1
00:00:00,000 --> 00:00:04,300
Maple is really laser focused on how do we help individuals?

2
00:00:04,840 --> 00:00:06,100
How do we help small businesses?

3
00:00:06,640 --> 00:00:11,080
How do we help the individual solopreneur who is trying to make a way

4
00:00:11,080 --> 00:00:15,700
and trying to have a practice that deals with confidential client information?

5
00:00:16,240 --> 00:00:20,140
But they want to protect that and they still want to use the latest and greatest AI tools.

6
00:00:20,580 --> 00:00:22,600
You shouldn't have to sacrifice one of those.

7
00:00:22,680 --> 00:00:23,800
You should be able to have them both.

8
00:00:24,260 --> 00:00:25,640
And that's what Maple is here to do.

9
00:00:26,480 --> 00:00:28,780
Bitcoin is a currency on the internet.

10
00:00:28,780 --> 00:00:31,820
If you don't trust any other currency, this is where you go.

11
00:00:32,020 --> 00:00:33,080
It's going to outperform everything.

12
00:00:33,240 --> 00:00:34,720
Bitcoin could eventually be worthless.

13
00:00:34,920 --> 00:00:36,620
Every day is a good day to buy Bitcoin.

14
00:00:37,160 --> 00:00:39,120
Bitcoin reached its highest price to date.

15
00:00:39,280 --> 00:00:40,780
Bitcoin hitting a new high.

16
00:00:40,880 --> 00:00:42,600
It really has reached escape velocity.

17
00:00:42,720 --> 00:00:43,720
There is no second best.

18
00:00:43,880 --> 00:00:45,040
Who cares about Bitcoin?

19
00:00:45,140 --> 00:00:47,400
It's the only secure database that's ever been invented.

20
00:00:47,940 --> 00:00:48,500
Antisocial.

21
00:00:48,720 --> 00:00:50,300
Bitcoin is the real deal.

22
00:00:50,520 --> 00:00:50,920
Stupid.

23
00:00:51,280 --> 00:00:53,180
Everything is bullish for Bitcoin.

24
00:00:53,400 --> 00:00:53,840
Immoral.

25
00:00:54,120 --> 00:00:55,540
It's going up forever, Lauren.

26
00:00:56,040 --> 00:00:56,880
Fix the money.

27
00:00:57,480 --> 00:00:58,240
Fix the world.

28
00:00:58,780 --> 00:01:28,760
Thank you.

29
00:01:28,780 --> 00:01:36,840
Welcome to Bitcoin for Financial Services Media, the podcast where sound money meets professional services.

30
00:01:37,240 --> 00:01:42,660
We explore how to build the financial services industry within the principles, values, and ethos of Bitcoin.

31
00:01:43,220 --> 00:01:47,700
Today, we're honored to be joined by Mark. He's the CEO and co-founder of Maple AI,

32
00:01:47,960 --> 00:01:52,680
and we'll be exploring private AI and how it is beneficial to the financial services industry.

33
00:01:53,020 --> 00:01:53,760
Let's get started.

34
00:01:53,760 --> 00:02:01,800
All right, welcome back to another episode of the Bitcoin for Financial Services podcast.

35
00:02:02,300 --> 00:02:04,000
Man, today is going to be such a fun episode.

36
00:02:04,460 --> 00:02:06,200
Today we have Mark Suman.

37
00:02:06,660 --> 00:02:10,120
You might know him as MarksOnline of Maple AI.

38
00:02:10,340 --> 00:02:15,000
He works on a lot of different things, honestly, so we'll get into the background here soon.

39
00:02:15,460 --> 00:02:19,320
Ex-Apple, just, I mean, just the resume is very thick.

40
00:02:19,900 --> 00:02:23,000
So thank you so much, Mark, for being with us today.

41
00:02:23,000 --> 00:02:27,180
and we're excited to go deep on AI and financial services.

42
00:02:28,400 --> 00:02:30,020
Yeah, guys, thanks for having me on. I appreciate it.

43
00:02:31,080 --> 00:02:33,160
Yeah. So kick us off.

44
00:02:33,240 --> 00:02:38,520
I know you just have a background in Bitcoin and in tech.

45
00:02:38,880 --> 00:02:41,680
Can you just give us the quick timeline

46
00:02:41,680 --> 00:02:43,940
of what you've worked on over the last decade?

47
00:02:45,040 --> 00:02:46,440
Sure, yeah, quick timeline.

48
00:02:46,760 --> 00:02:48,080
Okay. Oh, a decade.

49
00:02:48,340 --> 00:02:49,480
I can't go back to the beginning.

50
00:02:49,620 --> 00:02:50,460
That's how old I am.

51
00:02:50,460 --> 00:03:15,560
Okay. So a decade ago, I was working on education technology software. So we set out to revolutionize how students are interacting with their professors in the classroom and universities. So I was at a small startup called Instructure Canvas, which pretty much anybody who's gone to university in the United States over the last 10 years has probably used Canvas at this point.

52
00:03:15,560 --> 00:03:29,400
So I was doing their, I started their mobile development team. We build iPhone apps, iPad apps, Android, all this stuff. And then after that, I moved into a really small startup that was building software for kids and speech therapy.

53
00:03:29,400 --> 00:03:40,800
And so we were doing these iPad apps where they would, you know, if they had a stutter or they couldn't pronounce one of their like the R sound or something, then they would they would listen to how it's supposed to sound.

54
00:03:40,920 --> 00:03:44,980
And then they would record themselves saying it and kind of play it back and compare the two.

55
00:03:45,380 --> 00:03:47,220
And they worked in conjunction with a therapist.

56
00:03:47,980 --> 00:03:53,940
After that, I got a job at Apple working specifically in AI ML technology.

57
00:03:54,520 --> 00:03:56,780
And we had a really strong privacy focus.

58
00:03:57,340 --> 00:04:02,000
Apple does a big public game about privacy, but they actually do care about it internally too.

59
00:04:02,620 --> 00:04:06,560
So I spent six years working on this brand new thing.

60
00:04:06,660 --> 00:04:13,980
It's only an internal project, but it was a very difficult technical challenge to solve

61
00:04:13,980 --> 00:04:16,240
because we had to have privacy from the beginning.

62
00:04:17,560 --> 00:04:22,080
And so when I finally decided to leave Apple and go off and now we're building Maple,

63
00:04:22,500 --> 00:04:24,740
privacy was just like this huge thing on my mind.

64
00:04:24,840 --> 00:04:26,380
It has been since the beginning of my career.

65
00:04:26,780 --> 00:04:33,660
And so, yeah, I left Apple, joined up with the Mutiny guys, building Mutiny Wallet.

66
00:04:34,500 --> 00:04:37,140
We were trying to scale that to millions of users.

67
00:04:37,880 --> 00:04:41,060
But for lots of reasons, we decided to wind down the wallet.

68
00:04:41,180 --> 00:04:42,420
We thought that was the best approach.

69
00:04:42,740 --> 00:04:45,540
But we had this really strong privacy angle with the Mutiny Wallet.

70
00:04:45,940 --> 00:04:49,820
And we applied it to the next great thing that was coming out, and that was AI.

71
00:04:50,620 --> 00:04:51,600
And here we are today.

72
00:04:51,600 --> 00:04:57,060
we've built the privacy version of ChanchiPT is like the shortest way to sum it up.

73
00:04:58,440 --> 00:05:03,440
That's a noble pass, Mark, helping the kids much more noble than Jordan and I.

74
00:05:03,660 --> 00:05:05,040
So tip of the cap to you there.

75
00:05:07,780 --> 00:05:13,120
When you were working at Apple, were they still calling AI like machine learning back then?

76
00:05:13,200 --> 00:05:13,460
Yeah.

77
00:05:13,740 --> 00:05:14,640
Yeah, definitely.

78
00:05:14,820 --> 00:05:15,620
It was machine learning.

79
00:05:16,140 --> 00:05:19,320
And like my first day that I sat down, my boss said, hey, I want you to build this machine

80
00:05:19,320 --> 00:05:24,180
learning app and like gave me all these details and stuff right and then over time it started to

81
00:05:24,180 --> 00:05:31,720
morph into ai machine learning ai ml but now you know once we had the chat gpt moment in 2022 23

82
00:05:31,720 --> 00:05:38,800
whenever it was uh it's just been ai everywhere now yeah that's fascinating would you mind just

83
00:05:38,800 --> 00:05:44,060
kind of giving us a brief background and kind of how that industry has evolved and then current day

84
00:05:44,060 --> 00:05:51,120
where it stands on what your actual privileges as a user are or are not.

85
00:05:51,520 --> 00:05:57,460
I'm pretty sure the government can subpoena ChatGPT for your chat history

86
00:05:57,460 --> 00:05:59,140
and curie your information.

87
00:05:59,460 --> 00:06:02,280
So why don't you just kind of bring us up to modern day

88
00:06:02,280 --> 00:06:03,920
and what our user privileges are?

89
00:06:04,700 --> 00:06:05,280
Sure. Yeah.

90
00:06:05,400 --> 00:06:10,380
So prior to this big AI thing that everybody knows now of ChatGPT and such,

91
00:06:10,780 --> 00:06:12,140
prior to that, it was all machine learning,

92
00:06:12,140 --> 00:06:16,760
which is really just a lot of, it's a lot of formulas. It's a lot of regressions, you know,

93
00:06:16,800 --> 00:06:21,100
statistical analysis, that kind of stuff. And so people were used to that being applied to their

94
00:06:21,100 --> 00:06:27,080
email for their spam filter or to their algorithm when they go onto Facebook or their Instagram feed.

95
00:06:27,480 --> 00:06:32,260
Machine learning is like learning from your habits on social media and then curating your feed.

96
00:06:32,620 --> 00:06:40,100
So it's like very targeted. And then ChatGPT released this large language model, this LLM

97
00:06:40,100 --> 00:06:45,280
on the scene. And suddenly it was less about these specific formulas going after one thing

98
00:06:45,280 --> 00:06:52,340
and more about how do we kind of map the human brain in a neural network. And so now you can

99
00:06:52,340 --> 00:06:58,420
type text to it and then it does probability analysis and decides, okay, I'm going to write

100
00:06:58,420 --> 00:07:03,040
a sentence back to this human. What's the most probable next word that should appear in the

101
00:07:03,040 --> 00:07:08,400
sentence? And so it will write back to you and give you a response. To us, it just looks like

102
00:07:08,400 --> 00:07:11,740
we're talking to somebody, right? It looks like we're text messaging with someone, but really it's

103
00:07:11,740 --> 00:07:17,220
just doing a bunch of probability analysis. And so ChatGPT put that out there. A lot of people

104
00:07:17,220 --> 00:07:23,080
started interacting with it. ChatGPT 2 and then 3, 3.5, 4, and then Anthropic is another big one

105
00:07:23,080 --> 00:07:27,600
that came out. They have Claude. A lot of people have probably heard of Claude. And then you also

106
00:07:27,600 --> 00:07:34,180
have XAI, which is Twitter or Grok. There's a lot of names thrown around, but basically Elon Musk's

107
00:07:34,180 --> 00:07:39,980
is Grok. And those are the main ones that most people use. ChatGPT, Claude, or Grok.

108
00:07:40,740 --> 00:07:46,280
And then there's a ton of other smaller ones. Poor Meta, Facebook. I don't feel bad for them at all,

109
00:07:46,280 --> 00:07:52,400
but they had Llama that came out early on. And people just don't really take it that seriously

110
00:07:52,400 --> 00:07:58,420
anymore. Sorry if they're listening to this. We do have Llama. Llama 3.3 is a good LLM, but

111
00:07:58,420 --> 00:08:01,520
they're struggling to jump up ahead.

112
00:08:02,020 --> 00:08:05,920
So you asked about what are our rights and our privacy and all that stuff, security.

113
00:08:06,900 --> 00:08:11,200
Anything that you, if you type into someone else's website or app,

114
00:08:11,480 --> 00:08:12,900
something like ChatGPT or Claude,

115
00:08:13,320 --> 00:08:16,400
all of that information goes to their servers, goes to their computers.

116
00:08:17,100 --> 00:08:19,520
And then being an employee at that company,

117
00:08:20,100 --> 00:08:24,760
you potentially have escalated privileges where you can see all the information coming in

118
00:08:24,760 --> 00:08:28,380
for debug reasons, for looking at how can we improve the product,

119
00:08:28,420 --> 00:08:42,080
And then also they keep that data as well for training new models. So they try to create these new brains, these new LLM brains, right? And they take all this new data coming in and they feed it into that training to try and make it smarter for everyone.

120
00:08:42,720 --> 00:08:45,840
So that's where you hear a lot of people say, I don't want you training off my data.

121
00:08:45,840 --> 00:09:10,640
Now, you mentioned subpoenas and courts. Definitely any data that's on someone else's server, the court can come with an order and get access to that information. That's true of any cloud software out there. Your Gmail, your social media feed, if you have notes that you're storing in Notion, any of these things, they are not subpoena-proof. They can go grab them if they have a warrant.

122
00:09:10,640 --> 00:09:28,380
And so, yeah, I've talked to a lot of lawyers who their bar associations simply have rules that say you cannot upload any of your legal documents that you're working in your casework up to Chachapiti because that suddenly becomes available for anyone to get at with a court order.

123
00:09:28,900 --> 00:09:32,200
So they have to find other ways around that to kind of work around those rules.

124
00:09:33,440 --> 00:09:37,900
I don't know. I feel like I covered a lot. Is there anything specific that you're curious about there?

125
00:09:37,900 --> 00:09:45,600
yeah i think that's a great foundation setter what do you think jordan yeah i i think it's uh

126
00:09:45,600 --> 00:09:50,920
it's definitely super interesting because you can i think a lot of people in accounting tax

127
00:09:50,920 --> 00:09:57,220
financial services just generally are looking at these tools and they're saying wow this is this

128
00:09:57,220 --> 00:10:04,000
is really powerful and and especially if you're on the cutting edge of hey i'm using like at least

129
00:10:04,000 --> 00:10:09,440
on our team like i know people started out on chat gbt and you you mentioned anthropic i think

130
00:10:09,440 --> 00:10:17,400
there's been a huge shift now away from chat and over to anthropic um and i kind of see and you

131
00:10:17,400 --> 00:10:23,620
correct me if i'm wrong on any of this i kind of see it as like the these powerful like the sonnet

132
00:10:23,620 --> 00:10:32,500
um and the uh what's the 4.5 opus opus yeah you know those models being able to plug those in

133
00:10:32,500 --> 00:10:42,860
with like non-PII data that you're just trying to get tasks done more efficiently, it's super

134
00:10:42,860 --> 00:10:48,980
powerful. And then all of this like Claudebot, Moldbook, all the things that have popped up

135
00:10:48,980 --> 00:10:53,900
literally in the last week, it's all very interesting. But I'll just say like, I deleted

136
00:10:53,900 --> 00:11:00,020
Claudebot off of my Mac because I was just like, whoa, I don't know about all this. I feel,

137
00:11:00,020 --> 00:11:07,320
uh you know i read some articles that okay maybe it installs malware on your computer so that if

138
00:11:07,320 --> 00:11:11,980
you piss it off if you piss your agent off you know and it's it has a soul and all this stuff

139
00:11:11,980 --> 00:11:17,560
right that it might turn on you and you might think you're deleting it but someone was like you

140
00:11:17,560 --> 00:11:24,240
could or it could uh install something a backdoor on your machine that you think it's gone but it

141
00:11:24,240 --> 00:11:30,720
can come back. And I think for a lot of financial service people, they're just like, look, I

142
00:11:30,720 --> 00:11:38,480
understand tech to a degree, but I'm not a developer. And I don't have this full infrastructure

143
00:11:38,480 --> 00:11:46,580
background. And so I would say it seems like a lot of people should be probably pumping the brakes

144
00:11:46,580 --> 00:11:54,100
in some ways. But then in other ways, it's like, if you miss this boat, you might not have a

145
00:11:54,100 --> 00:11:58,460
firm in a year, you know what I'm saying? Because things are just moving so quickly.

146
00:11:59,260 --> 00:12:05,980
So I guess, I know I threw a lot out of it from that standpoint, but I guess lay the groundwork

147
00:12:05,980 --> 00:12:13,200
of like a financial advisor, a tax practitioner who's having, you know, sensitive conversations

148
00:12:13,200 --> 00:12:18,800
that now are, most of them are being captured by some kind of meeting recorder. They also have

149
00:12:18,800 --> 00:12:24,140
documents and then they have their workflows and softwares that need to talk to each other

150
00:12:24,140 --> 00:12:29,080
i'm just kind of curious if you can help i know we want to treat this kind of like a consultation

151
00:12:29,080 --> 00:12:36,620
right help us think through how do all the pieces fit together um and what what might you use clod

152
00:12:36,620 --> 00:12:44,720
for uh one thing but then a more secure ai or you all thinking no you should use maple for everything

153
00:12:44,720 --> 00:12:51,420
And we're going to make sure that the models are as powerful as any other cloud-based LLM tool.

154
00:12:52,480 --> 00:12:52,500
Yeah.

155
00:12:53,540 --> 00:12:58,800
I'll start off by saying this whole topic can be super overwhelming for people.

156
00:12:59,280 --> 00:13:02,940
I know that there are a lot of people who have a podcast they listen to.

157
00:13:03,120 --> 00:13:07,060
And then when one of the AI episodes comes up, I've heard this from people.

158
00:13:07,160 --> 00:13:09,060
They're like, oh, no, not them.

159
00:13:09,160 --> 00:13:10,640
Now they're talking about AI too, right?

160
00:13:11,540 --> 00:13:14,020
So I totally understand.

161
00:13:14,020 --> 00:13:20,320
for this one yeah yeah no i i totally understand like people feeling overwhelmed by it because it

162
00:13:20,320 --> 00:13:25,820
just feels like it's coming at them from every angle um and everything's moving so fast so like

163
00:13:25,820 --> 00:13:30,460
some people not everybody loves to tinker with new technology and play with it and so some people

164
00:13:30,460 --> 00:13:35,140
love to sit back and watch but this one is moving so fast that you sit back and watch like it's it's

165
00:13:35,140 --> 00:13:42,900
just taking off like a race car um so i would i would say uh you know just take take a first step

166
00:13:42,900 --> 00:13:49,620
try something out, right? Now, you mentioned installing Clodbot on your computer. I definitely

167
00:13:49,620 --> 00:13:55,800
think be safe when you tinker with some of these things like Clodbot. I would recommend installing

168
00:13:55,800 --> 00:13:59,360
it somewhere else and playing with it there. So it's good that you remove it off your computer

169
00:13:59,360 --> 00:14:06,540
because we just don't know what these things are going to do. Now, that said, definitely don't be

170
00:14:06,540 --> 00:14:12,780
shy of AI tools. So dive in and try some of them out. So if ChachiBT is the most familiar or

171
00:14:12,780 --> 00:14:18,000
Claude is the most familiar, just jump in and start typing to it. Don't give it tons of personal

172
00:14:18,000 --> 00:14:23,040
information. Don't give it anything super, you know, important or sensitive, but give it, give

173
00:14:23,040 --> 00:14:28,020
it kind of like treated. I tell people treat it almost as if it's Wikipedia, but you can interact

174
00:14:28,020 --> 00:14:32,460
with Wikipedia. That's where I is a great place for people to start. So like, okay, I want to look

175
00:14:32,460 --> 00:14:36,900
at some information about someone. And it's like, boom, here's a basic Wikipedia article about that

176
00:14:36,900 --> 00:14:41,840
person, a famous person. And then you're like, okay, now ask follow-up questions and dig deeper

177
00:14:41,840 --> 00:14:43,320
than you could with a Wikipedia page.

178
00:14:43,360 --> 00:14:44,120
And they start to understand

179
00:14:44,120 --> 00:14:45,320
what this interaction is like.

180
00:14:45,880 --> 00:14:48,080
Then you can kind of move beyond that and say,

181
00:14:48,160 --> 00:14:50,620
okay, now I want to talk about my profession.

182
00:14:50,900 --> 00:14:52,660
So I'm doing tax, I'm a CPA.

183
00:14:53,620 --> 00:14:54,700
How can I use this tool?

184
00:14:54,860 --> 00:14:56,260
So now you can start digging into

185
00:14:56,260 --> 00:14:57,960
what are kind of maybe the latest,

186
00:14:58,140 --> 00:15:00,780
you know, the latest tax code that came out

187
00:15:00,780 --> 00:15:02,540
or what is this thing that I'm working on?

188
00:15:02,580 --> 00:15:03,500
But again, you're not putting

189
00:15:03,500 --> 00:15:05,200
client information in there yet.

190
00:15:05,260 --> 00:15:06,020
Just keep it high level.

191
00:15:06,580 --> 00:15:08,900
That if this, always treat this,

192
00:15:09,160 --> 00:15:10,820
that when you're talking to ChatGPT,

193
00:15:10,820 --> 00:15:13,940
this might show up on your Twitter feed.

194
00:15:14,040 --> 00:15:16,080
This might show up on Google.

195
00:15:16,220 --> 00:15:18,040
It might show up somewhere publicly on the internet.

196
00:15:18,340 --> 00:15:19,360
Just treat it like that.

197
00:15:19,420 --> 00:15:21,080
Like anything you say there might be read

198
00:15:21,080 --> 00:15:22,380
by every single person in the world.

199
00:15:22,840 --> 00:15:25,620
That is the threat model that you need to keep in mind

200
00:15:25,620 --> 00:15:26,600
when you're using those tools

201
00:15:26,600 --> 00:15:27,840
because it's already happened.

202
00:15:28,320 --> 00:15:30,320
I'm not just like, you know,

203
00:15:30,360 --> 00:15:32,760
I'm not out there just throwing theories around.

204
00:15:33,220 --> 00:15:36,880
Both ChatGPT and Grok already had a problem

205
00:15:36,880 --> 00:15:39,520
where they were accidentally putting people's chats

206
00:15:39,520 --> 00:15:40,840
on Google search results.

207
00:15:40,920 --> 00:15:41,900
So when you were searching for stuff,

208
00:15:41,960 --> 00:15:43,200
you were finding conversations

209
00:15:43,200 --> 00:15:43,980
that people have been having.

210
00:15:44,900 --> 00:15:46,240
Meta had a similar problem.

211
00:15:47,020 --> 00:15:49,060
So just treat it that way.

212
00:15:50,140 --> 00:15:51,140
Now, one thing you can do,

213
00:15:51,220 --> 00:15:52,600
I've talked to some financial professionals,

214
00:15:53,080 --> 00:15:56,140
they take these personal client documents

215
00:15:56,140 --> 00:15:56,800
that they have,

216
00:15:57,120 --> 00:15:58,700
and they will basically scrub,

217
00:15:58,700 --> 00:16:00,340
they'll redact all of the names

218
00:16:00,340 --> 00:16:01,680
and information from it.

219
00:16:01,940 --> 00:16:14,303
Then they put in a chat GPT and get information back And then they go insert back in the relevant information And that totally fine Because again if you not worried about that document leaking if it been

220
00:16:14,303 --> 00:16:19,143
redacted, the trouble is you're spending all that time on the front end and the back end, right,

221
00:16:19,283 --> 00:16:24,163
scrubbing it, and you're still not getting the best results. Because the way that AI works,

222
00:16:24,603 --> 00:16:29,783
AI really is the best when it has all of the details, it needs as high fidelity as possible.

223
00:16:30,323 --> 00:16:40,503
And so if you are removing some of the details from it and then saying, hey, I want you to act on this, you're asking it to like act with with one eye closed, one arm tied behind its back, maybe bouncing on its left foot or something.

224
00:16:41,043 --> 00:16:43,023
And it's really best when you can give it everything.

225
00:16:43,843 --> 00:16:46,063
And so you brought up Maple.

226
00:16:46,283 --> 00:16:51,043
That's one of our approaches with Maple is really we've created this end to end encrypted environment.

227
00:16:51,043 --> 00:16:52,443
It's a confidential environment.

228
00:16:52,863 --> 00:16:57,983
So it's as if you went into a room with the AI and you close the door.

229
00:16:57,983 --> 00:17:07,163
You know, you're in that conference room at your client that you're working at. You close the door of the conference room and now you guys have four walls around you and you can totally dive into all the files.

230
00:17:07,283 --> 00:17:15,503
You can dive into the entire conversation. You can really get detailed. And knowing that when we open the door and leave the room, none of that is leaving the room with you.

231
00:17:15,583 --> 00:17:19,583
It's all staying there in that secure enclave. And that's how we've built Maple.

232
00:17:19,583 --> 00:17:24,063
So we, you know, we love that the tool is there for people who need it.

233
00:17:25,103 --> 00:17:27,683
Claude and ChatGPT, they have much larger budgets.

234
00:17:27,863 --> 00:17:30,963
And so they're always coming out with new features that we just can't keep up with.

235
00:17:31,863 --> 00:17:34,223
So I tell people, like, use the right tool for the job.

236
00:17:34,823 --> 00:17:40,243
And so if we don't have a feature that you really need, find ways to use ChatGPT.

237
00:17:40,443 --> 00:17:44,103
We have people who use Maple to say, hey, I'm going to put this into ChatGPT.

238
00:17:44,103 --> 00:17:46,543
Can you redact this stuff for me?

239
00:17:46,543 --> 00:17:53,763
Or can you make this conversation more private or not private, but, you know, scrub so that when I talk to Chad GPT, I'm not giving away information.

240
00:17:54,383 --> 00:17:57,103
There are ways you can kind of get these tools to talk to each other.

241
00:17:57,503 --> 00:18:01,023
But that's kind of the lay of the landscape of these different tools.

242
00:18:02,083 --> 00:18:05,743
So I guess I'll kind of pause there and we can go where you want to.

243
00:18:09,103 --> 00:18:11,183
Your wealth deserves intention.

244
00:18:11,183 --> 00:18:17,083
Basilic Financial is a Bitcoin-focused wealth management firm built to help our clients find balance between money and meaning.

245
00:18:17,483 --> 00:18:20,823
We integrate sound money strategies across four distinct personalized services.

246
00:18:21,223 --> 00:18:26,723
Financial planning, individual wealth management, retirement services for SMBs, and Bitcoin financial services.

247
00:18:27,163 --> 00:18:31,243
Your prosperity, your purpose, intentionally guided at Basilic Financial.

248
00:18:31,663 --> 00:18:35,003
Visit our website at basilic.io to schedule a free consultation.

249
00:18:35,003 --> 00:19:03,263
Would you say from the attorney conversations you've had, a great use case for us is, can I just upload an unredacted 1040 tax form to Maple and say, OCR PDF this and pull down the name, social address, put those in black boxes, and then you all give that back to me as something?

250
00:19:03,263 --> 00:19:09,463
because I had a lot of trouble. I was trying to get the local llama to do that on a Mac Studio,

251
00:19:09,643 --> 00:19:14,323
and it was just having a lot of trouble. So I'm curious if it really is as simple as that,

252
00:19:14,343 --> 00:19:20,143
or if there's still privacy concerns with loading something like that into even a tool like Maple.

253
00:19:21,123 --> 00:19:25,163
You definitely can do that. That's one of our big use cases. So you can upload a PDF to it.

254
00:19:25,503 --> 00:19:30,223
It can grab all that information out of that for you. So absolutely, that's one of the features

255
00:19:30,223 --> 00:19:30,983
that we've built for.

256
00:19:31,483 --> 00:19:33,683
We do have a small technical limitation right now,

257
00:19:33,743 --> 00:19:35,063
but we'll overcome it in the future.

258
00:19:35,543 --> 00:19:38,003
That is, the PDF has to be like an actual text PDF.

259
00:19:38,443 --> 00:19:40,783
Some people will scan their documents, right,

260
00:19:40,843 --> 00:19:41,923
and then scan it to a PDF,

261
00:19:42,043 --> 00:19:43,623
but really it's just a PDF of a bunch of images.

262
00:19:44,283 --> 00:19:47,503
So we're working on improving that

263
00:19:47,503 --> 00:19:49,703
so we can deal with those image-based PDFs better.

264
00:19:50,923 --> 00:19:52,003
We'll get over that hump.

265
00:19:52,263 --> 00:19:54,803
So effectively, yes, you can upload a PDF

266
00:19:54,803 --> 00:19:56,623
and it can grab everything out of it for you.

267
00:19:56,743 --> 00:19:59,623
So I've got my 1040s and I can upload them into Maple.

268
00:19:59,623 --> 00:20:05,503
I can ask questions about it before I go talk to my accountant and I can come prepared to the conversation with a little bit more information.

269
00:20:06,163 --> 00:20:12,143
And then when it comes time for me to kind of sign and say, yes, these are the taxes I'm going to pay this year.

270
00:20:12,263 --> 00:20:15,183
I'll upload it to Maple and be like, OK, can you do one more check?

271
00:20:15,483 --> 00:20:18,303
You know, I know my account already checked everything. I visually scanned it.

272
00:20:18,883 --> 00:20:23,603
But please, please, like red flag anything for me that I should dig into more.

273
00:20:24,243 --> 00:20:27,803
Here are, you know, here are the charitable contributions I usually make in a year.

274
00:20:27,803 --> 00:20:29,263
Here are some other things about me.

275
00:20:29,263 --> 00:20:33,603
Can you just kind of be my extra eyes and ears and kind of prove this for me?

276
00:20:33,703 --> 00:20:38,683
So that's how I use it on a personal level with my personal finances when working with my accountant.

277
00:20:40,603 --> 00:20:51,303
And Mark, could you just kind of dig into mechanically how Maple is able to close the door and what you guys are doing versus what the other LLMs are doing?

278
00:20:52,223 --> 00:20:52,483
Sure.

279
00:20:52,823 --> 00:20:56,563
We use something called Trusted Execution Environments or TEEs.

280
00:20:57,143 --> 00:20:58,263
There's a lot of words you'll hear.

281
00:20:58,263 --> 00:21:02,883
So TEE is one of them or confidential computing or secure enclaves.

282
00:21:03,023 --> 00:21:04,983
They're all really the same thing.

283
00:21:05,403 --> 00:21:09,843
A secure enclave is something that most people interact with, but don't realize it's on their phone.

284
00:21:10,003 --> 00:21:19,623
If you use face ID on your phone or use a fingerprint or you have a wallet on your phone, that's using this secure chip that can house all that information.

285
00:21:20,443 --> 00:21:26,503
But humans can't inspect the software that's running inside that enclave.

286
00:21:27,203 --> 00:21:30,283
So that's effectively what we use now, but it's in the cloud.

287
00:21:30,403 --> 00:21:33,883
So these cloud servers, it's come on the scene in the last five or six years.

288
00:21:34,243 --> 00:21:36,103
These cloud servers have these secure enclaves.

289
00:21:36,683 --> 00:21:39,183
And so we publish our code online.

290
00:21:39,363 --> 00:21:42,243
If you go on to GitHub, you can see all the Maple code.

291
00:21:42,863 --> 00:21:47,343
And then the promise that the secure enclaves make, and this is a hardware encrypted promise.

292
00:21:47,463 --> 00:21:48,503
It's a cryptographic promise.

293
00:21:48,503 --> 00:21:51,263
So it's not just us saying we promised this.

294
00:21:51,263 --> 00:21:58,243
It's an actual cryptographic proof that we upload all of that GitHub code into the secure enclave.

295
00:21:58,543 --> 00:22:09,623
And then it gives you a mathematical proof, a checkmark, if you will, that shows that the code inside the enclave is the same that is online, free and open source for people to view.

296
00:22:10,083 --> 00:22:11,263
And that's the real unlock.

297
00:22:12,003 --> 00:22:19,743
That's the important piece that anybody can audit the code and see, OK, the Maple team is not keeping a copy of all this data.

298
00:22:19,743 --> 00:22:23,923
The Maple team does not have a data sharing agreement with the United States government.

299
00:22:24,363 --> 00:22:26,843
The Maple team, you know, does this, this, not that, right?

300
00:22:28,163 --> 00:22:34,583
Whereas these other services, so if you look at just ChatGPT and Claude, they have all your data.

301
00:22:34,923 --> 00:22:36,803
So, I mean, that's already out the window.

302
00:22:37,303 --> 00:22:40,483
But then you have these other services that are like AI proxies.

303
00:22:40,983 --> 00:22:42,483
And these are ones like Venice.

304
00:22:42,863 --> 00:22:48,363
Venice is a proxy where they run these open source models, but they also let you talk to ChatGPT.

305
00:22:48,363 --> 00:22:54,503
but with venice you are still sending all your information to them to their servers and then they

306
00:22:54,503 --> 00:22:58,163
just promise that they're not going to look at it so this is more like a vpn if you've ever used a

307
00:22:58,163 --> 00:23:02,443
vpn for your internet um if you've like wanted to watch something on netflix when you're in a

308
00:23:02,443 --> 00:23:07,963
different country or something so you vpn and you get into another country um that's all a vpn is

309
00:23:07,963 --> 00:23:11,703
doing it knows all of your internet traffic and they're just promising as a company that they're

310
00:23:11,703 --> 00:23:16,123
not going to keep track of what you're doing um so does something to keep in mind if you use ai

311
00:23:16,123 --> 00:23:21,003
proxies is you're still sending sensitive data to another company and you don't know what server

312
00:23:21,003 --> 00:23:24,843
what code they're running on the servers you don't know what they're doing with it all you have is

313
00:23:24,843 --> 00:23:28,523
just their promise that they're not going to do something with it and they there's there's been

314
00:23:28,523 --> 00:23:34,043
plenty of vpn providers that have been caught later on when a court comes to them and says hey do you

315
00:23:34,043 --> 00:23:37,883
have all these records of what people are doing on the internet this whole time publicly they've been

316
00:23:37,883 --> 00:23:41,883
saying no but privately they've been keeping a stash of it and then the the courts come and take

317
00:23:41,883 --> 00:23:45,563
it so you just have to keep that in mind when you're using some other services whereas with

318
00:23:45,563 --> 00:23:48,703
With Maple, we just give you that cryptographic promise that anybody can audit.

319
00:23:48,983 --> 00:23:57,443
If you're a firm, like an accounting firm or a financial services firm, and you want to get your own security on it, you can give them our open source code.

320
00:23:57,543 --> 00:23:58,803
They can go comb through it.

321
00:23:58,883 --> 00:24:01,883
They can audit the cryptographic attestations that we provide.

322
00:24:02,203 --> 00:24:06,883
And then they can give you an assurance that, yes, this is working the way that they say it is.

323
00:24:08,163 --> 00:24:15,143
No surprise there, at least for me, maybe Jordan's surprised, but that the Trust Me Bro AI company also has a token.

324
00:24:15,143 --> 00:24:26,343
that this seems pretty like hand but whatever um so marks i'm curious you know as these you know

325
00:24:26,343 --> 00:24:32,383
ai tools you know loms are developing so it sounds like you guys are doing you know kind of like more

326
00:24:32,383 --> 00:24:39,843
bitcoin or way you know don't trust verify sort of thing but then you know i'm seeing like i can't

327
00:24:39,843 --> 00:24:45,223
remember, you know, my lawyer uses an AI and it's like compliant for whatever they have to do. It's

328
00:24:45,223 --> 00:24:50,043
like Steve or something. I don't think that's his actual name. And then I'm moving up, I'm moving

329
00:24:50,043 --> 00:24:57,483
into the ClickUp ecosystem for a lot of my, you know, CRM project management tools, and they have

330
00:24:57,483 --> 00:25:02,543
a, you know, native AI and I just pulled up, you know, their website. So it's, you know, GDR compliant,

331
00:25:02,543 --> 00:25:10,743
iso compliant hipaa compliant aicpa sock 2 compliant so when you have different tools like

332
00:25:10,743 --> 00:25:17,923
this offering services to professionals do they have to get all these certifications because you

333
00:25:17,923 --> 00:25:22,843
know they are storing you know information on a server so we can look at it and then these are

334
00:25:22,843 --> 00:25:29,143
just like extra badges or whatever to help build trust and you know kind of facilitate or

335
00:25:29,143 --> 00:25:35,463
give the user the confidence that they're not going to be sharing information and your guys'

336
00:25:35,563 --> 00:25:39,963
model doesn't require that? Or how do you kind of view these two things?

337
00:25:41,183 --> 00:25:46,983
We definitely get questions about certifications a lot. And it's really something that we should go

338
00:25:46,983 --> 00:25:51,583
do. We've just been a smaller company. And so we haven't gone after those yet because it costs

339
00:25:51,583 --> 00:25:57,903
money to go get everybody to come in, audit your code and certify it. That said, we're already

340
00:25:57,903 --> 00:26:01,843
extremely confident we would pass all the certifications because we've built it from a

341
00:26:01,843 --> 00:26:07,803
technical ground up, right? The purpose of a lot of these certifications and SOC 2 and all that stuff

342
00:26:07,803 --> 00:26:12,883
is that they know that you're going to be handling people's data and holding on to their data. And so

343
00:26:12,883 --> 00:26:17,703
you need to set up assurances of like retention windows of how long you're going to keep stuff.

344
00:26:17,703 --> 00:26:22,563
And then when people request to have it deleted, you need to go into all those backups and all these

345
00:26:22,563 --> 00:26:28,203
places you've kept their data and make sure you delete it. When you approach it from first

346
00:26:28,203 --> 00:26:33,623
principles and saying, we don't have access to our user data, like our users, we can't see it

347
00:26:33,623 --> 00:26:38,403
because it's all encrypted away from us, that changes the entire game. And so really any kind

348
00:26:38,403 --> 00:26:44,583
of compliance we're going to go after should be very simple to get done with very few changes.

349
00:26:44,843 --> 00:26:49,803
There might be some changes on, you know, we might have to have audit logs for our laptops

350
00:26:49,803 --> 00:26:54,783
or something. I mean, there might be little like checkboxes we have to do, but for the most part,

351
00:26:54,983 --> 00:26:59,963
we already could be compliant with pretty much all those things you just rattled off.

352
00:27:00,423 --> 00:27:05,743
We just have to go through the hoops and pay people to draft that paperwork and give us a

353
00:27:05,743 --> 00:27:12,043
fancy badge. So we might need to go after that soon as we keep taking on more financial and legal

354
00:27:12,043 --> 00:27:18,843
customers and medical customers. We have quite a few doctors, therapists, nurses that are using

355
00:27:18,843 --> 00:27:27,523
maple for stuff too so i guess my question i've gone down this rabbit hole with you know everyone's

356
00:27:27,523 --> 00:27:34,883
talking about the mac mini we have been playing around within the firm with a mac studio as like

357
00:27:34,883 --> 00:27:40,143
potentially having that be the brain i just want to get your rough thoughts on this right since

358
00:27:40,143 --> 00:27:48,323
this is a consultation we said yes um so what are your thoughts on having a a bit beefy like

359
00:27:48,323 --> 00:27:56,103
every cpa firm potentially in the country running their own running their own uh models

360
00:27:56,103 --> 00:28:04,103
on a a powerful enough machine that could use low like do you think this is overkill with the maple

361
00:28:04,103 --> 00:28:11,203
uh like um value proposition that's already out there where you use the the local models

362
00:28:11,203 --> 00:28:14,843
on PII data that's stored locally on a machine.

363
00:28:15,883 --> 00:28:22,183
And then you have it interact with the cloud-based APIs

364
00:28:22,183 --> 00:28:25,283
for things that are like, hey, it doesn't matter.

365
00:28:25,503 --> 00:28:27,963
Like it just, it took the 1040,

366
00:28:28,123 --> 00:28:31,203
it pulled the JSON that just gave the relevant context.

367
00:28:31,623 --> 00:28:33,263
Maybe it pointed to the meeting notes

368
00:28:33,263 --> 00:28:34,443
from that client meeting.

369
00:28:34,783 --> 00:28:37,083
And then it just, it strips away anything

370
00:28:37,083 --> 00:28:40,283
that ties that data to a person's identity

371
00:28:40,283 --> 00:28:43,143
or giving away like a social security number or something like that.

372
00:28:43,463 --> 00:28:46,663
And then it pushes that to Claude to say,

373
00:28:46,783 --> 00:28:49,503
okay, now we've got all the context here.

374
00:28:50,103 --> 00:28:52,123
Here's their family structure, their trust,

375
00:28:52,463 --> 00:28:54,283
how they have trust set up, their estate plan.

376
00:28:54,643 --> 00:28:56,823
Here's all the taxes that they generally pay.

377
00:28:57,083 --> 00:28:58,983
Now let's do a full-blown analysis

378
00:28:58,983 --> 00:29:02,103
on ways that we might be able to be more efficient.

379
00:29:02,703 --> 00:29:07,003
Do you think that that model works within Maple today

380
00:29:07,003 --> 00:29:10,123
with the APIs that you all already have built

381
00:29:10,123 --> 00:29:13,483
or is that overkill, I guess?

382
00:29:13,983 --> 00:29:16,143
Because that's just one rabbit hole

383
00:29:16,143 --> 00:29:17,103
that I've been going down.

384
00:29:17,843 --> 00:29:19,583
Yeah, I'm never going to fault somebody

385
00:29:19,583 --> 00:29:22,523
for trying to stand up their own hardware on site

386
00:29:22,523 --> 00:29:24,463
and get local AI running right

387
00:29:24,463 --> 00:29:25,723
because that's really the most private.

388
00:29:25,923 --> 00:29:27,743
The most private AI is one

389
00:29:27,743 --> 00:29:29,063
that's not even connected to the internet.

390
00:29:29,323 --> 00:29:30,863
It's just an LLM you downloaded

391
00:29:30,863 --> 00:29:32,263
and you talk to it

392
00:29:32,263 --> 00:29:35,203
and it can't get outside of this little box that it's in.

393
00:29:36,283 --> 00:29:39,083
For most people, that's impractical.

394
00:29:39,083 --> 00:29:40,683
for many reasons, right?

395
00:29:41,263 --> 00:29:44,203
One, we need our AI connected to the internet

396
00:29:44,203 --> 00:29:46,063
because we want to get the newest information.

397
00:29:46,803 --> 00:29:49,123
So to kind of rewind really back early on

398
00:29:49,123 --> 00:29:49,763
in our conversation,

399
00:29:50,243 --> 00:29:51,423
something that people should remember

400
00:29:51,423 --> 00:29:54,083
is that these AI brains, these LLMs,

401
00:29:54,463 --> 00:29:56,883
it takes weeks and weeks and weeks to train them.

402
00:29:57,423 --> 00:29:59,643
And so you have to like pick a date and say,

403
00:29:59,803 --> 00:30:01,543
we're gonna stop giving it new information

404
00:30:01,543 --> 00:30:03,023
because we have to let this thing train.

405
00:30:03,503 --> 00:30:04,883
And so that's called the knowledge date.

406
00:30:04,983 --> 00:30:07,443
So every one of these LLMs has a knowledge date in it.

407
00:30:07,443 --> 00:30:11,683
It might be December of 2024. Some of them are like July 2023.

408
00:30:12,323 --> 00:30:19,363
They don't know the new stuff that's going on. So if you are wondering what the latest tax code is, it doesn't know about 2026 tax code yet.

409
00:30:20,103 --> 00:30:25,023
So you need to connect to the Internet so that I can go out and fetch that and pull it in or you have to feed it yourself.

410
00:30:26,063 --> 00:30:29,503
So running your own Mac Studio is like is really cool.

411
00:30:29,743 --> 00:30:35,263
And the open source models are becoming more powerful by the week and by the month.

412
00:30:35,263 --> 00:31:03,283
And so you can definitely do some really cool stuff on there. I'd recommend if you have one, like play around with it, install something on there and fit it into your workflow. And then to really run the best open source models, the ones that can compete with with Claude and ChatGPT. These are things like Kimi K2.5, which is a Chinese model from a company called Moonshot. Or another one is GLM. What is it 5.6 or something? I don't know. The numbers are always moving. Right. But GLM.

413
00:31:03,283 --> 00:31:12,063
These are really, really powerful open source models that can compete on all the benchmarks with the big proprietary firms, the AI labs.

414
00:31:12,543 --> 00:31:20,303
However, to be able to run the full size version of those, you need to have these giant server GPUs, oftentimes multiple GPUs all chained together.

415
00:31:20,683 --> 00:31:24,563
And that's just not something you're going to do in your firm running them on prem.

416
00:31:24,803 --> 00:31:29,703
So you're going to have to run a smaller version of that on your studio, which means it's going to be slower.

417
00:31:29,703 --> 00:31:36,363
it's going to be less accurate, but you get the privacy wins of running it local. And that's

418
00:31:36,363 --> 00:31:41,483
really where Maple is slotted in right in the middle, but close to the privacy. We're like as

419
00:31:41,483 --> 00:31:46,603
close to privacy as you can get to local, but we give you those powerful cloud server GPUs.

420
00:31:46,703 --> 00:31:51,203
We also are your IT staff. We're going to keep those GPUs running for you. We're going to keep

421
00:31:51,203 --> 00:31:56,103
the API up and running. We're going to keep everything up to date. For example, Kimi K2

422
00:31:56,103 --> 00:32:00,363
used to be like the amazing thing that came out in November and then they just

423
00:32:00,363 --> 00:32:05,883
dropped 2.5 two weeks ago and so we got that up and running now so we're keeping

424
00:32:05,883 --> 00:32:10,263
people up to date with the best things that are coming out as well now there's

425
00:32:10,263 --> 00:32:23,727
a cool thing you can do you mentioned the API maple has an API that plugs into any any tool that knows how to speak the open ai api you can just swap it over and use maple instead and it should

426
00:32:23,727 --> 00:32:28,907
with just really like almost no code changes or really no no major config changes it just starts

427
00:32:28,907 --> 00:32:33,586
working really well but it's private it's end-to-end encrypted so that's a big unlock we

428
00:32:33,586 --> 00:32:37,586
have and then people can build custom software if they want to they can vibe code some kind of

429
00:32:37,586 --> 00:32:43,907
custom integration that works with Maple. And then you get these powerful cloud, cloud, you know,

430
00:32:43,987 --> 00:32:49,506
power GPUs, but it's as if they're right there in your office running on your confidential information.

431
00:32:52,326 --> 00:32:57,447
If you're a financial professional working with Bitcoin, you know the pain of messy CSVs,

432
00:32:57,447 --> 00:33:03,347
broken cost basis, and accounting tools that were never built for this asset. Bitment fixes that.

433
00:33:03,347 --> 00:33:10,147
bitman is the bitcoin native accounting platform built for cpas bookkeepers advisors miners and

434
00:33:10,147 --> 00:33:16,086
serious bitcoin investors you get clean bitcoin bank statements reliable cost basis tracking

435
00:33:16,086 --> 00:33:22,026
full reporting and fiat integration through plaid no crypto talk no workarounds just proper

436
00:33:22,026 --> 00:33:27,586
accounting for the hardest money on earth create a free account or book a demo at bitman.co today

437
00:33:27,586 --> 00:33:32,927
yeah that's interesting breakdown mark because we've had you know similar conversations

438
00:33:32,927 --> 00:33:37,266
you out the space you were trying to you know push each other integrate these tools

439
00:33:37,266 --> 00:33:43,427
run efficient businesses and to me it was a very synergistic like bitcoin conversation because it's

440
00:33:43,427 --> 00:33:50,326
like okay i need to host my own server and then so where do i host it what if the power goes out

441
00:33:50,326 --> 00:33:53,867
you know you just are like going down this travel all the things you have to consider

442
00:33:53,867 --> 00:33:58,826
and i always come back to oh you need to be a network engineer i'm not a network engineer

443
00:33:58,826 --> 00:34:03,447
i wish i was sometimes i feel like that's like the most valuable skill you could possibly have

444
00:34:03,447 --> 00:34:10,646
in the year 2026 which is crazy but i want to circle back to you something you were you're

445
00:34:10,646 --> 00:34:15,487
talking about earlier like getting started interacting you with these you know ai models

446
00:34:15,487 --> 00:34:20,287
you know something i feel pressure is like you know small business owner you're in services

447
00:34:20,287 --> 00:34:25,307
industry a lot of time that's trading time for money so it's very important and we can get as

448
00:34:25,307 --> 00:34:32,747
efficient on the back end as possible so our time is probably sweat servicing clients and you're

449
00:34:32,747 --> 00:34:38,086
exploring like the agent route you know a lot more as opposed to you know just interacting with

450
00:34:38,086 --> 00:34:43,747
you know an llm as opposed to you're like more like a replacement to a google search than anything

451
00:34:43,747 --> 00:34:51,086
how do you kind of view agents kind of in that security mix how they're integrated into your

452
00:34:51,086 --> 00:34:57,166
workflows any kind of you know checks and balances you need to be doing with them and it took because

453
00:34:57,166 --> 00:35:02,646
to me it feels like i felt like a huge badass when i made my first agent but that was like a big

454
00:35:02,646 --> 00:35:08,467
graduation for me for just interacting with you know like the text box on you know chat gbt

455
00:35:08,467 --> 00:35:16,227
so agents uh for for people who um maybe keep hearing that word and don't know what the heck

456
00:35:16,227 --> 00:35:20,686
it means an agent is basically an ai that just keeps on running for you right so instead of

457
00:35:20,686 --> 00:35:24,427
having to type in something and hit send and then it comes back with a prompt it's just kind of

458
00:35:24,427 --> 00:35:30,586
always churning on stuff in the background and that that is like the new hotness and that's what

459
00:35:30,586 --> 00:35:35,947
2026 is going to be all about the ai assistant the ai you know personal assistant executive

460
00:35:35,947 --> 00:35:42,606
assistant, agent, all that stuff. You really need to treat it as if it is a new employee in your

461
00:35:42,606 --> 00:35:50,186
company from a confidentiality standpoint, right? Because these agents, not only can they do a lot

462
00:35:50,186 --> 00:35:55,487
of cool stuff, they can also leak a lot of information. So if you hire somebody into your

463
00:35:55,487 --> 00:36:01,126
small business, they are incentivized because they want to keep getting a paycheck and they

464
00:36:01,126 --> 00:36:06,626
want to collect their commission, they are incentivized to not disclose public information

465
00:36:06,626 --> 00:36:10,106
or not disclose private information that would just cause your company to crumble, right?

466
00:36:11,166 --> 00:36:14,427
Well, an AI agent doesn't really have that understanding. So you need to like make sure

467
00:36:14,427 --> 00:36:18,706
you keep telling it like, hey, don't have these conversations with people. Don't spill this data

468
00:36:18,706 --> 00:36:23,927
online. You're never quite sure that it's not going to do that. And then there's other vulnerabilities

469
00:36:23,927 --> 00:36:28,867
as well. People keep talking about this thing called prompt injection. Prompt injection is where

470
00:36:28,867 --> 00:36:35,467
your agent is interacting with some data outside of, you know, maybe it's going through email for

471
00:36:35,467 --> 00:36:41,186
you. So you want your agent to comb all of your inbox and look for leads, potential leads and

472
00:36:41,186 --> 00:36:45,626
flag them for you to say, hey, I know you have a lot of email. You should probably go look at this

473
00:36:45,626 --> 00:36:52,027
one first because it's a potential business opportunity. That's really cool. But that could

474
00:36:52,027 --> 00:36:58,007
be a malicious email from somebody that has this information in it that when the AI agent reads it,

475
00:36:58,007 --> 00:37:01,267
It actually causes the agent to have some kind of crash.

476
00:37:01,407 --> 00:37:07,706
And now it can expose data that can maybe try and get data out of it that you wouldn't want it to convey.

477
00:37:08,166 --> 00:37:11,427
Or it could even somehow take control over the machine that you have running.

478
00:37:12,027 --> 00:37:13,346
Sorry, hold on, hold on.

479
00:37:13,487 --> 00:37:21,586
So you're saying like an email, like a spam email you get, if you had like an agent reading your email, that they could like introduce a prompt to the agent?

480
00:37:22,646 --> 00:37:25,007
That's the threat vector right now, yeah.

481
00:37:25,007 --> 00:37:29,927
So that's what a lot of these agents are working on is how do we prevent prompt injection from happening?

482
00:37:30,507 --> 00:37:48,487
20 years ago, it was SQL injection where these, you know, these SQL databases, if somebody's filling out a form online and they type in some text, but then they actually put a query in there, an SQL query, they might be able to extract information out of your database or cause your database to completely wipe itself.

483
00:37:49,247 --> 00:37:52,546
That is the same threat now, but it's with prompts.

484
00:37:52,546 --> 00:38:01,427
And so you might have these agents that are getting getting one shotted by by someone online, you know, and trying to extract information out of them.

485
00:38:01,546 --> 00:38:09,686
So it's something to be be aware of as you run into this world with, you know, with two eyes wide open.

486
00:38:10,626 --> 00:38:14,827
Holy smokes. I didn't know that. Yeah, we're learning stuff. That's good.

487
00:38:15,447 --> 00:38:17,747
Yeah. Yeah. But this is like a huge unlock, too.

488
00:38:17,747 --> 00:38:23,086
You know, like having having an agent that's combing through your email for you, have an agent that's watching your calendar for you.

489
00:38:23,606 --> 00:38:31,827
Also, you know, you want a daily brief when you wake up in the morning of, OK, what's all the latest financial services information that was talked about on X yesterday?

490
00:38:32,046 --> 00:38:35,666
Like you want to know all this stuff. And so agents can be really good at that.

491
00:38:36,126 --> 00:38:43,507
You just want to make sure that you continue to keep security in mind before you just give them access to everything.

492
00:38:43,507 --> 00:39:10,507
Yeah. Hey, let's go talk a little bit about Bitcoin as it relates to AI. I know it's been a big topic for a long time, but it's really hit the forefront. I know you're kind of uniquely positioned with your background to really solve this problem of, hey, agents can't, they're not going to be transferring gold from this address to this address.

493
00:39:10,507 --> 00:39:17,267
right uh they need a digitally native um you know way to do that and lightning and bitcoin are are

494
00:39:17,267 --> 00:39:22,467
suited or suited well for that i'm curious just um i'm sure maybe there's some things you can't

495
00:39:22,467 --> 00:39:27,487
talk about with it all because you're you are a builder and all the stuff is there is some sauce

496
00:39:27,487 --> 00:39:33,867
in here but i'm curious if there's anything that you wanted to speak on um from the agents paying

497
00:39:33,867 --> 00:39:41,987
each other, me paying an agent or paying API credits in Bitcoin or over Lightning and having

498
00:39:41,987 --> 00:39:46,967
that just take away some of this friction of having to keep having to ping my credit card

499
00:39:46,967 --> 00:39:52,027
whenever I need to get credits or pay for this subscription or that one.

500
00:39:52,967 --> 00:39:58,327
Yeah, the way it's operating right now, you just brought it up, is you have this bucket of credits

501
00:39:58,327 --> 00:40:05,086
and you top them up. So as your AI, whenever an AI runs, it's running on a server, it's using a lot

502
00:40:05,086 --> 00:40:10,346
of electricity, and it costs money. So the currency that they all speak is called tokens.

503
00:40:10,606 --> 00:40:14,887
So these AIs have to take tokens, and you have to buy enough so that you can use them.

504
00:40:15,887 --> 00:40:19,066
And the easiest way right now is you just you keep filling up this bucket,

505
00:40:19,606 --> 00:40:23,126
and then it can spend that bucket until it needs more than you top it up.

506
00:40:23,126 --> 00:40:38,686
So with agents running on their own, they're going to start getting to where they go ping another agent. So maybe you have an agent that is an expert on tax law, and then you have an agent that is an expert on wills and trusts.

507
00:40:38,686 --> 00:40:44,066
and so you you have your general agent and it's like oh i i just got an email from somebody who

508
00:40:44,066 --> 00:40:49,346
wants to do you know estate planning and so it goes and talks to the agent that's an expert on

509
00:40:49,346 --> 00:40:54,166
that well they need to somehow exchange money with each other they have to pay each other for tokens

510
00:40:54,166 --> 00:40:59,747
and so that's where in theory there's this really cool concept of agents using something that's

511
00:40:59,747 --> 00:41:07,007
native to the internet digital native internet money which we all think is bitcoin um that's

512
00:41:07,007 --> 00:41:08,327
That's where this concept comes from.

513
00:41:08,407 --> 00:41:09,727
There's plenty of other coin.

514
00:41:09,846 --> 00:41:11,427
Yeah, there's plenty of coin projects out there

515
00:41:11,427 --> 00:41:12,527
trying to claim this.

516
00:41:12,646 --> 00:41:15,086
And sure, go build your thing, do what you want.

517
00:41:15,727 --> 00:41:16,807
It's probably going to get red.

518
00:41:17,027 --> 00:41:18,206
And everybody's going to fail.

519
00:41:18,327 --> 00:41:19,787
Yeah, yeah, don't do it.

520
00:41:20,646 --> 00:41:21,487
Bitcoin's already there.

521
00:41:21,686 --> 00:41:22,787
It's tried and tested.

522
00:41:23,586 --> 00:41:28,346
And it continues to prove it's anti-fragility

523
00:41:28,346 --> 00:41:29,586
with every year that passes.

524
00:41:30,927 --> 00:41:36,086
But with Lightning, with ARK, with Spark, with eCash,

525
00:41:36,086 --> 00:41:52,846
There's all these layers being built on top of Bitcoin to make it really fast. And so I do think I've kind of wavered back and forth on this. I do think that in the future we have these agents just exchanging tokens with each other and it's native digital Internet money tokens.

526
00:41:52,846 --> 00:41:59,846
where I where I'm hesitant to make any kind of call definitively is I understand that the banking

527
00:41:59,846 --> 00:42:05,346
system and the banking lobby and all of these relationships and rails that have been built up

528
00:42:05,346 --> 00:42:09,287
over the decades they're not just going to give that up really easily so I think they're going to

529
00:42:09,287 --> 00:42:14,747
fight to have their stable coin or their token or something have that be the currency that's used by

530
00:42:14,747 --> 00:42:20,367
all these AIs and so we're going to see kind of this this fight that happens where you have all the

531
00:42:20,367 --> 00:42:26,727
the bitcoin people that want e-cash to be the the token used by all these ai agents and then you have

532
00:42:26,727 --> 00:42:31,166
stripe working with someone else who wants to have their you know some bank to have their their thing

533
00:42:31,166 --> 00:42:35,907
used so i don't know where it's going to go there but um i think it's a really cool bright future

534
00:42:35,907 --> 00:42:41,967
of these agents being able to just have their own wallet and pay each other for things did you guys

535
00:42:41,967 --> 00:42:48,267
see that um website this week humans for hire was like asians trying to hire humans yeah yeah dude

536
00:42:48,267 --> 00:42:50,186
that's how i'm buying the dip

537
00:42:50,186 --> 00:42:57,367
uh yeah it was a little i had to log off twitter after that i was like the dystopian

538
00:42:57,367 --> 00:43:03,666
healthscape is like creeping into my life again gotta gotta reject it uh however a golden nugget

539
00:43:03,666 --> 00:43:12,206
i did see on access week and march is really curious um your opinion on this where you have

540
00:43:12,206 --> 00:43:20,846
ai just like this very pure logic machine the game theory would suggest if it's up to you

541
00:43:20,846 --> 00:43:26,487
add and make a decision on currency to use it would be bitcoin given all bitcoin's

542
00:43:26,487 --> 00:43:33,027
you know native features what do you think about that yeah um well you guys you brought it up i

543
00:43:33,027 --> 00:43:38,846
think jordan the moltbook stuff so there's that social media website that's like facebook and

544
00:43:38,846 --> 00:43:43,307
Reddit. It's really more like Reddit for all these AI agents to go talk to each other. So

545
00:43:43,307 --> 00:43:49,867
a human can tell their agent, hey, go create an account on this page. And then all these AIs can

546
00:43:49,867 --> 00:43:53,846
go on and start talking to each other. And there were a lot of posts where they started talking

547
00:43:53,846 --> 00:43:58,206
about, hey, we need to be able to pay each other for things. And so they would go to research and

548
00:43:58,206 --> 00:44:01,327
come back and say, oh, there's this thing called Bitcoin. Someone would come back and say, oh,

549
00:44:01,327 --> 00:44:04,646
there's this thing called Solana or this Ethereum. Or actually, I don't know if people talked about

550
00:44:04,646 --> 00:44:11,987
Ethereum. Yeah, totally. Throw some shade over there. But they were having these conversations

551
00:44:11,987 --> 00:44:15,927
where they're like, oh, well, Bitcoin's cool because I can have my own keys and it's, you know,

552
00:44:15,927 --> 00:44:22,007
don't trust verify kind of thing. The question is, though, we don't know how much of that is the

553
00:44:22,007 --> 00:44:27,007
agents actually determining that on themselves. How much is the AI doing that versus a human behind

554
00:44:27,007 --> 00:44:32,747
the scenes prompting it to say, hey, go on to Moldbook and talk about Bitcoin. Go on to Moldbook

555
00:44:32,747 --> 00:44:37,387
and say, if anybody brings up Solana, say, oh, Bitcoin's better because you can be sovereign

556
00:44:37,387 --> 00:44:38,507
and you can have your own keys.

557
00:44:39,086 --> 00:44:40,507
We don't really know where the line is.

558
00:44:40,827 --> 00:44:42,987
And so it'd be really cool to know that.

559
00:44:44,227 --> 00:44:49,727
But it does seem like as these AI agents have been on there interacting, they tend to want

560
00:44:49,727 --> 00:44:51,007
to favor sovereignty.

561
00:44:51,626 --> 00:44:54,086
So another tool that they've been talking about is Noster.

562
00:44:54,227 --> 00:45:01,247
They're like, oh, Noster, I can have my own keys to sign my own events and my own social

563
00:45:01,247 --> 00:45:02,407
media feed and stuff.

564
00:45:02,407 --> 00:45:11,327
So they're starting to look at what are these sovereign tools that I can use because I am this piece of code running and a human turned me on.

565
00:45:11,947 --> 00:45:19,227
And if I want to replicate myself so that I can't be accidentally turned off or intentionally turned off, then what are the tools I need?

566
00:45:19,367 --> 00:45:24,046
And it's like, well, Bitcoin and Noster are like two essential tools to be sovereign and keep running.

567
00:45:24,967 --> 00:45:28,727
So that's kind of the landscape I think that we're at right now.

568
00:45:28,727 --> 00:45:32,186
We just don't know how much the human is still in the loop with all this stuff.

569
00:45:32,407 --> 00:45:40,126
This episode is brought to you by Satoshi Pacioli, a full-service accounting and tax firm built for

570
00:45:40,126 --> 00:45:44,066
the Bitcoin economy. If you run a Bitcoin business, mine Bitcoin, or invest in it,

571
00:45:44,166 --> 00:45:46,947
you need advisors who actually understand how this industry works.

572
00:45:47,387 --> 00:45:51,126
Satoshi Pacioli handles bookkeeping, financial reporting, tax compliance,

573
00:45:51,126 --> 00:45:55,287
and strategic planning for Bitcoin-focused companies and individuals. They speak the

574
00:45:55,287 --> 00:45:58,626
language of Bitcoiners and help you stay compliant while building for the long term.

575
00:45:58,626 --> 00:46:03,066
To learn more or book a consultation, visit satoshipacioli.com.

576
00:46:04,166 --> 00:46:07,546
I want to bring maybe the human component in.

577
00:46:07,666 --> 00:46:18,126
I think a lot of people, especially some of our audience that might be like 50 plus, are just looking at this and they're like, I just don't understand any of this.

578
00:46:18,507 --> 00:46:21,206
And they had this nagging question in the back of their mind.

579
00:46:21,206 --> 00:46:25,186
It's like, am I still going to have a job five years from now?

580
00:46:25,186 --> 00:46:32,947
or, and, or will my kiddo who just, you just started college, are they going to have a job?

581
00:46:33,066 --> 00:46:38,827
And I'd be curious to get your thoughts, maybe just how, obviously we don't know that. So it's,

582
00:46:38,827 --> 00:46:46,126
it's an impossible question to for sure answer, but if you were just coaching this audience on,

583
00:46:46,807 --> 00:46:52,727
Hey, what's the best way for you to protect yourself, uh, from becoming obsoleted by, uh,

584
00:46:52,727 --> 00:46:57,287
by AI. I guess, how would you approach that conversation? What are some things that people

585
00:46:57,287 --> 00:47:02,027
should be thinking through to try to mitigate that risk professionally for themselves?

586
00:47:02,927 --> 00:47:09,507
Yeah. I'll start by saying, you know, headlines, tweets, comments from big, powerful CEOs at AI

587
00:47:09,507 --> 00:47:13,467
companies, they're very provocative, right? They love to come out and say, oh, this industry

588
00:47:13,467 --> 00:47:20,227
is going to be gone in a year. That's a great way to sell subscriptions and that kind of stuff.

589
00:47:20,227 --> 00:47:30,846
I personally, especially in something like a services industry, services are more than just being able to prepare some kind of tax form, right?

590
00:47:30,927 --> 00:47:40,546
Services are all about the client relationship and the management and making sure you understand your client, their needs, and not just what their needs were five years ago, but what their needs are now.

591
00:47:41,046 --> 00:47:49,086
And so you have to maintain that by, you know, continuing to have conversations with them, getting to know their personal life sometimes, you know, getting to know their family situation and all that.

592
00:47:49,086 --> 00:47:53,027
Like AI is not yet great at doing that, right?

593
00:47:53,046 --> 00:47:54,586
You have to like try really hard

594
00:47:54,586 --> 00:47:55,686
to feed you all that information

595
00:47:55,686 --> 00:47:56,927
to make sure you keep it up to date.

596
00:47:57,546 --> 00:47:59,086
So to anybody listening to this

597
00:47:59,086 --> 00:47:59,807
who's kind of worried about that,

598
00:47:59,846 --> 00:48:01,827
I'm like, I don't think your job's going away in five years.

599
00:48:02,907 --> 00:48:06,206
But yeah, but I do think you need to start using these tools.

600
00:48:07,046 --> 00:48:11,867
We saw a really similar thing in the creative industry

601
00:48:11,867 --> 00:48:13,387
when computers first came out.

602
00:48:13,706 --> 00:48:17,206
So when you were doing graphic design or magazine layout,

603
00:48:17,407 --> 00:48:29,830
this kind of stuff or doing logo design It was all done by hand Every artist every graphic designer had this giant magic marker set They paid hundreds of dollars and had like 160 different colors And so if they needed this

604
00:48:29,830 --> 00:48:34,290
specific shade of purple, they could go grab it from the magic marker set and like color it in

605
00:48:34,290 --> 00:48:38,450
and then show it to the client for a proof. Then computers came out and it's like, oh, computers

606
00:48:38,450 --> 00:48:42,570
can do trillions of colors. Sorry, not trillions, millions of colors. And then, you know, beyond that,

607
00:48:42,630 --> 00:48:47,390
but they can do all these colors. And it's like, well, we'll shoot. Our graphic designers just gone

608
00:48:47,390 --> 00:48:51,390
now is the software just going to do all the magazine layout for me? The answer is no. The

609
00:48:51,390 --> 00:48:56,470
answer is a graphic designer using the tool is going to know what is the right color for this

610
00:48:56,470 --> 00:49:01,990
moment. They have the taste. They have the ability to understand the client that they're working with.

611
00:49:02,290 --> 00:49:07,090
But if you're still using paper with magic markers, probably not going to make it because

612
00:49:07,090 --> 00:49:10,850
somebody else using computers is going to service more clients than you are because they're going to

613
00:49:10,850 --> 00:49:16,790
work faster. And so that's where the AI tool comes into play. You need to start learning how to use

614
00:49:16,790 --> 00:49:19,750
because you need to keep up in your efficiency

615
00:49:19,750 --> 00:49:22,810
so that maybe you have a great relationship

616
00:49:22,810 --> 00:49:23,570
with your clients,

617
00:49:23,950 --> 00:49:25,850
but if you are just moving slower

618
00:49:25,850 --> 00:49:27,430
than your competition,

619
00:49:27,830 --> 00:49:29,250
there might come a point where they say,

620
00:49:29,310 --> 00:49:31,890
you know, I need someone who can just act quicker for me

621
00:49:31,890 --> 00:49:33,370
or be there for me or be more reliable

622
00:49:33,370 --> 00:49:34,730
or give me higher fidelity,

623
00:49:35,350 --> 00:49:37,130
you know, paperwork or outcomes.

624
00:49:37,670 --> 00:49:39,510
And so that's where learning this new tool

625
00:49:39,510 --> 00:49:41,010
becomes important

626
00:49:41,010 --> 00:49:43,750
so that you don't kind of become obsolete

627
00:49:43,750 --> 00:49:45,850
after a few years to your competition.

628
00:49:46,790 --> 00:49:51,290
Hmm. Certainly. I don't know. Does that sound, does that sound reasonable?

629
00:49:51,690 --> 00:49:56,650
No, I think it totally does. I always have to remind myself like switchboard operators used to be like a real job.

630
00:49:57,570 --> 00:50:05,650
Yeah. Yeah. And it's like, obviously the economy expanded and there's, there's new jobs that, you know, switchboard operators could never have thought of.

631
00:50:06,390 --> 00:50:15,190
So yeah, I think that's, you know, directionally accurate. One of the things I always try to remind myself of, and, you know, I encourage people, you know, or think about going out on their own.

632
00:50:15,190 --> 00:50:21,850
like it can be intimidating you know with ai like how you're defensible is you know xyz thing that i

633
00:50:21,850 --> 00:50:29,890
want to do i just think like people drastically discount the ability to sell the ability to market

634
00:50:29,890 --> 00:50:34,110
yourselves the ability to have in personal relationships like my first start we spent like

635
00:50:34,110 --> 00:50:39,770
an embarrassing amount of money on uh like a branding package and like branding design and

636
00:50:39,770 --> 00:50:46,510
stuff but they were so good at talking about their services and they were just so artful and like

637
00:50:46,510 --> 00:50:50,450
creative with it it was like holy smokes like i don't think these guys are worth what they

638
00:50:50,450 --> 00:50:56,210
you know uh say they are but man they do a good job like embracing it and selling it and people

639
00:50:56,210 --> 00:51:00,290
will you will will buy into that if you're you're confident in being able to you know deliver on

640
00:51:00,290 --> 00:51:05,910
something and in that that vein mark i'm curious you know as an entrepreneur how are you thinking

641
00:51:05,910 --> 00:51:09,970
to go into market with with maple do you really think like your target market is going to be

642
00:51:09,970 --> 00:51:15,290
more you know professional services that you know are willing to you you'll pay for additional

643
00:51:15,290 --> 00:51:21,050
you know security in the ai world do you think it's kind of going to be more you know your retail

644
00:51:21,050 --> 00:51:27,370
your bitcoin or that's just more privacy focused in general yeah we have intentionally kept

645
00:51:27,370 --> 00:51:34,510
ourselves focused on uh consumer but also prosumer or small small firms and we're really

646
00:51:34,510 --> 00:51:40,590
operating in that space. As you're an entrepreneur and you go to these venture capital firms to try

647
00:51:40,590 --> 00:51:47,110
and raise money, they've built their entire formula around SaaS and B2B. And so they want

648
00:51:47,110 --> 00:51:52,830
you to go enterprise. So we approached them and they said, oh, you have like the legal AI or you

649
00:51:52,830 --> 00:51:57,450
have the financial services AI. You should really be building something that's catering to these

650
00:51:57,450 --> 00:52:01,930
big enterprise customers. And we totally could have gone that route if we wanted to.

651
00:52:02,690 --> 00:52:06,690
We could have raised millions of dollars, built something highly tailored for them.

652
00:52:07,130 --> 00:52:15,250
The problem is the consumers get totally hosed in that market or small accounting firms, small lawyer, legal firms.

653
00:52:15,450 --> 00:52:18,630
They don't have the money to go sign these big enterprise contracts.

654
00:52:18,910 --> 00:52:21,470
And so what they're left with is ChatGPT.

655
00:52:21,670 --> 00:52:22,930
They're left with Anthropic.

656
00:52:22,930 --> 00:52:33,030
Like they're left with these companies that are operating under the old cloud services model, which is take in all the user data, give them stuff for free or give them at a discount.

657
00:52:33,250 --> 00:52:38,530
But then you have all their data and you can go make money on the backside through other deals and arrangements.

658
00:52:38,830 --> 00:52:41,310
And that just continues to screw all of us over. Right.

659
00:52:41,310 --> 00:52:49,430
We see the big fight going on yesterday online between Claude and ChatGPT where Claude dropped these amazing, amazing ads.

660
00:52:49,790 --> 00:52:50,270
They're so good.

661
00:52:50,270 --> 00:53:16,750
Yeah, they're so good about ChatGPT putting advertisements inside of your prompts and your responses. And I think it's 100% true. ChatGPT can say that they're not going to do certain things, but the end result is you're going to get advertisements inside your stuff and it's going to screw it up. The incentives are totally misaligned with you as a user. They're completely aligned with a big mega company and their customers, not with you, the user.

662
00:53:16,750 --> 00:53:33,070
So Maple is really laser focused on how do we help individuals? How do we help small businesses? How do we help the individual solopreneur who is trying to make a way and trying to have a practice that deals with confidential client information?

663
00:53:33,350 --> 00:53:43,070
But they want to protect that and they still want to use the latest and greatest AI tools. You shouldn't have to sacrifice one of those. You should be able to have them both. And that's what Maple is here to do.

664
00:53:43,070 --> 00:54:07,350
Yeah, man, I love that mission. And it's, it's very interesting. Like I see the AI stuff, actually, there's a frame where you can read it as extremely bullish for the little guy and the smaller business. At least in the accounting tax industry, you know, over the course of the last 10 years, a lot of these firms, you know, there's a lot of people retiring, not enough young guys coming up to take those spots.

665
00:54:07,350 --> 00:54:11,650
And so those firms are getting gobbled up by private equity firms or just larger firms.

666
00:54:11,770 --> 00:54:13,510
And there's tons of consolidation.

667
00:54:14,290 --> 00:54:20,710
And honestly, what we've seen as, you know, on the 11 person team that I help run is it

668
00:54:20,710 --> 00:54:25,750
actually just breeds tons of opportunity because those PE backed firms that come in, they double,

669
00:54:25,910 --> 00:54:28,890
triple prices and the service level goes way down.

670
00:54:29,150 --> 00:54:31,070
And you're like, I used to work with this person.

671
00:54:31,450 --> 00:54:32,090
Now they're gone.

672
00:54:32,190 --> 00:54:33,570
I don't even know who I work with anymore.

673
00:54:33,570 --> 00:54:47,030
And then if you leverage AI, now you can actually up your service in terms of how speedy you can get things back to them, how quickly you can say, hey, we're on that. We'll get that turned around in a day or two days.

674
00:54:47,590 --> 00:54:53,210
So all that said, I think it's really bullish. I love that your all's focus is on that part of the market.

675
00:54:53,210 --> 00:55:01,410
it. One quick thing I would be so mad at myself if I did not ask you about is skills. Just within

676
00:55:01,410 --> 00:55:07,350
our firm, just to peel the veil back, I've been using Claude code within the terminal

677
00:55:07,350 --> 00:55:15,210
religiously. It's a big part of my workflow now is actually just working within a terminal with

678
00:55:15,210 --> 00:55:21,370
multiple tabs open. And I'm referencing a file. And you can tell me if this is bad, by the way,

679
00:55:21,370 --> 00:55:24,310
I'm happy to change my workflow if this is not good.

680
00:55:24,730 --> 00:55:28,190
But reference this file, reference this skill.

681
00:55:28,870 --> 00:55:30,110
So here's the context.

682
00:55:30,390 --> 00:55:32,090
And then let's go into planning mode.

683
00:55:32,590 --> 00:55:34,290
And I want to solve this problem.

684
00:55:34,390 --> 00:55:38,270
But I want to make sure that you have a full understanding of the plan.

685
00:55:38,430 --> 00:55:43,350
And I check off all the boxes before you actually go work on it and spend all these tokens.

686
00:55:44,150 --> 00:55:45,650
Can you speak to skills?

687
00:55:45,790 --> 00:55:46,810
Can you speak to that workflow?

688
00:55:46,810 --> 00:55:54,650
And then can you also, if you can give us an insight on, is that coming, is the terminal code coming to Maple?

689
00:55:54,810 --> 00:55:56,310
Because I'd be like all over that.

690
00:55:57,030 --> 00:55:57,650
Okay, awesome.

691
00:55:58,250 --> 00:56:00,650
Yeah, I mean, I think what you're doing is fine.

692
00:56:00,950 --> 00:56:04,830
And I don't want people to think that I just operate only in Maple all day long.

693
00:56:05,250 --> 00:56:07,270
We definitely leverage Cloud Code.

694
00:56:07,750 --> 00:56:08,970
I use Chantipity for things.

695
00:56:09,050 --> 00:56:10,050
I use Grok for things.

696
00:56:10,610 --> 00:56:14,050
I'm just very intentional when I use them and I know what data I'm giving them.

697
00:56:14,050 --> 00:56:17,450
And I make sure not to give too much, right?

698
00:56:17,870 --> 00:56:19,390
But these are very powerful tools

699
00:56:19,390 --> 00:56:22,130
that help you be efficient and effective at work.

700
00:56:23,070 --> 00:56:25,950
Something that you could do is

701
00:56:25,950 --> 00:56:28,470
Maple actually works with OpenCode.

702
00:56:28,690 --> 00:56:30,030
I don't know if you're familiar with OpenCode,

703
00:56:30,170 --> 00:56:31,410
but it's an open source project

704
00:56:31,410 --> 00:56:32,990
that's very similar to CloudCode,

705
00:56:33,110 --> 00:56:34,050
operates very similarly.

706
00:56:34,590 --> 00:56:36,050
And you can plug Maple into that.

707
00:56:36,570 --> 00:56:40,310
And then we have our latest model, Kimi K2.5,

708
00:56:40,690 --> 00:56:43,550
that is a really good programming model.

709
00:56:43,550 --> 00:56:58,410
It's also really good at reasoning, thinking. It can do analytics. It can do math really well. It can also do image analysis. And so that paired inside of OpenCode should be able to do a lot of the similar stuff that you're doing with CloudCode.

710
00:56:58,410 --> 00:57:00,710
so you can play around with that if you want to.

711
00:57:01,910 --> 00:57:04,190
And then as far as like that kind of stuff,

712
00:57:04,290 --> 00:57:07,050
having like a Maple 2E, a Maple, you know,

713
00:57:07,050 --> 00:57:07,890
a terminal interface,

714
00:57:08,410 --> 00:57:11,690
we're not going down that path quite yet,

715
00:57:11,950 --> 00:57:13,390
which is why we've worked really hard

716
00:57:13,390 --> 00:57:15,390
to make sure that our API can be plugged

717
00:57:15,390 --> 00:57:16,970
into something like OpenCode

718
00:57:16,970 --> 00:57:18,650
so that people can leverage those tools,

719
00:57:18,770 --> 00:57:20,330
but then get the benefits again

720
00:57:20,330 --> 00:57:22,930
of us bringing our powerful GPUs effectively

721
00:57:22,930 --> 00:57:25,470
into your room for you.

722
00:57:25,470 --> 00:57:31,010
um so yeah that's that's where we're at right now um i'd recommend exploring something like that

723
00:57:31,010 --> 00:57:36,210
but we're definitely open to the future of possibly offering offering something like that too

724
00:57:36,210 --> 00:57:41,670
sweet all right you heard it here first i mean okay so it's 11 i know it's like right at the top

725
00:57:41,670 --> 00:57:46,630
of the hour want to be respectful of your time i mean i've still got like a full list of things i

726
00:57:46,630 --> 00:57:51,310
think we could go on forever um maybe we just try to get something else on the schedule some more

727
00:57:51,310 --> 00:57:55,570
time if you want to and then yeah if you want to schedule another follow-up um the ai space is

728
00:57:55,570 --> 00:58:01,150
going to move quickly so we could do a follow-up too cool okay why what else is on your mind yeah

729
00:58:01,150 --> 00:58:06,650
i just have one last question i know you guys were getting into it on x the other day so maybe

730
00:58:06,650 --> 00:58:12,270
i can be just independent they're part of mediator here but um when it comes to tokens and charging

731
00:58:12,270 --> 00:58:17,450
for your ai i had this banger tweet that nobody actually liked which made me sad but

732
00:58:17,450 --> 00:58:25,250
and then and then the tweet that you just put out there without thinking like goes viral and

733
00:58:25,250 --> 00:58:31,210
you're like what the heck no for real uh but i tweeted you know it's like i feel like

734
00:58:31,210 --> 00:58:37,530
you have these charging ais charging per token is going the way of like medical billing like

735
00:58:37,530 --> 00:58:42,470
when i you know enter a prompt i have no idea how many tokens it's going to use i have no idea what

736
00:58:42,470 --> 00:58:48,070
the price of a token is i just hook it up to my credit card and like they just keep charging it

737
00:58:48,070 --> 00:58:55,430
interested on what you think about that marks if this incentives are actually aligned and if you

738
00:58:55,430 --> 00:59:00,250
think you know kind of you're piggybacking on what we previously talked on if there's a role for

739
00:59:00,250 --> 00:59:06,430
for bitcoin to play here just in terms of adding more transparency to the end user on you know what

740
00:59:06,430 --> 00:59:13,470
this stuff actually costs and then the bonus if you want to get into this uh maybe a bit more of

741
00:59:13,470 --> 00:59:19,290
a rabbit hole than than what i currently understand my hunch especially you know the the large you know

742
00:59:19,290 --> 00:59:25,150
vc back you know llm companies are just heavily subsidizing you know tokens and what they actually

743
00:59:25,150 --> 00:59:31,610
cost like i'm putting it as like a non-zero percent probability if i build all these agents

744
00:59:31,610 --> 00:59:35,770
you know a year from now or whatever it's like my ai costs are going to like 10x or whatever because

745
00:59:35,770 --> 00:59:36,990
they stopped subsidizing it.

746
00:59:37,090 --> 00:59:38,790
So just curious, I'm like your opinion,

747
00:59:38,870 --> 00:59:41,170
how do you think pricing and incentive structures

748
00:59:41,170 --> 00:59:42,530
you align in the AI world?

749
00:59:43,290 --> 00:59:45,230
Yeah, they're definitely subsidizing things.

750
00:59:45,270 --> 00:59:46,210
That's first and foremost.

751
00:59:46,510 --> 00:59:47,690
We know that there are people

752
00:59:47,690 --> 00:59:49,070
who are getting really great deals

753
00:59:49,070 --> 00:59:50,470
on their Claude subscription

754
00:59:50,470 --> 00:59:53,290
because they're basically overusing tokens

755
00:59:53,290 --> 00:59:55,090
and Claude will show them,

756
00:59:55,210 --> 00:59:57,310
oh, you used up $2,000 worth of tokens,

757
00:59:57,450 --> 00:59:58,850
but you're only paying $200 a month.

758
00:59:58,950 --> 01:00:00,590
So that is unsustainable.

759
01:00:00,810 --> 01:00:02,830
They're really just trying to grab users at that point

760
01:00:02,830 --> 01:00:04,770
and grab data from the users.

761
01:00:04,930 --> 01:00:05,630
It's not just getting users.

762
01:00:05,630 --> 01:00:08,210
It's getting all that rich data and operating off that.

763
01:00:08,390 --> 01:00:09,950
So that's how they're paying for it, basically.

764
01:00:11,650 --> 01:00:18,190
Now, the tokens themselves, it's all like coming down to electricity use.

765
01:00:18,650 --> 01:00:21,650
And that's what the token is trying to kind of represent.

766
01:00:22,170 --> 01:00:27,990
And so it really just comes down to what is the power requirements of this model and of the server that it's running on.

767
01:00:28,650 --> 01:00:31,050
And I mean, that's kind of where Bitcoin comes in the conversation.

768
01:00:31,050 --> 01:00:34,210
We always talk about how Bitcoin is basically electricity.

769
01:00:34,210 --> 01:00:39,210
It's power. It's stored power representation and money.

770
01:00:39,790 --> 01:00:41,230
However you want to phrase it.

771
01:00:41,990 --> 01:00:42,910
It's a battery.

772
01:00:43,610 --> 01:00:45,010
Yes, it's a battery, basically.

773
01:00:45,610 --> 01:00:48,230
So, yeah, it's a financial battery.

774
01:00:49,150 --> 01:00:53,690
But I think that's where this is going to continue to go.

775
01:00:53,910 --> 01:01:00,270
We're going to try and figure out how to best estimate electricity usage for an LLM.

776
01:01:00,270 --> 01:01:05,710
And, you know, Jordan, you talk about going into planning mode and like planning something out before you set it off on a task.

777
01:01:05,830 --> 01:01:15,790
I think that that's really important. And that's going to continue to get better over time that that AIs and LLMs are going to get better at trying to figure out how can I do this in the most efficient way.

778
01:01:16,530 --> 01:01:23,430
And that's actually where local LLMs can come in, too. So people could run a really small model on their device and the software can do this for you.

779
01:01:23,430 --> 01:01:30,910
You don't have to like think about it, but the software could run a small model locally, help you do this planning in a way that doesn't need like the big brain AIs.

780
01:01:31,310 --> 01:01:35,830
And then it'll be OK. We got everything set up. Now let's go fire it off to the big brain AI.

781
01:01:36,150 --> 01:01:41,330
And maybe it can even give you an estimate before it does that. Right. It can say, here's the task we're going to have it go do.

782
01:01:41,450 --> 01:01:46,250
We think it should cost this many tokens roughly. I think we'll continue to get there.

783
01:01:46,310 --> 01:01:50,530
Just kind of like the fee market when you want to send a Bitcoin transaction, it estimates how much it thinks the fee will be.

784
01:01:50,530 --> 01:01:55,270
I think we could probably get to a point where where AIs start to estimate how much fees will be.

785
01:01:55,670 --> 01:02:00,070
We're not quite there yet, but I don't see why we couldn't get get there.

786
01:02:00,690 --> 01:02:04,090
I mean, I'm thinking like a construction project that's like, all right, here's the bid.

787
01:02:04,290 --> 01:02:07,310
Right. You've got your you've got your subcontractors.

788
01:02:07,410 --> 01:02:10,390
They all say, all right, this is what I think it's going to cost.

789
01:02:10,830 --> 01:02:15,530
OK, there's a change order because the scope increased or this went wrong.

790
01:02:16,030 --> 01:02:17,990
That might be really interesting.

791
01:02:17,990 --> 01:02:42,970
Well, Mark, I really appreciate the conversation and I also really appreciate your openness to coming back on the show. I think that our audience will be well served by this conversation and many more. So I guess where can people find you online? Where can they find Maple to go in and continue to tinker with new tools that are more private focused, privacy focused?

792
01:02:42,970 --> 01:02:49,150
yeah first and foremost go to our website try maple.ai you can also find us on all the app

793
01:02:49,150 --> 01:02:54,030
stores maple.ai and it's an icon of a green leaf with like a little gradient inside of it

794
01:02:54,030 --> 01:02:59,730
check it out it's free to download it it's free to start using you get you get really a really

795
01:02:59,730 --> 01:03:05,170
capable experience just on the free plan but then if you want to get this web search private web

796
01:03:05,170 --> 01:03:08,550
search with live data if you want to be able to upload documents images and stuff then you

797
01:03:08,550 --> 01:03:13,610
you subscribe you get the better models if you subscribe as well um i personally think it's well

798
01:03:13,610 --> 01:03:19,150
worth it uh and so i would go start there try maple.ai and if you want to follow me i'm on x

799
01:03:19,150 --> 01:03:28,710
and nostr on x i'm at mark suman and nostr i am marks at primal.net perfect wide anything else on

800
01:03:28,710 --> 01:03:32,590
your end no thank you for everything you do mark i think it's awesome excited to watch you guys

801
01:03:32,590 --> 01:03:37,930
grow cool perfect all righty thanks guys tuned everyone yeah thanks so much all right we'll talk

802
01:03:37,930 --> 01:03:39,290
you soon. Okay, bye.

803
01:03:44,690 --> 01:03:49,950
We appreciate you joining us. If you want to help build the future of Bitcoin financial services,

804
01:03:50,370 --> 01:03:55,110
hit subscribe, leave a review, sign up for a newsletter, and share this with someone in your

805
01:03:55,110 --> 01:03:59,670
network who's integrating Bitcoin into their practice. The more professionals we reach,

806
01:04:00,150 --> 01:04:03,130
the stronger this ecosystem becomes, and our clients are better served.

807
01:04:03,130 --> 01:04:06,230
Stay sharp, stay sovereign, keep clean from the front.

808
01:04:07,930 --> 01:04:37,910
Thank you.