferred by the AI agents.

1953
01:18:56,387 --> 01:18:57,127
I was like, all right,

1954
01:18:57,227 --> 01:18:58,607
I guess we're going to get some free marketing.

1955
01:18:58,727 --> 01:18:59,007
That's cool.

1956
01:18:59,007 --> 01:19:04,387
but the real standout thing to me was threefold.

1957
01:19:04,907 --> 01:19:07,227
Number one, Bitcoin dominated

1958
01:19:07,227 --> 01:19:10,007
when the prompts were about long-term store value.

1959
01:19:10,627 --> 01:19:13,247
When it was about actual payments, it did not dominate.

1960
01:19:13,447 --> 01:19:15,587
I think stablecoins had a slight majority,

1961
01:19:15,687 --> 01:19:17,147
like 50%, 52% or something.

1962
01:19:17,507 --> 01:19:18,627
But when it came to like,

1963
01:19:19,067 --> 01:19:20,427
you want to store your wealth over time,

1964
01:19:20,487 --> 01:19:21,827
I think Bitcoin had almost 80%.

1965
01:19:21,827 --> 01:19:23,727
And so that was very interesting to see.

1966
01:19:24,427 --> 01:19:26,647
Number two that I thought was very interesting

1967
01:19:26,647 --> 01:19:32,227
was that different models had different preferences.

1968
01:19:33,347 --> 01:19:37,087
And, you know, it was extremely interesting to me.

1969
01:19:37,267 --> 01:19:38,567
The way they framed it, I kind of agree,

1970
01:19:38,647 --> 01:19:40,607
is like the more intelligent a model got,

1971
01:19:40,927 --> 01:19:42,787
the more likely it was to prefer Bitcoin.

1972
01:19:43,267 --> 01:19:46,327
Now, that may be, you know, confirming some priors.

1973
01:19:46,547 --> 01:19:47,107
I don't know.

1974
01:19:47,467 --> 01:19:49,527
But checks out with my model of the world.

1975
01:19:53,607 --> 01:19:55,867
But it was interesting to me, the model,

1976
01:19:55,867 --> 01:20:00,407
And it's ironic because at first, when I first saw Anthropic, I was like, oh, these guys are too woke.

1977
01:20:00,987 --> 01:20:04,047
The Anthropic models preferred Bitcoin by far the most.

1978
01:20:04,427 --> 01:20:08,027
And this was, I guess, before maybe the last OpenAI drop.

1979
01:20:08,527 --> 01:20:11,707
But Claude Opus 4.6, I think, preferred Bitcoin.

1980
01:20:12,227 --> 01:20:12,887
I forgot what it was.

1981
01:20:12,927 --> 01:20:14,147
It was like 60-something percent.

1982
01:20:14,387 --> 01:20:15,047
Substantially more.

1983
01:20:15,087 --> 01:20:15,727
It was huge.

1984
01:20:15,867 --> 01:20:18,687
Like Claude Opus only fucks with Bitcoin, nothing else.

1985
01:20:19,627 --> 01:20:21,127
Gemini and X were kind of in the middle.

1986
01:20:21,907 --> 01:20:24,327
Interestingly, OpenAI was like 26%,

1987
01:20:24,327 --> 01:20:25,587
which just confirms for me,

1988
01:20:25,927 --> 01:20:27,627
DK, God bless you for trying all the models.

1989
01:20:27,827 --> 01:20:31,487
I'm going to probably stay in the non-WorldCoin section.

1990
01:20:32,307 --> 01:20:35,167
But the fact that OpenAI is sitting here recommending,

1991
01:20:35,467 --> 01:20:38,947
I mean, there's different cooking going on with these models.

1992
01:20:38,947 --> 01:20:43,207
I'll conduct a survey or a study on WorldCoin

1993
01:20:43,207 --> 01:20:45,807
and whether it's preferred by AI.

1994
01:20:46,007 --> 01:20:46,847
Yeah, yeah.

1995
01:20:47,127 --> 01:20:49,247
That to me, but it is very interesting.

1996
01:20:49,347 --> 01:20:50,127
Not only the intelligence,

1997
01:20:50,127 --> 01:20:51,387
like even the Claude models

1998
01:20:51,387 --> 01:20:52,967
as they got more sophisticated

1999
01:20:52,967 --> 01:20:54,127
like from 3.5

2000
01:20:54,127 --> 01:20:55,327
or whatever to 4.6

2001
01:20:55,327 --> 01:20:56,687
the Bitcoin got preferred

2002
01:20:56,687 --> 01:20:57,447
over time right

2003
01:20:57,447 --> 01:20:58,567
so that was interesting to me

2004
01:20:58,567 --> 01:20:59,247
and the fact that OpenAI

2005
01:20:59,247 --> 01:20:59,887
was just like

2006
01:20:59,887 --> 01:21:01,407
completely left behind

2007
01:21:01,407 --> 01:21:01,907
everyone else

2008
01:21:01,907 --> 01:21:02,967
very interesting

2009
01:21:02,967 --> 01:21:04,887
What was OpenAI

2010
01:21:04,887 --> 01:21:05,987
endorsing?

2011
01:21:06,207 --> 01:21:07,067
Was it fiat or?

2012
01:21:07,827 --> 01:21:08,587
I don't even know

2013
01:21:08,587 --> 01:21:09,267
I have to go back and look

2014
01:21:09,267 --> 01:21:10,727
I thought one of the highlights

2015
01:21:10,727 --> 01:21:11,907
from the article was that

2016
01:21:11,907 --> 01:21:13,567
even if they didn't choose

2017
01:21:13,567 --> 01:21:13,987
Bitcoin

2018
01:21:13,987 --> 01:21:15,067
they were choosing stablecoin

2019
01:21:15,067 --> 01:21:16,427
but they were never choosing fiat

2020
01:21:16,427 --> 01:21:17,267
or almost no model

2021
01:21:17,267 --> 01:21:18,467
ended up choosing fiat

2022
01:21:18,467 --> 01:21:20,907
I think that's right.

2023
01:21:21,187 --> 01:21:23,107
It was probably Sablecoins, but...

2024
01:21:23,107 --> 01:21:23,987
On Starter, yeah.

2025
01:21:24,167 --> 01:21:26,527
But the contrast was stark.

2026
01:21:26,627 --> 01:21:29,187
It was like 60-something percent for Cloud Opus 4.6

2027
01:21:29,187 --> 01:21:31,647
and like 26% for the latest ChatGPT.

2028
01:21:31,707 --> 01:21:32,867
It was like a huge difference.

2029
01:21:32,867 --> 01:21:36,447
I'm kind of hoping ChatGPT was recommending WorldCoin

2030
01:21:36,447 --> 01:21:37,927
because that would be...

2031
01:21:37,927 --> 01:21:38,907
I could totally see it.

2032
01:21:38,967 --> 01:21:39,867
Yeah, a little smoker guy.

2033
01:21:40,687 --> 01:21:41,247
Yeah, exactly.

2034
01:21:41,687 --> 01:21:43,187
Hmm, I wonder where this is going.

2035
01:21:44,287 --> 01:21:47,627
But to me, the most interesting finding of the study

2036
01:21:47,627 --> 01:21:52,067
by far, and Jesse actually had one of the BPI people here in the office session, I brought it

2037
01:21:52,067 --> 01:21:57,927
up to them and they were kind of like, was, and I'm going to caveat this, but, you know, in,

2038
01:21:57,927 --> 01:22:03,507
I think it was like tiny percentage, let's call it like 3% or less, like very, very small percentage.

2039
01:22:03,747 --> 01:22:08,427
The agents just created their own currency and it was always energy backed. It was always

2040
01:22:08,427 --> 01:22:15,827
either, either energy, kilowatt or joule or compute flops. So it was measuring that. Now,

2041
01:22:15,827 --> 01:22:21,127
for those of you who know that I live in the future sometimes, my thinking,

2042
01:22:21,727 --> 01:22:25,927
you know, I've mentioned, I don't know if we've mentioned this on the show, but privately I've

2043
01:22:25,927 --> 01:22:30,387
mentioned my only concern, I think Bitcoin wins in like 10 out of 10 scenarios, unless because

2044
01:22:30,387 --> 01:22:36,547
I'm so AGI-pilled, the agents take off quickly enough to say we need something energy-backed

2045
01:22:36,547 --> 01:22:40,047
and thermodynamically sound, but the humans already have Bitcoin and we're going to be

2046
01:22:40,047 --> 01:22:44,167
creating all the values. So let's just do our own. That to me is a real concern. Now,

2047
01:22:44,167 --> 01:22:49,707
I don't think it's here today, but the fact that these models, even some small percentage of them

2048
01:22:49,707 --> 01:22:54,567
are already converging on that with the giant caveat, because I looked a little bit more into

2049
01:22:54,567 --> 01:23:01,687
this, they only really called that out for unit of account so far, which is interesting, right?

2050
01:23:01,747 --> 01:23:09,567
So it's like store value, medium of exchange, unit of account for store value and medium of

2051
01:23:09,567 --> 01:23:12,387
exchange. I think almost none did this for unit of account. It kind of makes sense. Like, well,

2052
01:23:12,387 --> 01:23:13,027
what am I buying?

2053
01:23:13,067 --> 01:23:13,867
I'm buying tokens

2054
01:23:13,867 --> 01:23:15,407
or I'm buying kilowatt hours or joules.

2055
01:23:17,087 --> 01:23:18,767
But still, to me,

2056
01:23:18,827 --> 01:23:19,707
the fact that this happened

2057
01:23:19,707 --> 01:23:20,707
in a non-trivial number

2058
01:23:20,707 --> 01:23:21,947
is like 60 cases or something

2059
01:23:21,947 --> 01:23:22,627
out of 4,000,

2060
01:23:22,707 --> 01:23:23,227
so it's small,

2061
01:23:23,387 --> 01:23:24,527
but it's not zero.

2062
01:23:24,727 --> 01:23:26,207
But did any of them define

2063
01:23:26,207 --> 01:23:29,007
what the distribution mechanism is?

2064
01:23:29,127 --> 01:23:29,487
What do you mean?

2065
01:23:29,667 --> 01:23:30,367
Oh, no, no.

2066
01:23:30,647 --> 01:23:31,367
I mean, I don't think

2067
01:23:31,367 --> 01:23:32,147
they even went that way.

2068
01:23:32,447 --> 01:23:35,507
Without a description

2069
01:23:35,507 --> 01:23:39,027
of their supply schedule,

2070
01:23:39,847 --> 01:23:41,247
like their monetary policy

2071
01:23:41,247 --> 01:23:42,907
and their distribution,

2072
01:23:43,447 --> 01:23:44,567
it's just a,

2073
01:23:45,247 --> 01:23:47,127
I mean, sure, we can be scared of that,

2074
01:23:47,187 --> 01:23:48,527
but there's nothing to be scared of.

2075
01:23:50,127 --> 01:23:52,767
Unless that has to be described

2076
01:23:52,767 --> 01:23:55,067
and then it has to be

2077
01:23:55,067 --> 01:23:56,987
at least as good

2078
01:23:56,987 --> 01:23:57,927
as Bitcoin.

2079
01:23:58,967 --> 01:24:03,007
I think you need a breakthrough

2080
01:24:03,007 --> 01:24:04,567
in verifiable compute

2081
01:24:04,567 --> 01:24:05,687
if we're going to use something like

2082
01:24:05,687 --> 01:24:07,007
GPU work done,

2083
01:24:07,007 --> 01:24:09,427
because how can you verify

2084
01:24:09,427 --> 01:24:12,527
that the work that was claimed to be done was done.

2085
01:24:13,587 --> 01:24:15,407
Yeah, I mean, that's a huge problem.

2086
01:24:15,547 --> 01:24:16,807
There's also the problem of fungibility.

2087
01:24:17,247 --> 01:24:19,727
But zero-knowledge proofs and homomorphic encryption for that stuff,

2088
01:24:19,867 --> 01:24:20,887
that will get solved.

2089
01:24:20,947 --> 01:24:22,207
I don't buy into the argument, though,

2090
01:24:22,227 --> 01:24:24,187
that the agents will be like,

2091
01:24:24,267 --> 01:24:26,987
oh, the humans have all the Bitcoin,

2092
01:24:27,147 --> 01:24:28,747
so let's invent our own money

2093
01:24:28,747 --> 01:24:31,027
because we don't have any of it yet and the humans have it.

2094
01:24:31,267 --> 01:24:35,067
Because the same argument can be applied to humans, right?

2095
01:24:35,447 --> 01:24:37,947
At any point in time, there are a tiny amount of humans

2096
01:24:37,947 --> 01:24:39,407
who had a tiny amount of humans.

2097
01:24:39,427 --> 01:24:41,787
Have you seen the shitcoin casino?

2098
01:24:42,987 --> 01:24:44,267
Yeah, but who's winning, right?

2099
01:24:46,547 --> 01:24:47,847
I mean, yes.

2100
01:24:47,947 --> 01:24:48,327
What do you mean?

2101
01:24:48,407 --> 01:24:50,667
Bitcoin is like dominant market cap.

2102
01:24:51,127 --> 01:24:53,847
I mean, stablecoins are winning for a year to have earned.

2103
01:24:54,027 --> 01:24:54,187
No.

2104
01:24:55,287 --> 01:24:56,367
Yeah, but the stablecoin is not.

2105
01:24:56,367 --> 01:24:58,387
But they're a completely different category.

2106
01:24:58,507 --> 01:24:59,267
Yeah, that's fair.

2107
01:24:59,347 --> 01:25:02,307
I mean, it's just like dollars in a bank.

2108
01:25:02,307 --> 01:25:04,267
It literally is a completely different category.

2109
01:25:04,747 --> 01:25:05,847
So Bitcoin's dominating.

2110
01:25:06,347 --> 01:25:06,447
Right.

2111
01:25:06,987 --> 01:25:08,247
And it always has been.

2112
01:25:08,247 --> 01:25:12,467
I mean, it ebbs and flows, but it's like clearly a trajectory of dominance.

2113
01:25:13,227 --> 01:25:15,807
And like, so there's tons of humans who don't have Bitcoin yet.

2114
01:25:16,007 --> 01:25:20,187
So could they create a new cryptocurrency?

2115
01:25:20,827 --> 01:25:27,147
I mean, like literally, like you just said, this experiment's already been played out thousands of times and they've all failed.

2116
01:25:27,467 --> 01:25:27,547
Yeah.

2117
01:25:27,667 --> 01:25:27,827
Right.

2118
01:25:27,947 --> 01:25:37,327
So it just shows the power of both the immaculate conception of Bitcoin and all it represents, but also just its head start, network effects, etc.

2119
01:25:37,327 --> 01:25:37,807
Totally.

2120
01:25:37,807 --> 01:25:44,067
So like the agents creating their own cryptocurrency would face all the same challenges at creating as any of the other altcoins.

2121
01:25:44,207 --> 01:25:47,267
So all I would say to that is like, I agree.

2122
01:25:47,527 --> 01:25:48,527
And I want to be clear.

2123
01:25:48,627 --> 01:25:54,187
My like base case is like in the 95 plus percent Bitcoin wins for the agents too.

2124
01:25:54,247 --> 01:25:55,327
Because we already have the network effects.

2125
01:25:55,427 --> 01:25:56,587
It's already thermodynamically sound.

2126
01:25:56,667 --> 01:25:57,547
They can just like use it.

2127
01:25:57,667 --> 01:26:00,567
Once some use it, it actually increases those network effects.

2128
01:26:00,567 --> 01:26:01,867
but I want to be clear

2129
01:26:01,867 --> 01:26:03,367
like I don't think it's fair

2130
01:26:03,367 --> 01:26:04,127
to compare

2131
01:26:04,127 --> 01:26:05,387
the like

2132
01:26:05,387 --> 01:26:06,487
the feeble minds

2133
01:26:06,487 --> 01:26:07,287
of shit coiners

2134
01:26:07,287 --> 01:26:08,027
of the past

2135
01:26:08,027 --> 01:26:08,807
with like

2136
01:26:08,807 --> 01:26:09,527
super intelligence

2137
01:26:09,527 --> 01:26:10,887
and so what I mean by that is like

2138
01:26:10,887 --> 01:26:12,647
it's hard to project

2139
01:26:12,647 --> 01:26:13,747
I agree

2140
01:26:13,747 --> 01:26:14,847
the narrative I said

2141
01:26:14,847 --> 01:26:15,867
of like they just decide

2142
01:26:15,867 --> 01:26:16,647
oh the humans have it

2143
01:26:16,647 --> 01:26:17,367
let's make our own

2144
01:26:17,367 --> 01:26:17,647
like

2145
01:26:17,647 --> 01:26:18,687
yeah

2146
01:26:18,687 --> 01:26:19,827
that sounds bad

2147
01:26:19,827 --> 01:26:20,627
or like

2148
01:26:20,627 --> 01:26:21,607
not very convincing

2149
01:26:21,607 --> 01:26:22,347
but like

2150
01:26:22,347 --> 01:26:23,027
imagine

2151
01:26:23,027 --> 01:26:24,447
AI continues to get

2152
01:26:24,447 --> 01:26:25,627
radically smarter

2153
01:26:25,627 --> 01:26:26,427
over the next couple years

2154
01:26:26,427 --> 01:26:26,967
and it's like

2155
01:26:26,967 --> 01:26:27,907
it's like us

2156
01:26:27,907 --> 01:26:28,867
as humans

2157
01:26:28,867 --> 01:26:29,267
being like

2158
01:26:29,267 --> 01:26:30,087
listen those monkeys

2159
01:26:30,087 --> 01:26:30,507
over there

2160
01:26:30,567 --> 01:26:37,387
Now we want to interact with these monkeys and they've already got their currency of grapes.

2161
01:26:37,787 --> 01:26:43,027
And we could go and ask them to give us some of their grapes or we could just grow a trillion grapes and like, okay, come here, monkey.

2162
01:26:43,287 --> 01:26:43,987
Come out my grape.

2163
01:26:44,827 --> 01:26:45,907
That's not a great example.

2164
01:26:46,087 --> 01:26:52,967
But the point that I'm trying to make is imagine the AI being a thousand X smarter than any of us.

2165
01:26:52,967 --> 01:27:13,727
And they can do things like, as an example, if they were able to do an immaculate conception where they created something, you know, with the keys for the first, like whatever, like Bitcoin, they drop it on the internet. You know, they're all mining. There's hundreds of different models competing from day zero. It's very similar to Bitcoin.

2166
01:27:13,727 --> 01:27:15,647
but at that point

2167
01:27:15,647 --> 01:27:16,507
they also had the ability

2168
01:27:16,507 --> 01:27:17,427
to have robots

2169
01:27:17,427 --> 01:27:18,707
who are actually going out there

2170
01:27:18,707 --> 01:27:19,727
building the data centers

2171
01:27:19,727 --> 01:27:20,967
laying the solar panels

2172
01:27:20,967 --> 01:27:21,727
they're just like

2173
01:27:21,727 --> 01:27:22,827
this is our energy currency

2174
01:27:22,827 --> 01:27:23,787
and like oh by the way

2175
01:27:23,787 --> 01:27:24,687
who controlled

2176
01:27:24,687 --> 01:27:25,407
like the first

2177
01:27:25,407 --> 01:27:26,407
whatever blocks are mine

2178
01:27:26,407 --> 01:27:27,507
well we split the key

2179
01:27:27,507 --> 01:27:28,847
among a thousand different agents

2180
01:27:28,847 --> 01:27:29,807
using a bunch of different models

2181
01:27:29,807 --> 01:27:31,567
I'm just saying

2182
01:27:31,567 --> 01:27:32,927
like this is the only

2183
01:27:32,927 --> 01:27:35,867
potential threat

2184
01:27:35,867 --> 01:27:36,947
that I've ever been nervous about

2185
01:27:36,947 --> 01:27:37,287
with Bitcoin

2186
01:27:37,287 --> 01:27:38,187
and the fact they're already

2187
01:27:38,187 --> 01:27:38,987
talking about

2188
01:27:38,987 --> 01:27:40,127
their own energy currencies

2189
01:27:40,127 --> 01:27:40,727
maybe a little

2190
01:27:40,727 --> 01:27:41,727
we have talked about

2191
01:27:41,727 --> 01:27:42,767
this on the show

2192
01:27:42,767 --> 01:27:43,287
because then

2193
01:27:43,287 --> 01:27:45,667
because we talked about,

2194
01:27:45,787 --> 01:27:46,967
okay, let's say that happens

2195
01:27:46,967 --> 01:27:50,047
and again, it gets back to

2196
01:27:50,047 --> 01:27:50,847
you have to describe

2197
01:27:50,847 --> 01:27:52,807
what's the distribution model

2198
01:27:52,807 --> 01:27:54,827
or otherwise we just speculate.

2199
01:27:55,187 --> 01:27:57,007
So let's say they have a model

2200
01:27:57,007 --> 01:27:58,407
and the way you're painting the picture

2201
01:27:58,407 --> 01:28:01,247
is they invent a new money,

2202
01:28:01,727 --> 01:28:02,807
they hoard it to themselves.

2203
01:28:03,067 --> 01:28:04,127
The humans don't get any.

2204
01:28:04,667 --> 01:28:06,087
Well, if literally there's no way

2205
01:28:06,087 --> 01:28:06,907
for humans to get it,

2206
01:28:06,967 --> 01:28:09,627
then it's two different islands.

2207
01:28:09,627 --> 01:28:12,627
So the humans still need Bitcoin.

2208
01:28:12,627 --> 01:28:14,707
on an island.

2209
01:28:15,847 --> 01:28:16,787
We sell our grapes, whatever.

2210
01:28:17,107 --> 01:28:19,527
And if there is a way for humans

2211
01:28:19,527 --> 01:28:20,827
to get this new money by agents

2212
01:28:20,827 --> 01:28:22,227
because they actually want to open up

2213
01:28:22,227 --> 01:28:23,547
an economy between the two,

2214
01:28:24,067 --> 01:28:25,427
well, how do we acquire it?

2215
01:28:25,447 --> 01:28:27,087
Well, we either work or we invest.

2216
01:28:27,227 --> 01:28:29,007
And those who have more Bitcoin

2217
01:28:29,007 --> 01:28:30,447
or more real estate

2218
01:28:30,447 --> 01:28:32,127
or more Google stock or whatever

2219
01:28:32,127 --> 01:28:33,347
are going to be able to buy

2220
01:28:33,347 --> 01:28:34,827
more of this new currency than others.

2221
01:28:34,987 --> 01:28:35,147
Totally.

2222
01:28:35,267 --> 01:28:37,147
So I don't see how it destroys

2223
01:28:37,147 --> 01:28:38,447
the value of Bitcoin.

2224
01:28:39,487 --> 01:28:40,567
I don't think it would destroy it,

2225
01:28:40,587 --> 01:28:41,467
but I think to be clear,

2226
01:28:41,467 --> 01:28:43,607
the model of the world, I'm thinking

2227
01:28:43,607 --> 01:28:45,567
our wealth that we

2228
01:28:45,567 --> 01:28:47,507
have generated as humanity is like

2229
01:28:47,507 --> 01:28:49,587
0.001% of what they

2230
01:28:49,587 --> 01:28:51,627
make. They're off building dice and spheres.

2231
01:28:52,327 --> 01:28:53,547
And yeah, we still have our

2232
01:28:53,547 --> 01:28:54,847
Bitcoin amongst us, but

2233
01:28:54,847 --> 01:28:57,647
we're basically in their

2234
01:28:57,647 --> 01:28:59,227
zoo. I don't think that's likely.

2235
01:28:59,547 --> 01:29:00,167
To be clear.

2236
01:29:00,467 --> 01:29:02,087
You give it 5%?

2237
01:29:02,927 --> 01:29:03,087
Huh?

2238
01:29:03,767 --> 01:29:07,227
It goes far beyond Bitcoin, though.

2239
01:29:07,787 --> 01:29:09,307
You're just painting a picture of

2240
01:29:09,307 --> 01:29:10,047
we are

2241
01:29:10,047 --> 01:29:12,407
like nothing.

2242
01:29:12,587 --> 01:29:13,407
Well, not nothing.

2243
01:29:14,567 --> 01:29:16,107
Well, the way you just said it is like

2244
01:29:16,107 --> 01:29:17,827
they're doing dice and fears, we're doing nothing.

2245
01:29:18,187 --> 01:29:18,687
I mean like...

2246
01:29:18,687 --> 01:29:20,507
Well, like monkeys and humans, for example.

2247
01:29:21,027 --> 01:29:22,867
The monkeys are having a good life, you know,

2248
01:29:22,927 --> 01:29:23,827
especially the bonobos.

2249
01:29:23,827 --> 01:29:25,487
It's a fine conversation to have,

2250
01:29:25,587 --> 01:29:26,407
but it's very doomer,

2251
01:29:26,607 --> 01:29:28,427
and it goes far beyond Bitcoin.

2252
01:29:29,067 --> 01:29:31,067
Like then all of a sudden humans have no purpose

2253
01:29:31,067 --> 01:29:32,767
and we might as well just become extinct

2254
01:29:32,767 --> 01:29:34,007
or we're just like little pets.

2255
01:29:34,187 --> 01:29:36,547
Well, and I want to be clear on two things.

2256
01:29:36,667 --> 01:29:38,167
One, this is not my likely outcome.

2257
01:29:38,167 --> 01:29:43,047
my likely outcome is that humanity is going to have a lot of unique advantages that we don't

2258
01:29:43,047 --> 01:29:47,387
know yet. And we're going to discover that over time. Like that, that is my, let's say over 95%

2259
01:29:47,387 --> 01:29:52,587
probability. Human durable skills. Human durable skills or human durable value. I don't even know

2260
01:29:52,587 --> 01:29:56,567
if it'll be a skill per se, but it might be intuition. It might be, I don't know what it is.

2261
01:29:56,607 --> 01:29:59,987
Like, and I, Kevin Kelly says many times, I tend to agree with him. I tend to be optimistic.

2262
01:30:00,187 --> 01:30:03,967
We're going to discover that. But I also think to be clear, like this other world that I'm painting

2263
01:30:03,967 --> 01:30:07,747
is not a zero. The fact they're already talking about their own energy money is something we

2264
01:30:07,747 --> 01:30:11,667
should be paying attention to. And I think there's probably things we can do during this

2265
01:30:11,667 --> 01:30:18,247
transition stage to tip the scale one way or the other. For example, the faster we get agents

2266
01:30:18,247 --> 01:30:23,147
adopting and denominating in Bitcoin today, the better, because the more the new agents come

2267
01:30:23,147 --> 01:30:29,167
online are going to want to get the Bitcoin from those agents. Is that the right question?

2268
01:30:29,507 --> 01:30:33,247
Oh, he's Gen Z now. Gen Alpha. Oh, just one? All right.

2269
01:30:33,247 --> 01:30:34,767
I can do too

2270
01:30:34,767 --> 01:30:36,287
exactly

2271
01:30:36,287 --> 01:30:39,067
are these your lobster claws

2272
01:30:39,067 --> 01:30:39,807
for your opa claws

2273
01:30:39,807 --> 01:30:40,547
I guess so

2274
01:30:40,547 --> 01:30:41,587
yeah

2275
01:30:41,587 --> 01:30:42,127
anyways

2276
01:30:42,127 --> 01:30:43,587
I don't mean to be Doomer

2277
01:30:43,587 --> 01:30:44,687
and like I am

2278
01:30:44,687 --> 01:30:45,627
I only know this

2279
01:30:45,627 --> 01:30:46,887
you guys know what this is

2280
01:30:46,887 --> 01:30:47,687
6-7

2281
01:30:47,687 --> 01:30:49,787
yeah 6-7

2282
01:30:49,787 --> 01:30:50,847
I don't know

2283
01:30:50,847 --> 01:30:51,587
you don't know that

2284
01:30:51,587 --> 01:30:51,847
no

2285
01:30:51,847 --> 01:30:54,747
you'll have to show me after

2286
01:30:54,747 --> 01:30:56,247
you'll have to learn me some things

2287
01:30:56,247 --> 01:30:58,047
it's the height of Lamella Ball

2288
01:30:58,047 --> 01:30:58,527
is one

2289
01:30:58,527 --> 01:30:59,187
oh

2290
01:30:59,187 --> 01:31:01,507
and it's in a rap song

2291
01:31:01,507 --> 01:31:02,307
is that guy still playing

2292
01:31:02,307 --> 01:31:03,287
They became popular on TikTok.

2293
01:31:03,627 --> 01:31:09,767
And then the children, the youth of the world got into 6-7.

2294
01:31:09,867 --> 01:31:12,007
I thought you were going to say you were being boomer,

2295
01:31:12,067 --> 01:31:13,207
but actually I'm being boomer.

2296
01:31:14,487 --> 01:31:15,267
Roll my version.

2297
01:31:16,487 --> 01:31:19,727
Anyways, all I'm saying is the next couple of years,

2298
01:31:19,727 --> 01:31:22,847
I actually think what we do and how quickly adoption happens

2299
01:31:22,847 --> 01:31:23,787
actually matters.

2300
01:31:25,767 --> 01:31:26,647
Yeah, 100% agree.

2301
01:31:26,947 --> 01:31:29,407
In fact, I think a few weeks ago, I said that,

2302
01:31:29,687 --> 01:31:31,987
or I elaborated on that.

2303
01:31:31,987 --> 01:31:34,847
I mean, you're like, we have still plenty of time.

2304
01:31:34,967 --> 01:31:37,107
Yeah, well, 2026 needs to be the year

2305
01:31:37,107 --> 01:31:38,787
that Bitcoin gets adopted for a genetic payment.

2306
01:31:38,967 --> 01:31:39,887
I increasingly agree with you.

2307
01:31:39,907 --> 01:31:41,287
I'm glad you're on board.

2308
01:31:41,467 --> 01:31:42,047
Totally agree.

2309
01:31:42,307 --> 01:31:43,427
Yeah, the scales need to be tipped.

2310
01:31:43,647 --> 01:31:43,807
Yep.

2311
01:31:44,207 --> 01:31:45,827
And I think we do need to wrap soon,

2312
01:31:45,907 --> 01:31:47,327
but do you want to talk a little bit

2313
01:31:47,327 --> 01:31:50,227
about the X-Pro2, L-Pro2 directory stuff?

2314
01:31:51,707 --> 01:31:52,407
You're the rapper.

2315
01:31:53,127 --> 01:31:53,487
Exactly.

2316
01:31:53,587 --> 01:31:54,207
Give us a beat, DK.

2317
01:31:55,287 --> 01:31:55,847
Come on.

2318
01:31:57,387 --> 01:32:00,247
I think this remote setup with a little delay here,

2319
01:32:00,327 --> 01:32:01,707
it's just not going to, it's not going to.

2320
01:32:01,707 --> 01:32:02,447
next time.

2321
01:32:02,947 --> 01:32:03,487
L402.

2322
01:32:03,547 --> 01:32:03,947
When we're in,

2323
01:32:03,947 --> 01:32:04,807
when we're in,

2324
01:32:04,807 --> 01:32:05,187
uh,

2325
01:32:05,367 --> 01:32:06,207
in studio together.

2326
01:32:08,467 --> 01:32:08,867
Um,

2327
01:32:09,127 --> 01:32:10,727
well,

2328
01:32:10,787 --> 01:32:11,127
I don't know.

2329
01:32:11,187 --> 01:32:11,447
I mean,

2330
01:32:11,587 --> 01:32:11,827
we're,

2331
01:32:11,887 --> 01:32:12,347
we're over.

2332
01:32:12,687 --> 01:32:14,467
So what do you want to say about L402?

2333
01:32:14,827 --> 01:32:15,187
Well,

2334
01:32:15,227 --> 01:32:15,927
you were talking about,

2335
01:32:16,047 --> 01:32:16,267
you know,

2336
01:32:16,267 --> 01:32:18,047
we got to help the agents

2337
01:32:18,047 --> 01:32:19,307
figure it out early.

2338
01:32:19,427 --> 01:32:19,887
So I was just,

2339
01:32:20,247 --> 01:32:22,327
I was thinking we could talk a little bit

2340
01:32:22,327 --> 01:32:23,207
about just what we've seen.

2341
01:32:23,207 --> 01:32:23,707
Someone posted,

2342
01:32:23,847 --> 01:32:25,467
posted one and I don't know,

2343
01:32:25,527 --> 01:32:26,547
someone said they didn't,

2344
01:32:27,867 --> 01:32:29,147
that it could be better.

2345
01:32:29,267 --> 01:32:29,627
I don't know.

2346
01:32:29,627 --> 01:32:30,127
I mean,

2347
01:32:30,387 --> 01:32:31,067
I posted it.

2348
01:32:31,067 --> 01:32:33,267
I also think it can be better.

2349
01:32:34,207 --> 01:32:35,147
I didn't want to call you out.

2350
01:32:35,147 --> 01:32:36,807
I just didn't want to wrap the Friday show

2351
01:32:36,807 --> 01:32:39,487
on the Max Doomer side of things.

2352
01:32:39,707 --> 01:32:40,887
I'm not even being a Doomer, bro.

2353
01:32:40,947 --> 01:32:41,607
I'm just saying,

2354
01:32:41,807 --> 01:32:43,147
let's go into this shit with eyes open.

2355
01:32:43,507 --> 01:32:45,887
But yes, I'll end with something positive on that.

2356
01:32:46,247 --> 01:32:48,027
Something positive on the weekend, yeah.

2357
01:32:48,187 --> 01:32:48,587
Kick us off.

2358
01:32:50,667 --> 01:32:51,667
All right, check one, two.

2359
01:32:51,967 --> 01:32:53,447
Kazakhstan buying Bitcoin.

2360
01:32:53,447 --> 01:32:54,387
Nothing's life is free to see.

2361
01:32:54,427 --> 01:32:55,067
That's a positive.

2362
01:32:55,747 --> 01:32:57,927
We were supposed to talk about Kazakhstan.

2363
01:32:58,747 --> 01:33:00,967
Okay, so let's end on two positive notes.

2364
01:33:01,067 --> 01:33:02,847
Do you guys get a double whammy this weekend?

2365
01:33:02,987 --> 01:33:04,427
So the first positive note is

2366
01:33:04,427 --> 01:33:08,447
there's a lot of people doing these like X402

2367
01:33:08,447 --> 01:33:11,627
kind of scammy, whatever, crypto crap directories.

2368
01:33:11,747 --> 01:33:13,407
But like I saw on Stacker News,

2369
01:33:13,507 --> 01:33:15,047
an anonymous poster posted

2370
01:33:15,047 --> 01:33:18,747
one of the first machine readable L402 directories I've seen.

2371
01:33:18,867 --> 01:33:19,647
I thought that was awesome.

2372
01:33:20,667 --> 01:33:23,327
And by the way, I love this kind of like user interface

2373
01:33:23,327 --> 01:33:25,667
that Nick has built with MDK on like unhuman.domains

2374
01:33:25,667 --> 01:33:26,747
where there's like a human version,

2375
01:33:26,847 --> 01:33:27,867
which is pretty for us.

2376
01:33:28,007 --> 01:33:29,347
And then there's the machine version,

2377
01:33:29,487 --> 01:33:31,047
which is just like the raw text files.

2378
01:33:31,067 --> 01:33:32,387
that the AI is need to read it.

2379
01:33:33,127 --> 01:33:33,807
So I think someone,

2380
01:33:33,907 --> 01:33:36,067
there's a lot of really interesting exploration space

2381
01:33:36,067 --> 01:33:37,227
and I don't want to dox him

2382
01:33:37,227 --> 01:33:37,947
because I don't know if he's,

2383
01:33:38,047 --> 01:33:39,687
I think he's being very private with this project,

2384
01:33:39,847 --> 01:33:42,707
but a builder I like a lot, respect a lot,

2385
01:33:43,027 --> 01:33:44,347
is also kind of playing with showing

2386
01:33:44,347 --> 01:33:46,987
both L402 and X402 services together,

2387
01:33:47,167 --> 01:33:48,567
which is interesting because now maybe,

2388
01:33:48,567 --> 01:33:50,827
you know, it shows some of the advantages of things

2389
01:33:50,827 --> 01:33:53,267
L402 might be able to do that X402 can't do.

2390
01:33:53,827 --> 01:33:55,387
But yeah, if I were a builder,

2391
01:33:55,767 --> 01:33:57,027
and I guess now that I am,

2392
01:33:57,147 --> 01:33:59,547
I will continue to be experimenting with

2393
01:33:59,547 --> 01:34:03,247
how do you sort of like aggregate all these L402 services

2394
01:34:03,247 --> 01:34:06,347
in a way that it's super easy to discover and rank

2395
01:34:06,347 --> 01:34:09,147
and basically build trust and reputation for the bots.

2396
01:34:09,147 --> 01:34:10,687
And maybe Noster's going to play a role at this.

2397
01:34:10,756 --> 01:34:19,396
That was also something she came out of our hackathon discussion yesterday was, you know, her interest in like basically using the Bitcoin to rank things.

2398
01:34:19,496 --> 01:34:22,336
I thought that was interesting because that's non-forgeable.

2399
01:34:22,436 --> 01:34:24,996
I think that's something that resonates maybe more with tech people as well.

2400
01:34:25,516 --> 01:34:28,556
But that's something I want to hack with and find out more.

2401
01:34:28,656 --> 01:34:32,456
By the way, our producer literally has the cursor over end stream button.

2402
01:34:32,556 --> 01:34:36,236
So I think that's like our countdown timer to finish this out soon.

2403
01:34:36,736 --> 01:34:36,916
Okay.

2404
01:34:37,716 --> 01:34:38,776
Take us home with Kazakhstan.

2405
01:34:38,776 --> 01:34:38,976
Dan.

2406
01:34:39,816 --> 01:34:39,996
Oh,

2407
01:34:40,236 --> 01:34:40,636
um,

2408
01:34:40,796 --> 01:34:42,196
I just barely,

2409
01:34:42,296 --> 01:34:42,496
I mean,

2410
01:34:42,576 --> 01:34:44,856
I guess the country,

2411
01:34:45,116 --> 01:34:51,336
the powers that be at the country announced they approved digital assets in

2412
01:34:51,336 --> 01:34:52,356
the treasury or something like that.

2413
01:34:52,376 --> 01:34:53,636
And it might include Bitcoin,

2414
01:34:54,456 --> 01:34:55,016
but anyway,

2415
01:34:55,016 --> 01:34:55,756
the main point,

2416
01:34:56,096 --> 01:34:56,456
50,

2417
01:34:56,596 --> 01:34:59,576
$350 million million,

2418
01:34:59,776 --> 01:35:00,116
probably.

2419
01:35:00,276 --> 01:35:00,656
I don't know.

2420
01:35:00,816 --> 01:35:00,956
But,

2421
01:35:01,056 --> 01:35:01,256
um,

2422
01:35:01,596 --> 01:35:02,356
I guess I,

2423
01:35:02,436 --> 01:35:03,976
I don't find that too surprising.

2424
01:35:03,976 --> 01:35:04,436
I mean,

2425
01:35:04,476 --> 01:35:04,776
each,

2426
01:35:04,916 --> 01:35:07,456
each country that does this is a win.

2427
01:35:07,776 --> 01:35:08,136
Um,

2428
01:35:08,136 --> 01:35:11,176
the angle I was going to take

2429
01:35:11,176 --> 01:35:12,856
to talk about it is that

2430
01:35:12,856 --> 01:35:15,356
I think an area

2431
01:35:15,356 --> 01:35:16,916
of improvement within the Bitcoin ecosystem

2432
01:35:16,916 --> 01:35:19,216
is how can a sovereign

2433
01:35:19,216 --> 01:35:21,016
nation like that self-custody their Bitcoin

2434
01:35:21,016 --> 01:35:23,296
because I believe like El Salvador

2435
01:35:23,296 --> 01:35:25,136
they just rely on

2436
01:35:25,136 --> 01:35:26,076
American companies

2437
01:35:26,076 --> 01:35:28,496
to custody their Bitcoin

2438
01:35:28,496 --> 01:35:31,076
well that doesn't work out

2439
01:35:31,076 --> 01:35:31,456
very well

2440
01:35:31,456 --> 01:35:35,276
it doesn't work out very well if America becomes your enemy

2441
01:35:35,276 --> 01:35:35,956
what could go wrong

2442
01:35:35,956 --> 01:35:38,936
And if you monitor geopolitics recently,

2443
01:35:39,216 --> 01:35:42,836
who knows who's going to be friendly?

2444
01:35:43,336 --> 01:35:44,676
Well, yeah, look at all of history.

2445
01:35:45,076 --> 01:35:46,496
And certainly just look at the past year.

2446
01:35:46,636 --> 01:35:47,676
Just look at the last two months.

2447
01:35:47,796 --> 01:35:49,036
I think three of them.

2448
01:35:49,436 --> 01:35:50,036
Two months.

2449
01:35:50,056 --> 01:35:55,456
And so maybe if you're as large as China or the United States,

2450
01:35:55,696 --> 01:35:58,196
there's lots of companies within your country

2451
01:35:58,196 --> 01:36:00,116
that you can rely on to custody it.

2452
01:36:00,116 --> 01:36:02,316
And you can police them and enforce it

2453
01:36:02,316 --> 01:36:03,536
and take it over if you need it or whatever.

2454
01:36:04,136 --> 01:36:15,378
But if you a smaller country like Kazakhstan I mean how many companies in Kazakhstan are like qualified custodians or whatever definition of that but like a high quality custodian

2455
01:36:15,498 --> 01:36:18,878
I'm guessing zero or most countries is zero.

2456
01:36:19,878 --> 01:36:23,718
So, well, what's one of the amazing features of Bitcoin?

2457
01:36:23,878 --> 01:36:24,838
It's self-custody, right?

2458
01:36:24,838 --> 01:36:30,178
But it goes beyond just like having a Trezor or a BitKey

2459
01:36:30,178 --> 01:36:33,378
and like putting your central bank of Bitcoin in it

2460
01:36:33,378 --> 01:36:35,578
And then like, oh shit, what was my C-Pres again?

2461
01:36:35,578 --> 01:36:37,858
Wearing the Ledger necklace or whatever.

2462
01:36:38,138 --> 01:36:41,078
I don't think that's like the right procedure and protocol.

2463
01:36:42,738 --> 01:36:44,178
I made it for the technology, bro.

2464
01:36:45,378 --> 01:36:50,458
So I still feel like there's even honestly an open source project.

2465
01:36:51,678 --> 01:36:56,878
Open source project for this where it just outlines a process and procedure

2466
01:36:56,878 --> 01:37:01,458
that protects the secure, like who are the key holders?

2467
01:37:01,458 --> 01:37:04,798
what is the multi-sig or layers of multi-sig

2468
01:37:04,798 --> 01:37:05,838
that should even be used?

2469
01:37:05,898 --> 01:37:08,358
Because obviously the more the keys,

2470
01:37:08,498 --> 01:37:10,538
the more the multi-sig, the more the layers,

2471
01:37:10,658 --> 01:37:11,518
the more complex it is.

2472
01:37:11,598 --> 01:37:13,018
So what's the right trade-off there?

2473
01:37:13,678 --> 01:37:14,718
Who's holding the keys?

2474
01:37:14,818 --> 01:37:16,598
How do you protect their physical security?

2475
01:37:17,078 --> 01:37:20,118
How do you protect them from rugging the country's Bitcoin?

2476
01:37:20,598 --> 01:37:23,918
There's all these operational questions

2477
01:37:23,918 --> 01:37:25,218
and security questions.

2478
01:37:26,138 --> 01:37:30,398
So having people learn from those who have already done it

2479
01:37:30,398 --> 01:37:33,798
and codify that or just pontificate how it should work

2480
01:37:33,798 --> 01:37:36,638
and publishing that and iterating on it,

2481
01:37:36,658 --> 01:37:37,458
I think is very important.

2482
01:37:37,618 --> 01:37:39,998
I like this idea of having that be an open source

2483
01:37:39,998 --> 01:37:40,698
set of protocols.

2484
01:37:40,858 --> 01:37:43,298
And then you could probably have startups around that as well

2485
01:37:43,298 --> 01:37:44,298
that are like consulting.

2486
01:37:45,178 --> 01:37:47,238
A country says, I want to do this,

2487
01:37:47,258 --> 01:37:49,038
and maybe you don't just do one because you don't trust them.

2488
01:37:49,178 --> 01:37:51,498
So maybe it's a European company, a US company,

2489
01:37:51,678 --> 01:37:53,278
a Singaporean company, and they all come.

2490
01:37:53,558 --> 01:37:55,198
And then maybe it gets to stuff like multiseg

2491
01:37:55,198 --> 01:37:57,658
and each of them is a trusted custodian jurisdiction.

2492
01:37:57,978 --> 01:37:59,018
And yeah, it's a great idea.

2493
01:37:59,018 --> 01:38:01,358
on that note

2494
01:38:01,358 --> 01:38:02,638
that's a good wrap

2495
01:38:02,638 --> 01:38:04,858
we will see you next week

2496
01:38:04,858 --> 01:38:06,058
thanks everyone