CD197: MATT AHLBORG - PPQ.AI - AI AGENTS, PRIVACY, AND PAYMENTS

WEBVTT

NOTE
Transcription provided by Podhome.fm
Created: 03/25/2026 18:17:44
Duration: 4627.670
Channels: 1

1
00:00:32.280 --> 00:00:33.800
Happy Bitcoin

2
00:00:33.800 --> 00:00:38.440
Wednesday, freaks. It's your host, Odell, here for another Citadel Dispatch.

3
00:00:38.440 --> 00:00:42.360
The show focused on actual Bitcoin and Freedom Tech

4
00:00:42.360 --> 00:00:43.080
discussion.

5
00:00:43.535 --> 00:00:45.375
It is currently Wednesday,

6
00:00:45.375 --> 00:00:46.735
March 25,

7
00:00:47.055 --> 00:00:49.455
sixteen hundred UTC.

8
00:00:49.615 --> 00:00:52.735
You guys will probably be listening to this in a few hours.

9
00:00:53.135 --> 00:00:56.655
The current block height is nine four two one seven four.

10
00:00:57.370 --> 00:01:00.250
Current stats per dollar is one four one two.

11
00:01:01.530 --> 00:01:09.130
The price for 1 Bitcoin is now $70,795

12
00:01:09.770 --> 00:01:11.530
Bitcoin in gold terms.

13
00:01:12.355 --> 00:01:15.075
1 Bitcoin can now buy you 15.83

14
00:01:15.075 --> 00:01:17.635
ounces, but you might regret it if you do it.

15
00:01:18.675 --> 00:01:21.235
We're up on the week in gold 7%.

16
00:01:21.235 --> 00:01:23.795
We're up on the month 20.7%,

17
00:01:24.400 --> 00:01:27.040
and we're still down on the year 45%,

18
00:01:27.040 --> 00:01:28.880
but I expect that to turn around.

19
00:01:29.360 --> 00:01:31.439
Anyway, Freeze as always dispatch.

20
00:01:31.840 --> 00:01:37.200
It's purely funded by our audience, viewers like you. Thank you for sending your hard earned stats

21
00:01:37.439 --> 00:01:38.479
to support the show.

22
00:01:39.235 --> 00:01:40.674
It truly means a lot.

23
00:01:41.315 --> 00:01:47.314
The largest we have no ads or sponsors. You guys are the pure value for value experience.

24
00:01:47.395 --> 00:01:48.915
Largest Zap from

25
00:01:49.475 --> 00:01:54.990
last week was kind of light on Zaps from our show last week. I guess it was a Friday show.

26
00:01:55.469 --> 00:02:00.430
Maybe some of you guys haven't listened to it yet. If you haven't, make sure you go listen to it. Largest

27
00:02:00.829 --> 00:02:09.295
largest Zap was 10,000 sets from Ride or Die Freak, Mab 21. He said great rip. That rip, if you guys haven't listened to it yet is

28
00:02:09.455 --> 00:02:10.815
with Evgeny,

29
00:02:11.694 --> 00:02:14.735
who is the founder and maintainer of simplex.

30
00:02:14.735 --> 00:02:16.255
It was a conversation

31
00:02:16.815 --> 00:02:18.735
that I particularly enjoyed

32
00:02:18.815 --> 00:02:21.694
a lot. I mean, it's kind of like a celebrity to me,

33
00:02:22.860 --> 00:02:30.700
but it was a lot. It was a lot of fun. And I think you guys should listen to it if you haven't. But anyway, freaks, we have a great show lined up

34
00:02:31.260 --> 00:02:32.140
today.

35
00:02:32.379 --> 00:02:36.140
I'm trying to keep the ball rolling with high signal, good quality conversation.

36
00:02:36.615 --> 00:02:49.175
If I can't find something high signal, we just skip the week because I don't have ads or sponsors that require me to do three shows a week like all the other shows. But before we get started, always, all the relevant links are sealdispatch.com.

37
00:02:49.175 --> 00:02:52.055
Share with your friends and family available in every podcast app

38
00:02:52.940 --> 00:02:55.260
by just searching pseudo dispatch.

39
00:02:55.340 --> 00:02:58.780
We have a return guest today. We have Matt Alborg

40
00:02:58.780 --> 00:03:00.700
of ppq.ai.

41
00:03:00.700 --> 00:03:04.380
Yes. It's another AI focused show. How's it going, Matt?

42
00:03:04.860 --> 00:03:07.875
Yeah. It's good. Good. Very good, Matt.

43
00:03:08.195 --> 00:03:09.475
Thanks for having me back. The

44
00:03:10.115 --> 00:03:11.715
mats are in the house. I,

45
00:03:12.915 --> 00:03:14.995
we last had Matt on the show,

46
00:03:15.395 --> 00:03:17.235
07/14/2025

47
00:03:17.235 --> 00:03:18.835
episode one sixty eight.

48
00:03:19.920 --> 00:03:24.000
I was just starting to dip my toes into AI stuff.

49
00:03:25.519 --> 00:03:31.120
We've come a long way. It felt it felt like it was time. It was time for

50
00:03:31.535 --> 00:03:34.175
a return appearance and update conversation.

51
00:03:34.495 --> 00:03:36.015
What a year it's been, Yeah.

52
00:03:36.894 --> 00:03:39.855
Yeah. I feel about ten years older. So yeah,

53
00:03:39.935 --> 00:03:41.935
it's a lot is happening.

54
00:03:42.974 --> 00:03:46.575
I mean, you have been in Bitcoin for a while as well.

55
00:03:47.480 --> 00:03:50.920
Mean, to me, it feels very much like early Bitcoin, right?

56
00:03:52.120 --> 00:03:54.440
Yeah, it does. It feels very Wild West.

57
00:03:55.080 --> 00:03:57.720
The rules are being made as we go.

58
00:03:59.080 --> 00:04:04.235
And yeah, there's a lot of creativity out there on on what you can do what you can build,

59
00:04:04.395 --> 00:04:06.075
how you can use this stuff.

60
00:04:07.515 --> 00:04:11.035
Yeah, there's no template really for what's happening right now.

61
00:04:12.155 --> 00:04:13.995
And yeah, like early both

62
00:04:15.070 --> 00:04:16.590
big and fun.

63
00:04:17.070 --> 00:04:17.470
Yeah.

64
00:04:18.030 --> 00:04:18.510
Sorry.

65
00:04:19.790 --> 00:04:22.990
You're good. Go ahead, man. No, there's no veterans,

66
00:04:22.990 --> 00:04:28.430
right? There's no gray beards. There's no people who have been doing this for thirty years because the tech just hasn't existed.

67
00:04:29.035 --> 00:04:37.675
I feel like my mind is racing. Like you said, I feel like a year is a decade. That was to me, those are like the templates of my early Bitcoin experience.

68
00:04:38.155 --> 00:04:48.110
Yeah. So what I was gonna say, and I yeah, it's relevant here is just on the call with with the guys today, the teammates, we're each going over our development workflow.

69
00:04:48.270 --> 00:05:08.305
And, you know, in AI, there's, you know, you can find guides on how to, like, upscale your development workflow and such, but we're all just kind of inventing it as we go. You know, I'm using GitHub work trees and Versus code. And one of my teammates is using team team ox and he's generating like a femoral work trees. And then we

70
00:05:08.945 --> 00:05:22.510
all have different clot MD files, and we're using different testing agents and stuff like that. So there's really Yeah, it's it's we're all creating our own thing right now. And just trying to like, become as efficient as possible.

71
00:05:23.070 --> 00:05:23.630
Wild.

72
00:05:24.270 --> 00:05:25.389
How big is your team now?

73
00:05:26.725 --> 00:05:31.845
Three. We're actually I don't know what we were last time, but we were at five at one point.

74
00:05:32.245 --> 00:05:32.805
And

75
00:05:33.525 --> 00:05:37.845
two of the guys just weren't working out, and so we're down to a lean and mean three.

76
00:05:38.885 --> 00:05:44.200
But I actually feel like we're getting a lot more done now. We have a really good rhythm.

77
00:05:44.520 --> 00:05:45.160
And,

78
00:05:45.720 --> 00:05:46.440
kind of

79
00:05:47.240 --> 00:05:53.960
as we onboard people in the future, I just I really wanna, you know, make sure that it's the right chemistry and the right

80
00:05:54.360 --> 00:05:56.440
right mentality and all that stuff. So

81
00:05:57.985 --> 00:06:00.625
yeah, I mean, hiring is hard. I

82
00:06:01.824 --> 00:06:02.625
people,

83
00:06:02.625 --> 00:06:03.985
it's hard to, know,

84
00:06:04.785 --> 00:06:08.145
every people problems are hard problems. And then,

85
00:06:08.544 --> 00:06:13.520
I mean, I do think I mean, maybe we're, we seem to be entering the era of

86
00:06:13.920 --> 00:06:17.360
the lean team, the lean small team that gets shit done.

87
00:06:18.560 --> 00:06:20.720
We already kind of were with Bitcoin,

88
00:06:21.040 --> 00:06:21.760
I think,

89
00:06:22.320 --> 00:06:24.240
you know, particularly founders

90
00:06:24.240 --> 00:06:44.389
that I see this on the $10.31 side all the time. Founders that think in Bitcoin as opportunity costs, think very carefully about expanding their payroll, because it's less Bitcoin that they can stack and they're thinking long term. And then you couple that all of a sudden with these AI tools, and three people can do the work that maybe 30 or 50 people were doing previously,

91
00:06:44.389 --> 00:06:46.550
it's kind of a wild compounding effect.

92
00:06:47.510 --> 00:07:03.905
Yeah, I absolutely am a big believer in that, that maybe not three, I would like to get up to maybe 10 or so, but I think a team of 10 can do really, really crazy things now. And yeah, as you said, one person is now, I don't know if it's 30 or 50, but it's

93
00:07:03.985 --> 00:07:10.950
a lot. And yeah, you just have to know how to use AI. That's that's really the big thing. And that was kind of a decision maker.

94
00:07:11.190 --> 00:07:17.110
And, you know, when we hired people on and such, it's, you have to really want to learn AI, you have to

95
00:07:17.590 --> 00:07:25.565
live and breathe it in order to really level up. I think that there's probably a lot of people out there who they're just kind of casually using AI

96
00:07:25.885 --> 00:07:32.045
and they're not treating it like they're not giving it the respect that it deserves in terms of how much it can

97
00:07:32.699 --> 00:07:34.620
really increase your productivity.

98
00:07:34.620 --> 00:07:38.380
And so, yeah, I was just thinking the other day about how,

99
00:07:39.259 --> 00:07:39.980
you know,

100
00:07:40.620 --> 00:07:46.300
I'm not I don't have a CS degree myself. I have a half of one, so I'm not totally nontechnical,

101
00:07:46.300 --> 00:07:58.225
but I did, you know, I went to have like half a degree, and then I went to a couple of the coding boot camps. And then, you know, I've just been kind of grinding on on stuff with AI since. And,

102
00:07:59.010 --> 00:08:03.170
yeah, my thought was, like, for your marketer, the marketers of the future,

103
00:08:03.490 --> 00:08:08.450
will they be the ones will they be the marketing native people who

104
00:08:08.690 --> 00:08:13.330
learn how to use AI? Or will it be the developers who learn the marketing side?

105
00:08:14.215 --> 00:08:23.015
Obviously, it's gonna be a little bit of both, but I do wonder, like, which trade is going to learn faster or have the the brain to, like,

106
00:08:23.175 --> 00:08:27.495
master the enough of the other trade to become, like, really potent. Right?

107
00:08:28.370 --> 00:08:31.810
Because yeah, I would love when I did hire on

108
00:08:31.970 --> 00:08:33.890
kind of a community manager person.

109
00:08:34.850 --> 00:08:37.410
There was a lot of blockers happening because

110
00:08:37.570 --> 00:08:44.095
he was not able to spin up his own like analytics database or his own analytics dashboards. And so

111
00:08:44.335 --> 00:08:48.415
it was always waiting on me, like, to create these dashboards and stuff.

112
00:08:48.735 --> 00:09:10.760
I've got a lot on my plate anyways. So it would have been amazing if, like, you know, you bring in a community person, a marketing person who can also, like, really vibe on Cloud Code or something like that. And then they're just doing all of the technical work themselves. That's what I really think is like, really important. If you're like trying to trying to get a job these days, you have to be really good at AI.

113
00:09:12.515 --> 00:09:13.555
Yeah.

114
00:09:13.555 --> 00:09:16.835
I mean, it's a weird thing to think about. I mean, at the

115
00:09:18.115 --> 00:09:26.035
definitely high level, like on the surface, like what what you see a lot of people is like, they're using these things as maybe like an advanced Google search.

116
00:09:26.600 --> 00:09:31.160
And that's just barely scratching the surface of what it's capable of doing.

117
00:09:31.880 --> 00:09:34.440
Yeah. To your point. But on the topic

118
00:09:34.680 --> 00:09:38.600
of what type of person does it benefit the most? I mean, it definitely depends on

119
00:09:39.605 --> 00:09:40.245
what

120
00:09:40.884 --> 00:09:48.805
role is trying to be filled or what qualities are needed in a person. But I'm not I don't have necessarily a hard answer there. I think

121
00:09:49.125 --> 00:09:51.764
obviously trained developers have

122
00:09:52.160 --> 00:09:58.320
an advantage right now just because you need to have if you have some level of code competency,

123
00:09:58.399 --> 00:10:12.265
you can really get super powered on these tools. But I do kind of see it I mean, I, for instance, I have no coding background whatsoever. I can kind of read code. I know how to, like, get around to GitHub repo. I'm, like, a

124
00:10:13.225 --> 00:10:17.945
technically competent non dev. And, you know, like, my big claim to fame was, like, I

125
00:10:19.720 --> 00:10:28.840
I found an issue with I was looking through the code of green wallet, which is now Blockstream Wallet, I found an issue that a bunch of their geniuses missed. But I'm

126
00:10:28.840 --> 00:10:30.680
definitely not a dev, but

127
00:10:31.080 --> 00:10:38.665
these tools really superpower me because I'm technical enough to start to leverage them and see what's happening and get my feet wet.

128
00:10:39.065 --> 00:10:56.140
And I think that type of person benefits a lot. I think that top tier devs benefit a lot. And then I think maybe, and once again, this is all just speculation. We're so early, and that's one of the things that's so fascinating to me. I think those like mid level devs are those are the ones that get really commoditized in this era.

129
00:10:56.860 --> 00:11:00.700
I don't know if they have such a huge advantage over,

130
00:11:01.500 --> 00:11:02.860
you know, a technically

131
00:11:03.019 --> 00:11:04.300
competent non dev.

132
00:11:05.365 --> 00:11:09.285
That makes sense. I don't even know if I would like bisect it that way.

133
00:11:09.605 --> 00:11:16.965
I mean, I think it just all comes down to, do you have a willingness to learn? Do you have a willingness to change up your current

134
00:11:17.285 --> 00:11:23.060
ways? I think that some of the folks on the senior side are struggling a little bit because

135
00:11:23.220 --> 00:11:25.700
they were really good at something for a long time.

136
00:11:26.660 --> 00:11:43.775
They do see a lot of the mistakes that AI makes, but they, I guess, I think they wrap their ego a little bit in in their abilities, and they're a little bit off put by the fact that, like, an AI can ship now 10 times more than you can. And, it does make errors

137
00:11:44.175 --> 00:11:48.170
and such, which you have to correct, but, like, it's still shipping 10x.

138
00:11:48.170 --> 00:11:53.290
I think that it doesn't matter who you are, whether you're a junior, mid level, whatever.

139
00:11:54.329 --> 00:11:57.449
It's it's like, do you have the humbleness

140
00:11:57.449 --> 00:12:02.585
to, like, learn and respect this stuff and just keep learning? And

141
00:12:03.065 --> 00:12:09.145
and like I said, you have to forge your own path. There aren't a lot of guides out there right now to do this, do that.

142
00:12:09.464 --> 00:12:16.450
There are, but I personally just rather dig in and like kind of figure things out myself. And that's

143
00:12:16.530 --> 00:12:20.610
really what it comes down to. I also think that maybe developers

144
00:12:20.610 --> 00:12:25.735
who don't have a little bit of the business mindset also may struggle.

145
00:12:26.615 --> 00:12:29.735
You know, if they're perfecting something that really,

146
00:12:30.214 --> 00:12:38.459
you know, maybe they're building something beautiful, but you should always be asking the question, how much time am I spending on this thing and how is it going to impact

147
00:12:39.100 --> 00:12:44.140
bottom line? Or like, how does it compare to all these other things we're working on? So,

148
00:12:45.259 --> 00:12:51.285
I think you can't just be the technical guy living in the basement anymore either. You have

149
00:12:52.085 --> 00:12:54.885
to be a little more aware of the other facets as well.

150
00:12:56.165 --> 00:13:04.005
Yeah, think that's a good point in perspective. It'll be very interesting watching how it all evolves. I mean, just to jump into the meat here, one of the key

151
00:13:04.250 --> 00:13:09.450
from what I remember on our last conversation, I I think one of the key conversation key conversation

152
00:13:09.450 --> 00:13:10.730
points we were having

153
00:13:11.209 --> 00:13:11.930
is

154
00:13:12.410 --> 00:13:18.490
from my seat, there's two ways people are interacting with these tools. Two main ways people are interacting with these tools.

155
00:13:20.125 --> 00:13:21.084
One is

156
00:13:22.764 --> 00:13:25.964
the relatively controlled and proprietary

157
00:13:25.964 --> 00:13:28.764
subscription model. So people are subscribing

158
00:13:28.764 --> 00:13:29.404
to

159
00:13:29.644 --> 00:13:34.365
Anthropic or OpenAI or Google directly or Grok directly.

160
00:13:34.930 --> 00:13:36.850
Usually that involves KYC,

161
00:13:37.090 --> 00:13:39.090
at least some level of KYC,

162
00:13:39.090 --> 00:13:42.210
but the anthropic require, you know, Claude requires

163
00:13:42.850 --> 00:13:45.490
phone number verification, they're actively blocking.

164
00:13:46.290 --> 00:13:46.850
They're actively

165
00:13:47.394 --> 00:13:57.235
burner numbers, obviously, then they have you pay with credit cards. Then that is like another level of KYC to identify who's using it. They also came out and said, specifically

166
00:13:57.635 --> 00:14:08.460
that some of these open source Chinese models are training against them, and called out by name certain researchers on the on the Chinese side that were using their accounts with NIMs.

167
00:14:08.540 --> 00:14:18.845
Because of their prompt history and how they were doing it, they were able to reverse calculate who they were, regardless of the other KYC elements. Have a bunch of privacy issues there, but as a result,

168
00:14:19.405 --> 00:14:26.925
you're getting a significant amount of value. I mean, I think the most the latest report said that your $200

169
00:14:26.925 --> 00:14:29.959
Claude subscription is getting you about $5,000

170
00:14:29.959 --> 00:14:38.839
worth of credits a month right now. So Anthropic is just subsidizing it. They're the drug dealer giving you free hookups until you're completely

171
00:14:39.160 --> 00:14:45.425
hooked on their drug. And then presumably I expect them to want profitability at some point and they'll jack up prices.

172
00:14:45.585 --> 00:14:47.825
And then you have a PPQ style,

173
00:14:49.425 --> 00:15:07.710
pay per token model where the more you use these tools, the more you pay, they can get quite expensive, but you have more choice and user sovereignty. You're able to switch between models on the fly. TPQ now at this point, I don't even know how many models you offer. I mean, I guess that's a question for you. But it's shit ton Yeah. And of

174
00:15:09.470 --> 00:15:12.590
you can just switch on the fly. How do you think about

175
00:15:13.685 --> 00:15:32.800
that dynamic? Because it is way, way more expensive as I'm like learning with these things and I'm having them run, especially if I'm I mean, as one thing, one cool thing about PBQ is you can offload a lot of things to the cheaper models because you have model choice. But I mean, if you're hitting Opus or God forbid, you're hitting OpenAI,

176
00:15:32.800 --> 00:15:36.640
they charge so much on a per token basis.

177
00:15:37.200 --> 00:15:39.360
How are you thinking about that dynamic?

178
00:15:39.360 --> 00:15:45.345
Because obviously it's key to your business and your sustainability and also personally, how are you using these tools?

179
00:15:46.225 --> 00:15:47.105
Yeah,

180
00:15:47.105 --> 00:15:52.385
I think that running subscriptions, that's the very tried and true method that's been

181
00:15:52.625 --> 00:16:10.630
employed for a couple of decades now. The one main benefit is you get to charge users money even when they don't use your product. And so that creates like an extra padding for you, I guess. But what I find is that I do think that the subsidies are happening a lot right now. And in some cases, you

182
00:16:11.269 --> 00:16:12.470
if you're just

183
00:16:13.245 --> 00:16:19.485
using cloud code all day every day, it makes sense to just get a cloud subscription and not use PBQ.

184
00:16:20.525 --> 00:16:21.085
But

185
00:16:21.565 --> 00:16:26.685
I think there's a tremendous amount of cost in maintaining that subscription as the company.

186
00:16:27.990 --> 00:16:34.150
You see every other day now, they have changes in the rate limits. They're sneaky. They're sneakily changing

187
00:16:34.310 --> 00:16:40.390
the terms. They're they're doing all sorts of stuff like that. They probably have entire teams dedicated

188
00:16:40.550 --> 00:16:53.705
to squeezing as much as they can out of the non users or the the casuals that that barely spend anything. And then and then, like, punishing the power users as much as possible. And that's just something that, like, I find that really boring

189
00:16:54.460 --> 00:16:59.020
to have to manage as a business. And I also think it's like not,

190
00:16:59.820 --> 00:17:04.539
it's the way that things are already done. So I would rather try to build something that

191
00:17:04.780 --> 00:17:19.645
actually is innovating in this area. But where it's just a very simple billing system, we focus on building features and products. And other companies out there are focusing on like rate limiting their subscriptions and setting up abuse

192
00:17:19.645 --> 00:17:20.445
detection

193
00:17:21.005 --> 00:17:21.805
and all this and that.

194
00:17:22.549 --> 00:17:30.629
And I've seen it happen to several kind of competitor companies out there where they will come up with a subscription with a certain amount of like monthly tokens.

195
00:17:31.190 --> 00:17:33.029
And then once abuse starts happening,

196
00:17:33.990 --> 00:17:44.455
then they have to change the terms and they make their users angry and then they have to, you know, win their users back and this and that. And I would just rather build the hard product and

197
00:17:44.695 --> 00:17:46.135
let it win on its merits.

198
00:17:46.295 --> 00:17:49.175
Yeah. I mean, there's a lot more to say about the,

199
00:17:49.815 --> 00:17:53.000
you know, trying to doing vendor lock in and such.

200
00:17:54.120 --> 00:17:57.399
But these models are becoming some of them are great

201
00:17:57.640 --> 00:17:59.480
and distinguished

202
00:17:59.480 --> 00:18:09.475
from others. But I think over the last year, they've actually come more alike than they have become different. Like, you know, Claude was the clear coding leader

203
00:18:09.715 --> 00:18:24.649
over the last year, and now Codex is doing very well. And now even the open source ones are doing well. So, I think it's going to be hard to lock people in. And I think the subsidies are not going to last and we're already seeing that. So, just I like running a paper

204
00:18:24.890 --> 00:18:37.385
query business and just charging an honest margin. I'm competitive on the power models, the ones that everyone uses. And then where I'm gonna make my money, hopefully, is on kind of maybe the more exotic features

205
00:18:37.385 --> 00:18:38.424
of our website.

206
00:18:39.385 --> 00:18:40.664
Exotic APIs

207
00:18:41.225 --> 00:18:42.024
using

208
00:18:42.184 --> 00:18:49.249
if you so if you're using our API, it's very cheap. But if you use our UI, we have a bit of a higher margin because people generally don't

209
00:18:49.649 --> 00:18:57.330
spend a lot of money on on UIs. They're spending more money in the in the IDEs and through the API. So that's where we'll make our money.

210
00:18:58.435 --> 00:18:59.075
And

211
00:18:59.395 --> 00:19:01.155
I also think that there's

212
00:19:01.955 --> 00:19:03.155
a future where

213
00:19:04.675 --> 00:19:08.915
we'll be able to create really specialized tools, AI tools,

214
00:19:09.155 --> 00:19:11.155
that can solve very niche problems

215
00:19:11.320 --> 00:19:19.799
in a way that a generic model cannot solve those problems. And when you create those types of tools, you could also put a nice margin on top of those as well.

216
00:19:21.720 --> 00:19:23.560
Yeah. That makes a lot of sense to me. I mean,

217
00:19:24.845 --> 00:19:27.325
this was always the Fiat Silicon

218
00:19:27.325 --> 00:19:28.205
Valley

219
00:19:28.605 --> 00:19:34.924
VC model. Right? Which is growth at all costs, spend a shit ton of money, subsidize user growth.

220
00:19:35.165 --> 00:19:42.830
I just don't think we've ever seen it at this level. I mean, if if those numbers are correct and it's, you know, power users are getting $5,000

221
00:19:42.830 --> 00:19:43.789
worth of

222
00:19:44.910 --> 00:19:48.029
clawed tokens for their $200 max subscription,

223
00:19:48.429 --> 00:19:52.990
that means that they're spending for growth. They're currently spending $60,000.

224
00:19:52.990 --> 00:19:54.910
They're losing $60,000

225
00:19:54.645 --> 00:20:02.804
per power user, which is just, I mean, I know there's a lot of capital sloshing around over an AI land. But it's hard to imagine them lasting

226
00:20:03.205 --> 00:20:14.809
very long with that model. And like you said, I mean, we have been seeing people get locked in their whole life is in Claude or Gemini or something, and then the terms change or they get a vague ban hammer or something.

227
00:20:15.690 --> 00:20:19.290
And then they're going to have a bad time. And it's probably that probably accelerates and gets worse.

228
00:20:19.924 --> 00:20:22.884
On the note of, you know, model competitiveness,

229
00:20:22.885 --> 00:20:26.244
I will say the single biggest thing I think we've seen over the last year,

230
00:20:26.645 --> 00:20:37.539
which has been really cool to see is the lighter weight open models, the floor has been risen. So there's still, I feel like a significant gap between the latest Claude or the latest

231
00:20:37.860 --> 00:20:39.220
OpenAI model

232
00:20:39.380 --> 00:20:42.820
and what's available out there in open source land.

233
00:20:43.700 --> 00:20:43.940
But

234
00:20:44.895 --> 00:20:47.455
even if we saw no improvements

235
00:20:47.455 --> 00:20:50.095
to the open source lightweight models,

236
00:20:50.095 --> 00:20:52.655
which is ridiculous, of course, we're going to see improvements.

237
00:20:52.655 --> 00:20:54.735
I think they change everything. Mean,

238
00:20:55.135 --> 00:20:59.855
they're already incredibly competent. So it really just changes how society functions.

239
00:21:00.440 --> 00:21:08.200
I I they're, I think they're certainly much smarter than they used to be. But I find that as the tools get more complicated,

240
00:21:08.600 --> 00:21:11.160
they really start showing their stupidity.

241
00:21:11.160 --> 00:21:15.455
For example, when you plug these open source models into OpenClaw,

242
00:21:15.455 --> 00:21:17.934
I know this story way too well because

243
00:21:18.414 --> 00:21:25.535
OpenClaw actually caused PBQ to pop off quite a lot over the last four months. And a lot of our users are

244
00:21:25.880 --> 00:21:41.560
spinning up OpenClaws and then they're throwing our API key in there and they're trying all different models and such. But you cannot run OpenClaw with on Sonnet or Opus purely unless you're like a very rich man. So they're using Kimi, they're using Minimax. And

245
00:21:41.945 --> 00:21:45.945
OpenClaw is actually pretty dumb with those models and it makes a lot of mistakes.

246
00:21:46.345 --> 00:21:52.345
And and then, yeah, unfortunately, I've been I'm I'm also head of customer service. So I'm the guy that

247
00:21:52.505 --> 00:21:54.745
everybody comes to and they're like, hey, this

248
00:21:55.065 --> 00:21:56.025
thing is not working.

249
00:21:56.570 --> 00:21:59.529
And it's super hard for me to solve because

250
00:21:59.850 --> 00:22:00.970
it's it's,

251
00:22:01.130 --> 00:22:02.809
it could be a PPQ problem.

252
00:22:03.050 --> 00:22:09.770
Could be some recent shipment we get, but it could also be the model itself is is messing up. It could be

253
00:22:10.304 --> 00:22:19.105
recent OpenClaw update. That thing is like OpenClaw itself is like a hot pile of trash in terms of bugs and such and sluggishness.

254
00:22:19.105 --> 00:22:19.904
And so,

255
00:22:20.225 --> 00:22:24.960
yeah, it's it's like it's a very it's like a black box of problems.

256
00:22:24.960 --> 00:22:27.200
Yeah. So I'm getting away from my point. But

257
00:22:27.360 --> 00:22:34.399
all I'm saying is that the open the open source models, like, when you've really put them to the test, and they really need a lot of context

258
00:22:34.915 --> 00:22:39.635
to make like decisions and such, you can really start to see them become dumb.

259
00:22:40.435 --> 00:22:42.755
Yeah. I mean, I've noticed that as well

260
00:22:43.155 --> 00:22:44.595
with my experimentation

261
00:22:44.595 --> 00:22:50.195
on Open Claw. And yet, to your point, my like beginning of Open Claw, I was just hitting Opus,

262
00:22:50.840 --> 00:22:56.360
and it just cost a shit ton of fucking money. And then of course I did the you know, humans

263
00:22:56.680 --> 00:23:01.880
love flip flopping from extreme. So then I like went to mini max and then it was retarded.

264
00:23:02.680 --> 00:23:10.065
And like, there's a happy path in the middle. But to your point, I mean, I think this is where something like PPQ really excels. Released

265
00:23:10.065 --> 00:23:10.785
a

266
00:23:12.065 --> 00:23:26.469
I guess like it's like a blended model. I don't know how how to explain it, but like I you you basically you set it up with open claw and then like it knows when it's supposed to hit the smart models, the more expensive models, and it knows when it should hit the cheaper,

267
00:23:26.789 --> 00:23:28.389
less smart models. How does that work?

268
00:23:29.345 --> 00:23:31.425
Yeah. So it's called AutoClaw.

269
00:23:31.425 --> 00:23:53.070
We actually borrowed it. Well, like, the the core of the code comes from an an open source project that we found. But we we added our PBQ tweaks to it and such. But, basically, it has a smart router on it. So it looks at the input. It looks at the question that the user's asking, and it does a keyword search essentially, a really extremely fast one millisecond keyword search on the input.

270
00:23:53.789 --> 00:23:54.830
Certain terms

271
00:23:55.149 --> 00:24:01.304
that exist in that in that input. And then if the terms kind of indicate that this is a complicated question,

272
00:24:01.465 --> 00:24:08.345
then it kicks it to the to the expensive model. And if it's a super easy question, then it kicks it to the cheap model.

273
00:24:08.665 --> 00:24:10.184
And it's actually been

274
00:24:10.825 --> 00:24:16.879
at first, it was wildly like popular people were loving it because it was saving them like 80%

275
00:24:17.440 --> 00:24:18.960
in costs or more.

276
00:24:20.240 --> 00:24:38.044
And that's really releasing that. It was our number one, like, blog post ever. Everybody loved it. And then OpenClaw did an update in the last month or so, and then it started breaking down and all sorts of problems were happening. Yeah. So I think I got it fixed, but it's still very murky. And, yeah,

277
00:24:38.605 --> 00:24:59.515
it it's it it takes a lot of my time to because I have to be an open claw expert now. But, you know, that's that's that's part of the job. So but, yeah, that's that's basically how it works. And that is the nice thing about PBQ is I have both the proprietary models and the open source models at my disposal. So I can throw in whatever models I want into that auto router,

278
00:24:59.755 --> 00:25:08.155
and and I can update them on the fly and and that the users don't have to do that. That's also very key. Right? They're not they're not on the cutting edge of

279
00:25:08.395 --> 00:25:19.519
of knowledge. So I'm kind of trying to make the decisions for them based on my best knowledge. And so, yeah, I'm plugging in the latest and greatest models the best that I can to keep their claws working.

280
00:25:24.159 --> 00:25:26.480
Yeah. I mean, that's wild. I think,

281
00:25:28.055 --> 00:25:32.615
by the way, to on the hiring point, like, customer service guys, the next

282
00:25:32.935 --> 00:25:35.654
thing you should hire for. So you're not doing

283
00:25:35.895 --> 00:25:36.774
it yourself.

284
00:25:37.735 --> 00:25:42.779
I like it, though. I really get a feel for I get a feel for the customers, and

285
00:25:44.780 --> 00:25:48.940
I I like my voice when I talk to the customers, but it is driving me

286
00:25:49.340 --> 00:25:53.259
slowly insane because it's, like, it's a twenty four hour job.

287
00:25:53.260 --> 00:25:53.420
Like,

288
00:25:54.265 --> 00:26:04.424
you know, I'm not I'm not waking up 3AM to answer messages or anything, but they do happen all the time. And I'm like, you know, focused in work and then a message pops up and I got a

289
00:26:05.145 --> 00:26:18.469
context switch to answer answer their problem or whatever. But yeah, there's definitely some merit to the founder managing support, and at least in the beginning, I mean, we're the largest investor in strike, and they're a significantly larger company than you.

290
00:26:20.070 --> 00:26:27.905
And they deal with people that's not answering the phone. He was but my point was he was until like six months ago, he was

291
00:26:28.545 --> 00:26:30.865
responding directly to support requests.

292
00:26:31.905 --> 00:26:36.545
And it was one of the things that was like, okay, you you got your plate full. It's

293
00:26:37.190 --> 00:26:45.669
not easy hiring for customer support because you need them to be technically competent and actually care about the business too, which is

294
00:26:46.790 --> 00:26:49.750
a difficult mix to find. To the point on the AutoClaw,

295
00:26:51.044 --> 00:27:01.764
do you think is this where we're going? I mean, to me, that feels like that's what the future looks like. It looks like I don't know if it looks exactly like that. But it looks like

296
00:27:02.085 --> 00:27:05.284
people using a variety of models where they make sense,

297
00:27:05.910 --> 00:27:07.190
rather than

298
00:27:07.750 --> 00:27:11.750
just having Claude, the top model Claude do everything for them.

299
00:27:12.230 --> 00:27:14.950
Yeah, certainly. That's absolutely the future.

300
00:27:15.190 --> 00:27:15.910
Yeah.

301
00:27:16.070 --> 00:27:22.365
It's all about the intelligence of the routing, right? That's the that's the key point. I think many users want,

302
00:27:22.685 --> 00:27:28.365
they would on the side of paying more and making sure that they were getting a quality answer

303
00:27:28.685 --> 00:27:31.325
than having catastrophic mistakes

304
00:27:31.325 --> 00:27:43.419
happen. Right? So you have to tweak the routing that way where I personally, I almost like we have different modes. We have eco mode, we have regular mode, and then we have premium mode. But I would almost like,

305
00:27:43.740 --> 00:27:49.900
I kind of want to retool the premium mode and say, like, the default should be use the expensive model.

306
00:27:50.495 --> 00:27:52.895
And then you have to find simplistic

307
00:27:54.335 --> 00:27:55.375
indicators

308
00:27:55.375 --> 00:27:56.655
to downgrade.

309
00:27:56.655 --> 00:27:58.815
Right? And that that would kind of guarantee

310
00:27:59.055 --> 00:28:03.455
that you're getting not maybe not guarantee, but much stronger accuracy.

311
00:28:03.455 --> 00:28:08.900
And but the routing is not easy. The way I'm doing it right now, it's just like basic keyword heuristics,

312
00:28:08.900 --> 00:28:18.420
which is like not scientific at all. It does a decent job for what it is, but the future it's gonna be much more. I think I can see a whole field

313
00:28:18.825 --> 00:28:29.705
of AI research going into this area. Is this not only that you have to make these decisions, but you have to make them super fast. You're basically running a query before you run the query. Right? What are your thoughts on like,

314
00:28:30.600 --> 00:28:33.639
so in my quest to save money, but also

315
00:28:33.880 --> 00:28:36.440
have competence. Cause I've, I've basically

316
00:28:36.840 --> 00:28:46.120
opted out of the entire subscription model. So I've been paying per token my entire self learned education experience, and it was very expensive at first. I've really grown to love

317
00:28:46.645 --> 00:28:47.845
specifically

318
00:28:48.085 --> 00:28:49.845
Gemini flash three.

319
00:28:50.005 --> 00:28:55.684
It's fast, it's cheap, and it's relatively competent. Like it doesn't make disastrous

320
00:28:56.005 --> 00:29:00.325
mini max mistakes or even Kimmy level mistakes and it's cheaper than Kimmy.

321
00:29:00.830 --> 00:29:16.510
Is the future something like using a model like that? That's like relatively competent, relatively fast, relatively cheap. And that's like pre screening stuff and then making the it's like the quarterback, you know, then it's like making the executive decision rather than like a basic keyword type of thing.

322
00:29:18.085 --> 00:29:19.044
I don't know.

323
00:29:19.765 --> 00:29:27.764
Again, like I said, I I think it's a complicated question. So I really don't know, but I'm glad you're enjoying Flash. It is very cheap.

324
00:29:28.005 --> 00:29:36.240
And it's actually one of the models that we have 30% off even below what you pay Google, because we have a bunch of free credits that we're trying to burn through.

325
00:29:36.800 --> 00:29:48.124
So that's why it's probably especially cheap. But yeah, I really don't know. Like I said, it's it needs to be done quickly. Right? So like, if you have to run a flash query before your real query happens,

326
00:29:48.445 --> 00:29:50.205
you're introducing latency,

327
00:29:50.205 --> 00:29:51.484
right? So it's definitely

328
00:29:51.885 --> 00:29:53.004
an art. Out

329
00:29:53.804 --> 00:29:58.365
of all the things, right? Like if you have cost, you have competency

330
00:29:58.365 --> 00:29:59.485
and you have latency,

331
00:29:59.970 --> 00:30:05.570
Like maybe those are the three main trade offs when you're like trying to do your setup. I think for end users,

332
00:30:05.970 --> 00:30:07.090
probably latency

333
00:30:07.090 --> 00:30:10.370
is the one they're most willing to take

334
00:30:10.370 --> 00:30:19.075
a hit on. You know, there's a bunch of tasks that like, at the end of the day, I just send the thing running, and then I go live my life and, you know, take care of kids and whatnot,

335
00:30:19.235 --> 00:30:20.674
or do another task.

336
00:30:21.155 --> 00:30:21.715
Yeah,

337
00:30:22.115 --> 00:30:27.955
I'd be willing to save money, but maintain competency. Like I would a 100% be willing to sacrifice.

338
00:30:27.955 --> 00:30:28.115
Yeah.

339
00:30:28.980 --> 00:30:29.779
Minutes.

340
00:30:30.260 --> 00:30:39.220
Yeah. You might be right there. And also in a world where you're, where you're setting it and forgetting it, and you're coming back fifteen, twenty five hours later,

341
00:30:39.940 --> 00:30:47.015
then it does make sense to let the dumb model, so to speak, have a crack at it. Because now we're getting to the point where

342
00:30:47.575 --> 00:30:52.935
you can even use the dumb models to test the dumb models in, you know, foolproof

343
00:30:52.935 --> 00:31:12.725
ways. Right? So if you can have the dumb model build something, and then you can write tests for it, or it will write tests for it. And then if it passes all the tests, then it determines that it did a good job. And if it doesn't pass the test, and maybe it upgrades the model and tries again, right? So yeah, I agree. Maybe latency is the least valuable of all those.

344
00:31:13.765 --> 00:31:17.684
You know, certain tasks, like if it's self driving, then obviously,

345
00:31:17.685 --> 00:31:18.804
latency matters.

346
00:31:19.765 --> 00:31:22.645
Because then you're in a car accident, if the thing has latency.

347
00:31:23.600 --> 00:31:35.280
But I think for a lot of things, people are willing to take the hit for it. The other big feature you launched recently that I was excited about is you added these models that are run-in secure enclaves similar

348
00:31:35.280 --> 00:31:37.520
to what Maple has been doing.

349
00:31:38.325 --> 00:31:39.525
How

350
00:31:39.525 --> 00:31:42.484
do you think about that in your full product suite

351
00:31:42.725 --> 00:31:44.164
and how you implemented I

352
00:31:45.044 --> 00:31:49.604
think it's I'm glad that we were able to use a provider that

353
00:31:49.765 --> 00:31:55.960
abstracts away a lot of the complexity there. Because what PBQ is trying to do many different things at a

354
00:31:56.600 --> 00:32:01.240
pretty good job on many things. And so we didn't have time to dig into

355
00:32:01.640 --> 00:32:14.914
how the secure enclaves work and this and that. And so we found a provider that did a lot of that for us. And so we were able to to get that out the door. So far, it's actually quite popular. It's hard to set up through API.

356
00:32:14.995 --> 00:32:16.914
So when you come to our UI,

357
00:32:16.915 --> 00:32:20.995
we do so the the thing is is the user needs to encrypt

358
00:32:20.995 --> 00:32:27.769
their content before they send it onto us. Right? And so when they're using, you know, PPQ in the browser,

359
00:32:28.490 --> 00:32:32.330
we can do that encrypting for them. But when they're using it through the API,

360
00:32:32.650 --> 00:32:37.130
we had to build out this special proxy that they basically clone themselves

361
00:32:37.665 --> 00:32:48.945
and deploy. So now every time they run an API request through us, it's running through that proxy first to encrypt everything. So it's a little bit extra setup on the API side. But, yeah, I

362
00:32:49.585 --> 00:32:54.750
I'm I am very excited about these models. I think that OpenClaw

363
00:32:54.750 --> 00:32:55.629
especially

364
00:32:55.630 --> 00:32:57.229
shows us that

365
00:32:58.269 --> 00:32:59.549
AI is gonna

366
00:32:59.950 --> 00:33:02.829
really hit mainstream when it can start solving

367
00:33:03.149 --> 00:33:08.110
personal problems. Right? But personal problems are where you want the most privacy.

368
00:33:08.725 --> 00:33:16.404
And so, these models, there are gonna be many use cases for them beyond what people are currently using them for.

369
00:33:16.725 --> 00:33:19.845
And that's gonna be very powerful, I think. You'll be able to

370
00:33:20.870 --> 00:33:24.390
run this stuff on a lot of your personal app data.

371
00:33:24.630 --> 00:33:32.470
And you will be able to do so with relative confidence that only you will have access to it. So yeah, I'm

372
00:33:32.470 --> 00:33:35.190
very bullish on these models. And yeah,

373
00:33:36.735 --> 00:33:42.575
we we've been we've been pleasantly surprised by how quickly people had adopted them. I

374
00:33:45.215 --> 00:34:06.265
yeah. I mean, look, I think that makes a lot of sense to me. I don't know how I like, I'm kinda interested, like, how many people are even aware of the concern at this point. Like, I think a lot of people are just, like, uploading their lives specifically to the proprietary models, usually under a subscription account that every search and prompt and data is all tied to the same user. But I do think the dream

375
00:34:07.065 --> 00:34:10.025
and I disagree actually with the Maple guys on this.

376
00:34:11.225 --> 00:34:13.385
The dream is to have it all in one place.

377
00:34:13.865 --> 00:34:16.665
You know? I think there's different schools of thought here.

378
00:34:18.309 --> 00:34:20.630
But I do like how you did it with PPQ.

379
00:34:20.630 --> 00:34:22.070
I mean, I noticed

380
00:34:22.150 --> 00:34:23.670
Venice, which is probably

381
00:34:23.910 --> 00:34:26.710
one of your bigger competitors has done similar.

382
00:34:26.950 --> 00:34:32.309
But I like the idea that you have the ability to use the proprietary models

383
00:34:32.484 --> 00:34:34.645
that are significantly more competent.

384
00:34:36.085 --> 00:34:38.085
Then you in the same interface,

385
00:34:38.244 --> 00:34:39.925
be able to use the,

386
00:34:40.325 --> 00:34:51.520
you know, open source models that are running in in these Ts, the secure enclaves for more private stuff. And I think it goes back to the to the earlier conversation with cost and competency

387
00:34:51.760 --> 00:34:52.880
and latency.

388
00:34:53.520 --> 00:34:57.520
Like my dream is that I have some kind of quarterback

389
00:34:57.520 --> 00:34:59.760
that's just automating in the background.

390
00:35:00.000 --> 00:35:01.840
Yeah. And it's probably a model, I think,

391
00:35:02.615 --> 00:35:04.775
which things are going private,

392
00:35:04.775 --> 00:35:09.895
which things are going proprietary and more expensive and less private, maybe even local.

393
00:35:10.135 --> 00:35:13.975
Right? You know, like mixing local in there. Maybe the quarterback's local,

394
00:35:14.375 --> 00:35:16.855
but it should all be kinda in one

395
00:35:17.890 --> 00:35:19.010
interface,

396
00:35:19.330 --> 00:35:23.330
one platform that is kind of doing it all in the background for me.

397
00:35:23.730 --> 00:35:24.290
Yeah,

398
00:35:24.530 --> 00:35:25.650
I've never

399
00:35:26.450 --> 00:35:29.330
thought about it that clearly. But I think you're absolutely right.

400
00:35:30.050 --> 00:35:45.785
That makes a lot of sense to me. These models have different capabilities and different trade offs. And you should most people will benefit from having access to all of them. And yes, the secret sauce will definitely be when there's a little quarterback making those decisions

401
00:35:46.030 --> 00:35:48.430
for you. So you don't have to have all this

402
00:35:48.670 --> 00:35:49.630
competency

403
00:35:49.630 --> 00:35:50.670
about

404
00:35:50.670 --> 00:35:52.350
what each model is doing.

405
00:35:53.470 --> 00:35:56.350
You'll always need have some of You'll

406
00:35:56.350 --> 00:36:05.435
always need to have the users, as you say, many of them are not even aware of the costs that they are bearing when they give up this data. But I think, you know,

407
00:36:05.755 --> 00:36:12.555
people will become more aware of that and they already are. And so, yeah, it's going to be this mental framework that everyone evolves to

408
00:36:12.875 --> 00:36:19.530
in the coming years. What data of mine is local only what data of mine is, am I sending to the cloud, etc?

409
00:36:20.810 --> 00:36:21.930
Yeah, I mean,

410
00:36:22.250 --> 00:36:31.345
like, this is a bigger this goes bigger than AI to me, you know, it's, it's my greater experience in the freedom tech movement

411
00:36:31.665 --> 00:36:32.785
and building

412
00:36:33.505 --> 00:36:38.385
projects and doing user education and dealing with users, end users directly.

413
00:36:39.105 --> 00:36:42.305
Like our biggest, whether that's Bitcoin or Noster or Tor,

414
00:36:42.910 --> 00:36:44.670
you know, message encryption,

415
00:36:45.470 --> 00:36:47.630
all these things historically

416
00:36:47.630 --> 00:36:49.950
have required immense personal responsibility.

417
00:36:50.350 --> 00:36:50.990
And then

418
00:36:51.310 --> 00:37:01.525
the result is personal responsibility is always a more difficult decision because you have to take responsibility, you don't have someone else doing it for you. So what happens is we have a ton of friction

419
00:37:01.845 --> 00:37:03.365
throughout these tools.

420
00:37:03.605 --> 00:37:05.605
And then because there's a friction,

421
00:37:06.485 --> 00:37:12.405
we'll never hit the level of scale that the controlled and surveilled alternatives have because they're just easy.

422
00:37:12.645 --> 00:37:13.445
And I always say,

423
00:37:14.760 --> 00:37:22.440
I love Ben, but if you have to watch a two hour BTC sessions video to use a tool, like we're already fucked. Like you're already screwed. And

424
00:37:22.920 --> 00:37:25.720
to me, one of the coolest aspects of AI is

425
00:37:26.040 --> 00:37:30.200
that we could reduce a lot of the friction across the stack

426
00:37:30.815 --> 00:37:37.375
by using open source models, by using models that are in secure enclaves to kind of remove

427
00:37:37.375 --> 00:37:44.255
a lot of that decision making from the user. But of course, there's still going to be a level of personal responsibility. I just think we can reduce that significantly.

428
00:37:44.255 --> 00:37:46.800
Yeah, for sure. Yeah, I agree.

429
00:37:50.480 --> 00:37:55.279
Yeah, I mean, I don't know how are you thinking about

430
00:37:58.035 --> 00:38:12.515
I mean, so we briefly mentioned OpenClaw. I mean, I think that was probably one of the big highlights of the year since we last recorded. I mean, didn't you clearly have your concerns about it, but I think as on a high level, just the idea of

431
00:38:13.010 --> 00:38:14.530
a open source

432
00:38:15.089 --> 00:38:16.530
configurable agent

433
00:38:16.530 --> 00:38:21.810
that lives on your computer and has, I mean, one of the fastest growing open source projects

434
00:38:21.970 --> 00:38:23.250
that we've ever seen

435
00:38:23.809 --> 00:38:43.515
is pretty cool to witness and changes a lot of the conversation and narrative around the space. Do you think it's a flash in the pan? Do you think it's actually a taste of what's to come? How do you think about it? Yeah. No. I don't think it's a flash in the pan at all. My gripes with it are just how buggy it is and how much customer service problems it causes me. It's a good problem because it's it's

436
00:38:43.970 --> 00:38:56.210
people are warming to it. They're trying they want to they want an AI personal assistant. That's what everybody wants. And so no, these things are only gonna get better. They're gonna get less buggy. They're gonna become more capable. And,

437
00:38:57.250 --> 00:38:57.490
I mean,

438
00:38:58.665 --> 00:39:08.665
normal people are going to be doing incredible things with their AI assistants. And it's gonna be removing a tremendous amount of friction in everybody's lives. When you can have

439
00:39:08.985 --> 00:39:13.670
something, schedule your flights for you, and actually do it correctly.

440
00:39:13.670 --> 00:39:21.670
And like, these are all the dreams, but it's not happened yet. But, you know, book your Airbnb for you and and do it correctly.

441
00:39:21.670 --> 00:39:22.710
Give you a

442
00:39:23.349 --> 00:39:29.464
the my my big want is I want a morning podcast that's generated for me with all the things that I care about,

443
00:39:29.704 --> 00:39:44.270
five minutes every morning. That's kind of my little personal project. But there's so many of these ideas out there, right? And it's only going to make us like able to live our lives doing more of the things that we want to do. And the AI assistants are gonna,

444
00:39:44.510 --> 00:39:46.190
you know, handle a lot of the

445
00:39:47.150 --> 00:39:50.190
for us. So no, I think that

446
00:39:50.350 --> 00:39:53.390
it's going to be a huge, huge movement,

447
00:39:53.470 --> 00:39:57.315
whether it's OpenCLOS specifically, maybe it will be. But there's already,

448
00:39:57.635 --> 00:40:02.995
you know, a dozen other personal assistant agents out there. So we'll see which ones win out.

449
00:40:03.395 --> 00:40:05.315
But, you know, it doesn't really matter.

450
00:40:05.635 --> 00:40:17.580
We know that it's all gonna get much better. Even in six months, I can't imagine how much better they're gonna Yeah. No, I mean, I agree. That's why I like this cadence. Oh, you know, hopefully we'll do it again in a year. So much happens. The

451
00:40:18.060 --> 00:40:21.500
it's funny you mentioned the personalized podcast

452
00:40:21.500 --> 00:40:24.380
to update you. Cause one of the first things I did was

453
00:40:25.075 --> 00:40:28.915
like, I built a AI news desk that is giving me

454
00:40:29.315 --> 00:40:30.115
updates,

455
00:40:30.115 --> 00:40:33.155
and then I made it public. It's at colyr.com.

456
00:40:33.155 --> 00:40:35.395
But it says text based. Open clock?

457
00:40:35.795 --> 00:40:37.715
Yeah. I have I have multiple.

458
00:40:38.600 --> 00:40:50.520
I have a top top news desk agent, and then he has sub agents underneath him. And he thinks he's like a Walter Cronkite type of individual. And he's like pulling in a bunch of different data. And then every three hours,

459
00:40:51.165 --> 00:40:53.885
they're sending out updates on Noster.

460
00:40:53.885 --> 00:40:57.405
So they're signed, hashed, hashed and signed Noster posts,

461
00:40:57.485 --> 00:40:59.485
just very brief text.

462
00:40:59.965 --> 00:41:01.885
And what's cool is every

463
00:41:03.245 --> 00:41:05.005
three hours they're making this

464
00:41:05.325 --> 00:41:12.740
newswire that they're sending out, and they're keeping all the old ones. So they're learning and getting more advanced over time.

465
00:41:13.539 --> 00:41:15.700
At first, I'm

466
00:41:15.700 --> 00:41:16.980
pretty sure they think,

467
00:41:17.619 --> 00:41:26.845
say he, I'm pretty sure these soulless robot slaves like think this is the first war ever that's ever happened, because they were born into the Iran conflict.

468
00:41:27.005 --> 00:41:36.525
So they're like, they're like, gold is crashing in the war era. Like, there's a little bit of work that needs to be done there. But it's quite cool. And to me, it's interesting, right? Because

469
00:41:37.299 --> 00:41:49.380
a lot of people have approached it, oh, monitoring the situation. We need to monitor the situation with AI. And they release, like, World Monitor, and there's a bunch of forks and all this other stuff. And what it does is it's all it's, like, gives you way too much information.

470
00:41:49.955 --> 00:41:51.555
And to me, the cool

471
00:41:52.275 --> 00:41:55.875
liberating aspect of this tech is actually the opposite of that.

472
00:41:56.435 --> 00:41:57.235
I don't wanna spend

473
00:41:58.755 --> 00:42:10.580
hours looking at, you know, ship patterns and all this other shit. Like, I want my agents to just tell me, yo, you need to know this. You don't have to watch any ads or go through any of this stuff, and then you can go live your life.

474
00:42:10.980 --> 00:42:17.860
So I kinda looked at it from the exact opposite of that. Let them deal with all the information and then let them give me a distilled kinda

475
00:42:18.260 --> 00:42:19.380
brief on it.

476
00:42:21.485 --> 00:42:31.645
But on the So you could You could use the next step could be for your Newswire thing, you could use our TTS model. We have both 11 Labs and Deepgram,

477
00:42:31.725 --> 00:42:40.280
where you could convert that to the thing that's missing. There's still a missing link in this whole awesome podcast generation thing is you can create a file

478
00:42:40.520 --> 00:42:42.040
of of a podcast,

479
00:42:42.280 --> 00:42:46.280
but it does not easily plug into Spotify or Apple Music or whatever.

480
00:42:46.600 --> 00:42:51.235
So there needs somebody needs to create, I already have a name for it, RSSify.

481
00:42:51.635 --> 00:43:01.155
So it basically takes a file, and then it puts it into some RSS feed that's kind of private where only the person with the link can access it. But then

482
00:43:01.370 --> 00:43:12.810
you could literally just have it auto uploading into your Spotify and and all of the friction would be gone. That's what's prevented me from, like, really, really Well, there's the bot tasks. Yeah. Go ahead.

483
00:43:13.210 --> 00:43:27.155
I mean, I think when you start playing with these things, you start to realize that open protocols are way easier to interact with. And what's nice is RSS is an open protocol. Noster's it's all these things work really well with Noster. Obviously not if you want privacy,

484
00:43:27.235 --> 00:43:31.475
but if you just have them upload the MP three to Blossom,

485
00:43:31.475 --> 00:43:32.515
which is the Noster

486
00:43:33.260 --> 00:43:34.940
media spec for

487
00:43:35.500 --> 00:43:36.300
videos,

488
00:43:37.099 --> 00:43:55.015
pictures, and and audio files, m p threes, it's really easy to turn a Noster feed then into an RSS feed. Like, Gigi has a simple tool that he vibed that just turns any Nostril feed into an RSS feed. What's nice is then you don't have to store it or distribute it. You could share with other people. Like, I think it's kind of cool that

489
00:43:55.575 --> 00:43:57.015
my little news desk,

490
00:43:57.335 --> 00:44:10.980
my parents use it. You know, I can just like send them a link. Obviously, anyone in an Oster client can use it. So just something to think about. But yeah, I understand why you would want it in your regular podcast app. But you could even like, I don't know Spotify.

491
00:44:10.980 --> 00:44:17.835
Spotify is like the evil empire in terms podcasting. Yeah, they let you just put in a bare RSS feed. But like Apple podcasts let you

492
00:44:18.555 --> 00:44:21.515
I think you can apply on Android lets you

493
00:44:21.835 --> 00:44:27.035
Yeah, I think I think if it's on an RSS feed, you can get it into Spotify. Yeah, I is the

494
00:44:27.330 --> 00:44:39.810
Blossom thing you're mentioning. Is that something that is, is that an app I can have on my phone where it automatically just loads into my phone? Because that's the thing for me. Blossom is a protocol for hashing

495
00:44:39.810 --> 00:44:40.050
and

496
00:44:41.065 --> 00:44:42.025
replicating

497
00:44:42.025 --> 00:44:44.425
media files. So you don't have to host them.

498
00:44:44.905 --> 00:44:50.425
Okay. So where's the app? Right. Where's the app that I can just press? Well, that's the cool piece is that,

499
00:44:53.680 --> 00:45:01.280
I mean, obviously, you could listen to them in a Nostrap, but Gigi has a tool. I I forget what it's called that converts any

500
00:45:01.440 --> 00:45:09.805
so, like, this right? So SIL dispatch, you can search for it in any podcast app. Right? It's just hosted like a normal RSS feed.

501
00:45:10.045 --> 00:45:12.845
But I also upload the signed and hashed

502
00:45:13.644 --> 00:45:16.605
audio files to Nostr using Blossom.

503
00:45:16.765 --> 00:45:19.805
Right? So I have a Nostr account primal.net/citadel.

504
00:45:19.920 --> 00:45:23.200
It has its relative, you know, n pub or whatever.

505
00:45:24.240 --> 00:45:29.280
GG's tool can convert that noster feed into a completely different RSS feed

506
00:45:29.599 --> 00:45:30.560
that is

507
00:45:30.799 --> 00:45:40.105
not reliant on a single RSS host. Like, I use pod home as my RSS host. I use a Bitcoin or it's a smaller RSS host. But if Pod Home goes down,

508
00:45:40.585 --> 00:45:42.825
my normal Citadel dispatch feed you lose.

509
00:45:43.065 --> 00:45:43.865
If

510
00:45:44.025 --> 00:45:56.060
you are listening through the RSS feed that is made from the Noster feed, those Blossom files are actually hosted in like six or seven different Blossom servers. And because they're hashed and signed, you can replicate them

511
00:45:56.540 --> 00:45:59.420
to infinity and not worry about verifiability.

512
00:45:59.420 --> 00:46:00.780
Right? Because they're hashed. Yeah.

513
00:46:02.255 --> 00:46:07.055
So it's kind of fascinating to me. So so it gives it gives dispatch for instance,

514
00:46:07.615 --> 00:46:14.975
some resilience and robustness from having a single RSS host. Yeah. Yeah. Makes sense. But anyway, we kinda

515
00:46:15.549 --> 00:46:22.590
I wanted to talk about OpenCLO a little bit more. I mean, so the reason so I don't think it's a flash in the plan. I agree with you.

516
00:46:23.470 --> 00:46:25.310
And I think it was really cool to see.

517
00:46:25.630 --> 00:46:28.109
And to me was like kind of an moment

518
00:46:28.109 --> 00:46:34.735
on, oh, maybe the open source side of this thing can really win out. But why I say a flash in the pan is

519
00:46:35.295 --> 00:46:36.815
we're already seeing,

520
00:46:37.455 --> 00:46:39.455
you know, anthropic has just been

521
00:46:39.855 --> 00:46:42.015
well, first they sent him a cease and desist.

522
00:46:42.655 --> 00:46:55.980
Then he went to open AI, he got like kind of Aqua hired, kinda what I don't even know. It's an open source project, so it's a little bit weird. But I get why he did it. And then Claude, they've just been releasing left and right. Basically, proprietary

523
00:46:56.460 --> 00:46:57.420
alternatives

524
00:46:57.420 --> 00:46:58.460
to Yeah.

525
00:46:59.365 --> 00:47:00.325
OpenClaw.

526
00:47:00.325 --> 00:47:01.125
Obviously,

527
00:47:01.525 --> 00:47:06.245
know, we were talking about personal assistance. This is what the selling point was for Siri

528
00:47:06.325 --> 00:47:21.040
all those years ago, ten years ago, like Siri's still retarded. But the idea was Siri was supposed to book reservations for you and stuff. Google's obviously working on Gemini stuff, already has people all in G Suite. So when I say flash in the pan, I don't mean like disappearing completely,

529
00:47:21.359 --> 00:47:21.920
but

530
00:47:22.720 --> 00:47:30.480
how do you think about that dichotomy? Like, I feel like the overwhelming majority of users are gonna use whatever Apple's

531
00:47:28.755 --> 00:47:40.515
you know, centralized proprietary alternative is, whatever cloud centralized proprietary alternative is, whatever Google's alternative is versus, you know, an open source, whether it's open claw or something else out there.

532
00:47:41.154 --> 00:47:47.130
Yeah. I don't know. I will say that my one angle of knowledge on this is that

533
00:47:47.530 --> 00:47:50.970
if our Open Claws or whatever you're gonna call them,

534
00:47:51.850 --> 00:47:52.970
if they require

535
00:47:52.970 --> 00:47:54.410
many modalities,

536
00:47:54.410 --> 00:48:01.665
right? If, you know, right now they're just chat heavy, but I just told you, like, can, you can use TTS, you can use STT,

537
00:48:01.665 --> 00:48:03.905
you can use video generation,

538
00:48:03.905 --> 00:48:05.105
image generation.

539
00:48:05.665 --> 00:48:06.945
There's all sorts of modalities.

540
00:48:07.890 --> 00:48:13.810
And what I've discovered running PBQ is that no one company is best in class.

541
00:48:13.970 --> 00:48:18.369
And even even a few of those things. Anthropic has the best coding models,

542
00:48:18.690 --> 00:48:25.255
but I don't even, they don't even have any media models, right? OpenAI kind of has maybe the best all around

543
00:48:26.135 --> 00:48:27.175
chat model.

544
00:48:27.494 --> 00:48:33.414
But its media is not that great anymore. In fact, they just scrapped Sora two, I think a day or two ago.

545
00:48:34.140 --> 00:48:35.740
They were losing a shit Yeah. Ton of money on

546
00:48:36.940 --> 00:48:40.700
So my point here is that if the best personal assistants

547
00:48:42.859 --> 00:48:45.820
are very multimodal in the future, then

548
00:48:46.539 --> 00:48:52.825
the best personal assistant will be the one that can connect to the technologies of many companies instead of one company.

549
00:48:53.145 --> 00:49:00.984
And so, that's an argument to say that maybe open source or, you know, build your own assistant kind of wins, right? Where you can just choose

550
00:49:01.640 --> 00:49:03.640
the best from each category.

551
00:49:03.640 --> 00:49:04.840
Yeah. That's

552
00:49:04.840 --> 00:49:08.280
my only one area of knowledge in that in that argument.

553
00:49:08.760 --> 00:49:10.440
I think so. I think that's a pretty

554
00:49:11.160 --> 00:49:14.040
I think that's a solid argument. I think the question then becomes,

555
00:49:14.515 --> 00:49:22.035
can we get the cost down? And I think the answer is probably yes, because obviously there's a big discrepancy between people using the

556
00:49:22.515 --> 00:49:24.595
proprietary stuff on subscriptions.

557
00:49:24.595 --> 00:49:25.715
But that makes sense to I

558
00:49:26.355 --> 00:49:34.300
think there's a unique advantage there that open source has that the proprietary stuff doesn't. I also think there's not as much vendor lock in as people realize.

559
00:49:35.020 --> 00:49:36.780
Yeah, agree with that too.

560
00:49:39.180 --> 00:49:40.220
Can move around more.

561
00:49:41.115 --> 00:49:43.195
You can, yeah. Like I said,

562
00:49:43.595 --> 00:49:52.955
Claude had the best coding model, but now there are several models out there that are pretty close. And we also learned that it's not just the model, it's also the tooling and stuff.

563
00:49:53.435 --> 00:49:55.595
And it's amazing how fast

564
00:49:56.400 --> 00:50:04.720
some new proprietary thing comes out and then there's a very strong open source competitor within a few days. And so, yeah, it's

565
00:50:04.720 --> 00:50:09.440
very interesting to watch. Just today, Anita Posh was saying how

566
00:50:09.920 --> 00:50:11.200
she is I'm

567
00:50:11.875 --> 00:50:16.435
guessing you read this also. It's one of the bigger, bigger notes today on Oster.

568
00:50:16.675 --> 00:50:21.315
But yeah, Anthropic was blocking her credit card because she's in Africa.

569
00:50:21.475 --> 00:50:25.635
And so now she can't use clock co work. And she's like, PBQ

570
00:50:25.635 --> 00:50:27.340
doesn't work for CloudCowork,

571
00:50:27.340 --> 00:50:34.060
which I looked into, and it seems that is true. You cannot bring in outside API API key into CloudCowork.

572
00:50:34.380 --> 00:50:41.144
But then I just did a quick search, and there's, like, a repo out there called OpenWork

573
00:50:41.144 --> 00:50:42.744
or OpenCowork.

574
00:50:42.904 --> 00:50:44.025
Open code?

575
00:50:44.184 --> 00:50:46.505
This is different. OpenCowork. OpenCowork.

576
00:50:46.585 --> 00:50:47.145
Yeah.

577
00:50:48.424 --> 00:50:52.490
And and so that is, like, I would say probably 95%

578
00:50:52.490 --> 00:50:57.210
as good as Claude Cowork. And it's just a matter of, can she spin it up quickly?

579
00:50:57.530 --> 00:50:58.170
And

580
00:50:59.130 --> 00:51:00.570
that's the challenge.

581
00:51:02.090 --> 00:51:08.305
Open source tools, they're always a little harder to get up and running. So can we lower that barrier of entry enough?

582
00:51:09.585 --> 00:51:11.185
Yeah. That makes sense to me.

583
00:51:14.224 --> 00:51:15.265
I saw

584
00:51:17.265 --> 00:51:21.184
I saw you popping off about

585
00:51:23.610 --> 00:51:29.450
I don't know. Would you what to call it? Bitcoin maximalism or something about the fact that you accept too many different payments.

586
00:51:29.930 --> 00:51:31.930
I think we talked about it last time too.

587
00:51:35.664 --> 00:51:44.545
I may have a lot of strong thoughts on that. Yeah. What you wanna discuss that briefly? I mean, you're you're accepting basically every payment under the sun right now. Right?

588
00:51:44.865 --> 00:51:47.105
Yes. Yeah. We we accept everything.

589
00:51:47.424 --> 00:51:59.090
Our focus is Bitcoin. We like to spend our development time building out new Bitcoin tech as much as possible. You're not gonna see us, like, building out Solana

590
00:51:59.490 --> 00:52:01.650
this or Ethereum that.

591
00:52:01.810 --> 00:52:25.160
We will slap on Solana and Ethereum payment methods because it's easy. But, yeah, my my my argument, my feelings are, yeah, very strong in this area because I think that Bitcoiners or Bitcoin only types of people. It's harmful to Bitcoin. And I can give very concrete examples of this as I've learned over the years in my work. So I started off

592
00:52:25.880 --> 00:52:46.100
well, I didn't start off at BitRefill, but I worked at BitRefill for several years. And most people know that they are kind of a Bitcoin first company, but they also accept pretty much everything. I I was there long enough and I was analyzing the data to see that the share of Bitcoin payments was gradually going down over years. And the share of altcoins and stablecoins and whatever was going up.

593
00:52:46.500 --> 00:52:59.245
And so if BitRefill had made the choice to be Bitcoin only, I don't think BitRefill would be around anymore. I think they would have been subsumed by competitors that accept all of these other coins. So do you want,

594
00:53:00.445 --> 00:53:08.285
you know, kind of a Bitcoin first company that makes rational decisions about this stuff? Or do you want kind of just like very

595
00:53:08.845 --> 00:53:09.645
purist

596
00:53:09.805 --> 00:53:13.150
type of companies out there who kind of struggle.

597
00:53:13.150 --> 00:53:16.190
And I faced this several times in marketing PBQ.

598
00:53:16.190 --> 00:53:16.750
So

599
00:53:17.230 --> 00:53:18.190
wanted to

600
00:53:18.350 --> 00:53:28.735
so when PBQ first came out, it was actually lightning only. That was not always my plan. I always knew that the other payment methods were gonna come because I worked at BitRefill and I understood the reality.

601
00:53:29.295 --> 00:53:49.890
But I started with Lightning because I I thought Lightning is the coolest use case for PBQ. But I tried to post it on the Bitcoin Reddit at first, and it was just completely ignored for the first, like, six months. And then by that time, I had started adding in other payment methods. And so I went back. And so when we got our round of funding, I think, is when I asked again.

602
00:53:50.450 --> 00:54:00.155
And then the guy finally answered me. I don't know who's in charge of it, but he answered me and said, well, no. You're you're pedaling shitcoins, so I'm not gonna post your your thing.

603
00:54:00.555 --> 00:54:06.155
And so then I would go over to the cryptocurrency redditr/r/cryptocurrency.

604
00:54:06.795 --> 00:54:11.530
And they don't like me because I'm Bitcoin first, So they're not posting me either.

605
00:54:12.650 --> 00:54:13.530
And and

606
00:54:13.770 --> 00:54:16.410
and this story has played out several times.

607
00:54:16.730 --> 00:54:28.595
I've tried to, like, I tried to make a posting on Bitcoin or jobs because I want to hire Bitcoiners. Right? But they won't post my jobs on Bitcoin or jobs because because I I pedal shitcoins.

608
00:54:28.675 --> 00:54:31.715
Even though some of the companies like Strike, for example,

609
00:54:32.195 --> 00:54:37.315
they've had, you know, they they coins on Ethereum in their app for years, I think.

610
00:54:38.870 --> 00:54:40.630
Yeah. We support Tether

611
00:54:40.710 --> 00:54:42.070
on Tron actually,

612
00:54:42.150 --> 00:54:44.790
globally. And that's because that's where the users are.

613
00:54:45.030 --> 00:54:52.550
Users use Tether on Tron as a dollar rail. And in America, people use Bank of America. And in Africa, they use Tether on Tron.

614
00:54:53.095 --> 00:54:54.615
We supported over there.

615
00:54:55.655 --> 00:54:56.935
Even when strike

616
00:54:57.015 --> 00:55:07.415
went into Argentina, they were originally like USDT on Ethereum as well. But anyways, it doesn't matter. But yeah, this has played out several times for me. And I do think that that attitude,

617
00:55:08.940 --> 00:55:12.220
it it people who actually care about Bitcoin are businesses.

618
00:55:12.700 --> 00:55:23.825
And I I don't believe that PBQ would survive being Bitcoin only, especially because PBQ is a scale business. The more volume we bring in, the more

619
00:55:24.145 --> 00:55:29.744
volume discounts we can get from our providers and stuff like that. So we have to go for the whole pie.

620
00:55:31.105 --> 00:55:33.744
Yeah, I'm not articulating it perfectly, but

621
00:55:35.840 --> 00:55:38.320
yeah, I I do think that it's

622
00:55:39.120 --> 00:55:44.800
I think that a lot of other Bitcoin companies out there have been kind of scared into not

623
00:55:45.120 --> 00:55:54.635
adopting these other payment methods or or doing practical things because they're afraid of being called out for this or that. And I'm just trying

624
00:55:54.795 --> 00:56:00.555
to do my best to get us past that where it's like, I I don't, you know, we

625
00:56:00.875 --> 00:56:06.930
we shouldn't be, like, beholden to these people who are, you know, strictly Bitcoin only.

626
00:56:07.010 --> 00:56:08.610
And, you know, I think

627
00:56:08.770 --> 00:56:21.625
maybe for a handful of businesses, you could do that, and it could work. But for for many other businesses, you just need to go after as many customers as possible. And you need to solve real problems in their lives.

628
00:56:21.945 --> 00:56:25.705
And yes, altcoins and stablecoins, stablecoins especially,

629
00:56:25.945 --> 00:56:30.984
do solve real problems for people all around the world. I look at Bitcoin more as a

630
00:56:31.470 --> 00:56:36.590
century long battle, and I do think we will end up at Bitcoin

631
00:56:36.670 --> 00:56:37.470
only

632
00:56:37.550 --> 00:56:43.230
at some point in our history or at some point in the future. You and I may be dead by then.

633
00:56:43.470 --> 00:56:45.310
But I do think that's

634
00:56:45.470 --> 00:56:54.085
it's it's, you know, it's a very long journey that we're going on. And I don't think that the right thing right now is to, like, self segregate.

635
00:56:54.085 --> 00:56:54.645
Yeah.

636
00:56:55.445 --> 00:56:59.125
Yeah. I mean, I think I align with your perspective more than I don't.

637
00:57:00.650 --> 00:57:01.690
And I think

638
00:57:02.490 --> 00:57:07.530
look. I think it's a natural response to so many scams in the ecosystem,

639
00:57:07.850 --> 00:57:09.690
in the greater crypto ecosystem.

640
00:57:09.690 --> 00:57:17.615
And it's a relatively easy litmus test, like, if people accept crypto versus Bitcoin only. But a perfect example in your own industry,

641
00:57:18.175 --> 00:57:20.815
I would say is Venice, who I mentioned earlier.

642
00:57:21.455 --> 00:57:23.535
Like they have a whole token scheme

643
00:57:23.615 --> 00:57:32.790
built into their app. Right? They have a pre mined Venice token that offers yield. It's operating on like a BNB style system.

644
00:57:32.950 --> 00:57:44.005
That is a core aspect of their entire app versus someone like you who is merely accepting it as payments. There should be a delineation there. Like, it's a completely different delineation.

645
00:57:44.325 --> 00:57:47.445
And when I if I remember from our last conversation,

646
00:57:48.164 --> 00:58:02.240
you know, you're a Bitcoin first company, you hold your profits in Bitcoin, you offer discounts in Bitcoin. Like, it should be very obvious to anyone who actually looks more than just surface deep that that of the differences between those two types of companies.

647
00:58:03.120 --> 00:58:06.640
Yeah. I I hope so. But, like, you know, not a

648
00:58:07.345 --> 00:58:22.305
event is helping people. And I do think they're helping some people. They're providing privacy for some folks, then that's great. But I am I am I am a Venice hater, I would say for the reason you brought up, they do have this, like, overly complicated token

649
00:58:22.490 --> 00:58:28.730
thing that basically is the entire design of it is to just bleed money from ignorant investors,

650
00:58:28.730 --> 00:58:45.135
right? That's the purpose of it. I've seen Eric himself, like describe the token mechanics and this and that. And even in his own tweet, he's like, I don't know what the price of this token asset should be. Then at the end of his tweet, it's like, Oh, I guess it should just be the cost of AI and then the capital.

651
00:58:46.175 --> 00:58:46.735
The

652
00:58:47.055 --> 00:58:48.495
what is that called the

653
00:58:48.750 --> 00:58:49.630
capital

654
00:58:49.710 --> 00:59:01.470
when you hold it for a period of time, the cost of capital, right? Yeah. And he's like, guess that's the true price of it. And it's I just wanted to reply like, good job. Figured it out. That's exactly what it is. There's really no need for a token.

655
00:59:02.670 --> 00:59:02.910
But

656
00:59:04.375 --> 00:59:05.255
But yeah,

657
00:59:05.495 --> 00:59:14.055
and then other things like that Venice has been they overly promise privacy, and they have for a long time. And he glosses over the privacy stuff.

658
00:59:14.295 --> 00:59:43.905
PBQ has essentially the same model that Venice has had, but we're just more honest about it. We tell our customers like, hey. You need to trust that we're not looking at your queries. You need to know that when you're using a proprietary model, your content is going to those guys. So just use that as you may. But Venice is just very loosey goosey with those things. And the thing that really kills me is, like, they it works for them. He's he's leveraging his million followers or whatever,

659
00:59:44.065 --> 00:59:44.865
and he

660
00:59:45.090 --> 00:59:47.970
his kind of reputation in the space as

661
00:59:48.050 --> 00:59:51.570
formerly, like, cypherpunk ish, you know, swap

662
00:59:51.890 --> 01:00:01.645
guy. And I don't think he actually, like, cares about privacy all that much for his users. They recently adopted TEE models just just like we did.

663
01:00:02.045 --> 01:00:07.244
Yep. And that's great. Like I said, I I hope I'm not coming on too too bitter,

664
01:00:07.325 --> 01:00:16.440
but it's like he only did it after people were really, really calling him out for a lot of the shenanigans that he was saying on Twitter. Right? So, yeah, I mean,

665
01:00:17.320 --> 01:00:26.360
it's a it's it grinds my gears that I'm the little guy over here trying to trying to fight the good fight. And then he's just like, you know, tweeting all these very

666
01:00:26.440 --> 01:00:58.490
brash things about about his product. But, yeah, it is what it is. And that's that's the best of it. You have to use the tools that you have. He's using his reputation, and I'm using whatever tools I have. You know, I can criticize him and, and I have right on my side. I'm not I don't have a large voice, but I can kind of pick pick apart things. So it is what it is. I mean, you can say a lot of things about Voorhees. And I tend to agree with pretty much everything you said, but he's really good at making money. He I mean, he had Satoshi dice.

667
01:00:58.985 --> 01:01:04.665
He sold it before transaction fees made it an untenable business model all for Bitcoin.

668
01:01:04.985 --> 01:01:06.185
Then he launched

669
01:01:06.425 --> 01:01:07.305
shapeshift

670
01:01:07.305 --> 01:01:17.920
right on the verge of, of shit coins, you know, hitting the populace made bank off of that. And then he seemed to you know, he's always in the right place at the right time and executes

671
01:01:17.920 --> 01:01:18.480
and

672
01:01:18.720 --> 01:01:22.560
he takes a lot of shortcuts, you know, but it's kind of crazy to watch.

673
01:01:23.040 --> 01:01:23.520
But

674
01:01:24.079 --> 01:01:27.760
I think that's just a good I mean, it's a perfect example for that conversation,

675
01:01:27.839 --> 01:01:28.000
right?

676
01:01:28.695 --> 01:01:33.255
Which is, if you can if you compare and contrast PPQ versus Venice,

677
01:01:33.335 --> 01:01:40.455
it's obvious which one is trying to be user focused and ethical and try and do things the right way,

678
01:01:40.775 --> 01:01:45.670
and be Bitcoin first, versus the one that is clearly a shitcoin operation.

679
01:01:47.190 --> 01:01:49.670
And and to be clear, Venice also,

680
01:01:49.750 --> 01:01:51.190
because he's a good businessman,

681
01:01:51.270 --> 01:01:57.750
like he self hosts BTC pay, he accepts Bitcoin, along with every other payment method. Yeah, he knows exactly what he's doing.

682
01:01:58.545 --> 01:01:59.265
Yeah.

683
01:02:00.145 --> 01:02:05.985
Matt, at one point there, I knew you as the you were the data guy, you know, you had useful tulips.

684
01:02:06.464 --> 01:02:12.385
You're the data guy. Do you have any data for us? That's interesting on the PPQ side? I mean, the two big things that I'm curious about

685
01:02:13.100 --> 01:02:15.740
is, like, user behavior on

686
01:02:15.980 --> 01:02:17.100
model selection

687
01:02:17.340 --> 01:02:18.060
and

688
01:02:18.540 --> 01:02:23.660
user behavior on the payments. Like, how are the Bitcoin payments your most common payment?

689
01:02:24.220 --> 01:02:31.855
Yeah. They are. Yeah. I'll I'm I'm trying I would love to share as much as I can, but I wanna hold a little bit back. But I will say Feel free.

690
01:02:32.175 --> 01:02:33.775
Yeah. Yeah. Bitcoin,

691
01:02:33.775 --> 01:02:37.695
Lightning specifically is our over half, just slightly over half.

692
01:02:38.735 --> 01:02:39.855
Monero's number two.

693
01:02:41.740 --> 01:02:47.820
Interesting. And number three and four are kind of a balance between Tether and credit cards.

694
01:02:48.140 --> 01:02:53.900
And then the rest are just various other stables and Litecoin and stuff like that.

695
01:02:55.415 --> 01:02:58.775
What's a high credit cards or or Tether?

696
01:02:59.735 --> 01:03:00.375
They

697
01:03:00.615 --> 01:03:09.735
I honestly don't know. They're they're pretty neck and neck, I think. It depends on the month, really. I think Tether is up and coming now more and more than it was. But I actually look at this kind of

698
01:03:11.620 --> 01:03:28.585
when you look at the BitRefill stats and BitRefill, in my opinion, is the best cross section of all crypto payments. They're the largest online crypto ecommerce website. Right? So Lightning is about five percent of their volume. Maybe maybe higher than that, but not not much higher. And so if if I want

699
01:03:29.224 --> 01:03:34.745
I wanna get into the pie that they have. Right? So I need to increase my awareness outside of just

700
01:03:34.905 --> 01:03:36.825
the Lightning and Noster circle.

701
01:03:36.984 --> 01:03:37.305
Right.

702
01:03:37.890 --> 01:03:38.530
So

703
01:03:38.690 --> 01:03:42.609
that's what I'm trying to do. I do wanna bring in those other users.

704
01:03:42.690 --> 01:03:44.609
And yeah, I I I

705
01:03:44.849 --> 01:04:11.760
think that over time, they will find that, like, there's a lot of really cool things that like lightning can do for them. It's our lowest minimum deposit because the tech allows for that. You can deposit as little as 10¢ in Lightning. And I think the other payments are all a dollar minimum because of the fees. Among many other things, it's very private. It's fits really well with a lot of these agentic payments coming up. Yeah. So, yeah, that that's that. In terms of models, again, I think

706
01:04:12.480 --> 01:04:21.040
I don't know. I don't know if I have really that great insight on model selection and whatnot. I think that the proprietary models are still the most popular.

707
01:04:22.395 --> 01:04:26.155
Aside from being more competent, they also have the brand recognition

708
01:04:26.155 --> 01:04:32.875
that people just they don't know what they never heard the word Kimi before. They're just gonna click on the word they have heard of called Claude.

709
01:04:34.810 --> 01:04:38.890
Stuff like that. But yeah, I think it really varies by business.

710
01:04:39.130 --> 01:04:45.290
I don't think PPQ is like an even I know that some of my competitors they

711
01:04:45.450 --> 01:04:56.265
have way more popular open source usage on their platforms and the proprietary models are not like the absolute highest. But that's the way it is for PVQ currently.

712
01:04:57.705 --> 01:05:00.585
Interesting. Yeah, I'm trying to think of other interesting

713
01:05:00.585 --> 01:05:08.620
things. I mean, there's a bunch of stuff. It just kinda caught me off guard here. So, you agree that the main user behavior right now in PBQ is people

714
01:05:09.500 --> 01:05:13.980
trying to use Clog or whatever with some additional privacy?

715
01:05:14.700 --> 01:05:16.780
I think privacy is a really big use case.

716
01:05:18.765 --> 01:05:21.565
Mixture. I think people do try different models

717
01:05:21.805 --> 01:05:29.724
often. I should come up with that stat. Like, what is the average how many models does the average user try? I bet it's Yeah. I'd be really interested in that.

718
01:05:30.204 --> 01:05:35.980
I think they try the models, but I do think they kind of settle into one or the other after a while. That's what I do anyways.

719
01:05:36.220 --> 01:05:39.580
That's what I do model for this and that. But yeah.

720
01:05:41.020 --> 01:05:49.615
Yeah, it's like I said, OpenClaw. I put this tweet out since OpenClaw came around, our revenue went up 400% in just a few months. Wow.

721
01:05:50.415 --> 01:05:58.655
Yeah. It was a huge, huge boon to us. I also I think I feel a little bit of a plateau coming because as I stated earlier,

722
01:05:59.380 --> 01:06:08.580
people really wanted to get their hands on OpenClaw, but then they start finding out that it's like really hard to get working in the way that you want it to. So they're kind of backing off now.

723
01:06:08.980 --> 01:06:10.340
But I do think that

724
01:06:10.660 --> 01:06:18.575
it's it's gonna be very short lived and like the better OpenClaw or OpenClaw itself is gonna approve. And like, we're gonna all be really

725
01:06:18.655 --> 01:06:21.135
diving into these AI personal assistants.

726
01:06:21.775 --> 01:06:27.215
Yeah. The other thing that kind of throws that stat for Loop is that

727
01:06:28.200 --> 01:06:34.760
the Open Claw is not very efficient. It's very easy to consume. The Open Claw is a great consumer.

728
01:06:34.920 --> 01:06:40.520
So like when I first played around with it, like I said, I just spent a shit ton of money in the beginning.

729
01:06:41.000 --> 01:06:42.840
It was a very expensive lesson.

730
01:06:44.175 --> 01:06:48.655
And then I fine tuned it to a degree and got my spending under control.

731
01:06:49.055 --> 01:06:52.015
Like when you're using like the just a normal chat interface,

732
01:06:52.015 --> 01:07:03.300
maybe hitting just hitting one model directly, you end up use it's way harder to spend as much money as you do with something like an agent just running wild in an inefficient manner. It consumes so hard.

733
01:07:04.260 --> 01:07:16.335
I yeah. My my open claw was very liberal the other day. I was telling it to so we just got our l four zero two endpoints up. And what that means is, like, you can now basically use a curl request

734
01:07:16.414 --> 01:07:22.975
to pay for one image at a time without even having to get an API key from us. But Yep. The agent just wants it.

735
01:07:23.375 --> 01:07:36.500
Yeah. It but the what what was funny is I told it to use one model that cost 29¢. It was like a 29¢ video. And it couldn't get that one to work. So it went with like a $3 model. I was like, I did not tell you to do that.

736
01:07:37.700 --> 01:07:39.860
It just cost him $3 So

737
01:07:40.175 --> 01:07:41.055
yeah,

738
01:07:41.055 --> 01:07:43.775
those are fun things that you learn along the way.

739
01:07:44.974 --> 01:08:08.960
Yeah. I mean, I had I have a dashboard that among other things I used to track all those stats that I read up at the beginning of the the beginning of the show and it updates every five minutes and uses a cheaper model to do that. But at one point, it just I don't know, just fucking crapped out and switched to Opus and was just updating my situation monitor dashboard every five minutes on Opus. By the time I realized that it spent like $250

740
01:08:10.244 --> 01:08:13.445
So there's a lot of those cases I think are happening out there.

741
01:08:13.845 --> 01:08:14.885
But it's interesting.

742
01:08:15.365 --> 01:08:17.844
Go on. So we introduced

743
01:08:18.005 --> 01:08:22.244
spending limits on our API keys. So you're very helpful. Yeah.

744
01:08:23.765 --> 01:08:24.085
I

745
01:08:24.660 --> 01:08:29.220
and that, by the way, that was I was on a flight too. I got off the flight. I was like, what the fuck?

746
01:08:31.220 --> 01:08:35.540
It's interesting that you brought up L four zero two, because that was my last question for you

747
01:08:35.780 --> 01:08:37.540
before we wrap is,

748
01:08:37.940 --> 01:08:39.780
you know, you come from the payment space.

749
01:08:40.695 --> 01:08:47.495
You're obviously also now neck deep in the AI space. The, you know, the hot button topic is agentic payments.

750
01:08:47.655 --> 01:08:51.415
How do you think of that? How is that going to play out in your opinion?

751
01:08:51.655 --> 01:08:52.775
What's your perspective there?

752
01:08:54.000 --> 01:09:02.720
Oh man. I don't know if I have that in grid of an insight. It is it's a lot to wrap your head around. I I implemented the L four zero two personally,

753
01:09:02.720 --> 01:09:10.695
and so like I'm figuring it out and at first sight, it's not a paradigm shift better than just going to PPQ

754
01:09:10.695 --> 01:09:16.375
and getting an API key. Because usually, like, it is a paradigm shift better than going to

755
01:09:16.695 --> 01:09:23.429
some proprietary website, creating a login account, signing up for a subscription, whatever. It's way better than that.

756
01:09:23.750 --> 01:09:32.630
But p b two never had that. Like, you can literally generate an API key just by sending a request, and then you could load money on that API key with another request.

757
01:09:33.815 --> 01:09:36.055
So doing it on a pay per

758
01:09:36.455 --> 01:09:37.655
request model,

759
01:09:38.615 --> 01:09:41.094
it's not that much better. Like but

760
01:09:42.055 --> 01:09:49.709
my partner was kinda mentioning today as I was kinda relaying these thoughts. He's saying, like, the power of it is gonna be when

761
01:09:50.349 --> 01:09:54.110
so when I use an API key on on PBQ,

762
01:09:54.110 --> 01:10:08.114
and generally, like, almost all the time, I know what models I'm the one choosing models. I'm the one in control. Right? But with the l four zero two, like, we're gonna create and maybe you've seen this, but Ryan Gentry released this website called four zero two index.

763
01:10:08.594 --> 01:10:10.594
And it's like thousands of different

764
01:10:10.835 --> 01:10:21.769
endpoints that have different models, different modalities. They do different tasks for you. Right? And so when the agent it's when when the human is no longer choosing the models or choosing

765
01:10:22.410 --> 01:10:26.489
this or that, and it's the agent doing that, then I think maybe

766
01:10:26.650 --> 01:10:37.284
these four zero two payment methods are gonna be coming way more handy. Basically, they're just gonna search a website like four zero two index. They're they're gonna be given a task first. Like, the human gives it a task,

767
01:10:37.684 --> 01:10:40.885
and then they're gonna go and they're gonna look at these thousands of endpoints,

768
01:10:41.445 --> 01:10:56.270
and they're going to solve that task by just, you know, consuming whatever endpoints they want to. So, also another thing that the four zero two payments may solve is like, let's say the future is where these 10,000, 15,000 endpoints exist,

769
01:10:56.510 --> 01:11:03.055
and it's literally with hundreds of different providers, which right now, that's what four zero two index shows. It's like there's almost 300 providers.

770
01:11:03.375 --> 01:11:06.735
You don't want to have a top up model when your agent

771
01:11:07.135 --> 01:11:15.070
is using all these different models, because you don't want to store balance on dozens of different providers, right? You're just leaving money all over the internet.

772
01:11:15.390 --> 01:11:20.030
It's more ideal if you just pay exactly what you need for that thing at that time,

773
01:11:20.430 --> 01:11:23.310
and then go about your business. Right? So that's another

774
01:11:23.645 --> 01:11:32.125
way that these four zero two payments could really come in handy. We will see how how it all develops. We don't know if the future is gonna be like, maybe

775
01:11:32.525 --> 01:11:44.429
maybe all AI really is just gonna come from a handful of companies, and you won't need to, like and and the the downside of leaving money all over the Internet is not as high. It's like like you said, you just go to OpenAI.

776
01:11:44.590 --> 01:12:00.955
That's where all your money is, and it gives you access to all the different tools. Right? But the more beautiful future is where all of these tools are served by hundreds, thousands, tens of thousands of different providers. It's much more democratic that way. Right? And and your your agent is just consuming as necessary.

777
01:12:02.730 --> 01:12:04.090
That makes sense to me.

778
01:12:05.770 --> 01:12:13.690
Yeah. I mean, the agent should just be able to go out there and just natively pay for the data it needs or the access it needs. I think I mean, the better question,

779
01:12:13.930 --> 01:12:15.690
given your background is,

780
01:12:16.435 --> 01:12:19.635
is this something that USD tokens are going to dominate?

781
01:12:19.635 --> 01:12:25.315
Is this I think Stripe released credit cards, like the agent credit cards, or is it Bitcoin?

782
01:12:25.875 --> 01:12:29.475
Obviously, it'll be a mix of the few, but how do you think about that competition

783
01:12:29.990 --> 01:12:38.550
brewing? I mean, even Gentry site, I think lists ones that are using USD tokens. I know OpenRouter, for instance, only accepts USDC

784
01:12:38.710 --> 01:12:39.510
on base. Yeah.

785
01:12:42.015 --> 01:12:47.215
Again, I don't have the yet. I don't have the best opinion on this, but I don't think it's the

786
01:12:47.535 --> 01:12:53.855
like, the the tokens, probably, it will be some USD token of some kind. But I think the protocols,

787
01:12:54.640 --> 01:13:04.720
that's where it's interesting to me. So I've learned a bit about x four zero two. That's that's what the USDC on base is. That's coin bases play for these agentic payments.

788
01:13:05.200 --> 01:13:08.960
I looked into that quite a bit. And I think

789
01:13:10.425 --> 01:13:12.425
there's a lot of things I don't like about it.

790
01:13:13.864 --> 01:13:23.465
It's not private. It's it's it literally is all on a blockchain. So you can literally, like, look up and and use heuristics and this and that to, like, de anonymize what's going on.

791
01:13:25.280 --> 01:13:35.360
Also, the fees are high on x four zero two. I was doing kind of some napkin math the other day. And for every query you send, you pay 2 tenths of 1p

792
01:13:35.520 --> 01:13:52.375
as a the gas fee right now on base. That doesn't sound like a lot, but it actually is a lot if you're if you're query is 1¢ and you're paying 2 tenths of a cent in gas, that's 20% fee. And and a lot of these AI queries really are just 1p. So I don't like x four zero two. And then MPP,

793
01:13:52.535 --> 01:14:00.370
machine payments protocol, that's the one that Stripe released last week. That one is actually interesting to me because it is a blockchain,

794
01:14:00.610 --> 01:14:03.650
but they also support the creation of payment channels.

795
01:14:03.970 --> 01:14:04.690
And so

796
01:14:05.330 --> 01:14:18.864
unlike Bitcoin, a Bitcoin has ten minute block confirmation time. And I think to even spin up a Lightning channel, you need, I think, three or six confirmations for a Lightning channel to be fully open. So you're talking about waiting thirty minutes to sixty minutes plus to to

797
01:14:19.185 --> 01:14:23.105
anyways, you can do a transaction on this MPP thing in under one second,

798
01:14:23.740 --> 01:14:27.020
and then it sets up a channel. You basically have a channel

799
01:14:27.180 --> 01:14:31.820
with the provider, and you can send payments over. But It's like their own version of Lightning.

800
01:14:32.140 --> 01:14:35.580
It's their own version of Lightning, but, yeah, it doesn't have network

801
01:14:35.755 --> 01:14:39.115
I'm sure that it's certainly much more proprietary.

802
01:14:39.115 --> 01:14:48.395
Honestly, I haven't dug in more, but I think yeah. I I still think that the openness of the Lightning Network can can lead to something much greater.

803
01:14:49.910 --> 01:14:52.310
And and so that's where I

804
01:14:52.550 --> 01:15:02.790
would put my money currently. But I do think that the MPP does some cool stuff. Like, I I'm I don't think the x four zero two is gonna do too well, unless they also start

805
01:15:03.065 --> 01:15:04.665
adopting payment channels.

806
01:15:06.505 --> 01:15:09.225
Fair enough. It should be interesting to see how it all plays out.

807
01:15:10.105 --> 01:15:14.985
Matt, this conversation has been great. Before we wrap, what's next for PPQ

808
01:15:14.985 --> 01:15:17.145
and final thoughts for the audience?

809
01:15:18.430 --> 01:15:19.070
Yeah.

810
01:15:19.790 --> 01:15:24.350
Yeah. Thank you for having me on. Again, I I agree it was a great discussion.

811
01:15:26.110 --> 01:15:31.790
There's a whole bunch of stuff coming for PPQ Next. We're just gonna keep grinding on products that people use.

812
01:15:32.430 --> 01:15:32.910
So

813
01:15:33.655 --> 01:15:38.695
I I think that we can we're we're only getting faster with the way we ship.

814
01:15:39.015 --> 01:15:39.575
And,

815
01:15:40.695 --> 01:15:43.255
yeah, many many cool new things in the future.

816
01:15:44.615 --> 01:16:01.950
I would love for us to start creating very specialized tools as well that that are doing really unique things. And that will also help us kind of establish a bit more of a moat other than just kind of being an aggregator of sorts. Yeah. That's pretty much all I have. I do hope, yeah, we can chat again at some point in the future

817
01:16:02.614 --> 01:16:04.375
and just keep wrapping on on a

818
01:16:05.895 --> 01:16:08.454
Yeah. Let's do this yearly cadence. I like it.

819
01:16:08.775 --> 01:16:12.534
Yeah. And I agree on the tools thing. I mean, I think that's where

820
01:16:13.175 --> 01:16:17.710
your model really starts to shine when I have like this multifunction

821
01:16:17.710 --> 01:16:23.229
brain that's just an API key that I can plug into a bunch of these end user tools and have a lot of flexibility.

822
01:16:23.230 --> 01:16:26.590
It gives me superpowers compared to people that are using

823
01:16:26.830 --> 01:16:38.304
it's that's the advantage over people that are using like a cloud subscription, for instance. Yeah. Yeah. Awesome, Matt. This is great. I'll put all relevant links in the show notes, freaks. I'm gonna go to ppq.ai.

824
01:16:38.304 --> 01:16:39.985
Check it out. Play around with it.

825
01:16:40.385 --> 01:16:49.020
Give Matt his feed. Give Matt feedback except for his payment methods. He appreciates it. Thank you for supporting the show, freaks. All relevant links are still dispatch.com.

826
01:16:49.020 --> 01:16:58.460
We're gonna go back to Bitcoin next week. I already have a bunch of shows lined up over the next few weeks, but we're gonna be oscillating basically between Bitcoin, Nostril, and AI if you haven't noticed.

827
01:17:00.625 --> 01:17:01.985
Love it. Thanks, Matt.

828
01:17:02.385 --> 01:17:07.505
Yeah. Thank you. Appreciate it. Thank you, freaks. Stay on the stack sets. Peace.