1
00:00:00,240 --> 00:00:04,000
Welcome back to Data Driven, the podcast that dives into the collision

2
00:00:04,000 --> 00:00:07,800
of data technology and occasionally geopolitics with

3
00:00:07,800 --> 00:00:11,600
the finesse of a caffeinated quantum computer. In this episode,

4
00:00:11,680 --> 00:00:15,120
Frank Lavine is joined once again by Christopher Nuland,

5
00:00:15,120 --> 00:00:18,840
AI technical marketing maestro at Red Hat, for a no holds

6
00:00:18,840 --> 00:00:22,240
barred breakdown of America's freshly minted AI action plan.

7
00:00:22,560 --> 00:00:25,920
From Cold War vibes and AI sovereignty to the CHiPs Act,

8
00:00:26,260 --> 00:00:30,020
GPU geopolitics, and existential musings on large language

9
00:00:30,020 --> 00:00:33,780
models, this episode has more hot takes than a GPU server farm

10
00:00:33,780 --> 00:00:37,500
in July. Plus, we debate whether Europe can still flex

11
00:00:37,500 --> 00:00:41,020
its AI muscle without surrendering to Silicon Valley, and whether

12
00:00:41,020 --> 00:00:44,860
AI models will ever truly think or just continue to be unreasonably

13
00:00:44,860 --> 00:00:48,540
effective spreadsheet goblins. So buckle up, data

14
00:00:48,540 --> 00:00:52,140
disciples. This one's dense, dynamic, and

15
00:00:52,140 --> 00:00:54,900
dangerously nerdy. Let's get into it.

16
00:00:59,040 --> 00:01:02,720
All right, that bouncy little pop number. That is a fun

17
00:01:02,720 --> 00:01:06,480
fact. AI generated can only mean one thing. It's time for

18
00:01:06,560 --> 00:01:10,160
a new edition of Frank's World TV Live or

19
00:01:10,160 --> 00:01:13,959
an episode of Data Driven, depending on where and how you're listening, slash

20
00:01:13,959 --> 00:01:17,412
watching. You can catch me at the following URLs. Franksworld.com

21
00:01:17,548 --> 00:01:20,900
data driven tv and impact

22
00:01:21,126 --> 00:01:24,248
quantum.com speaking of impact

23
00:01:24,392 --> 00:01:27,840
quantum.com my co host and I have launched another

24
00:01:27,920 --> 00:01:31,300
book and it's basically Quantum

25
00:01:31,300 --> 00:01:34,820
Curious, the Gateway to the Next Computing Revolution.

26
00:01:35,060 --> 00:01:38,660
And what that is is we basically took the third, the first

27
00:01:38,900 --> 00:01:42,580
13 some odd episodes of the season and distilled it down

28
00:01:42,580 --> 00:01:46,380
into a little book. It's 2.99 on Amazon, but if you join

29
00:01:46,380 --> 00:01:50,220
our mailing list, it's completely free. So go to Impact Quantum or scan

30
00:01:50,220 --> 00:01:53,860
the QR code to find out more. All right, now that I've gotten the

31
00:01:54,180 --> 00:01:57,890
commercialism ism out of the way, I'd like to welcome

32
00:01:58,050 --> 00:02:01,890
back our guest, Christopher Nuland, who is

33
00:02:01,970 --> 00:02:05,570
a peer of mine on the same team, an AI

34
00:02:05,570 --> 00:02:09,170
technical marketing manager at Red Hat. How's it going? Good,

35
00:02:09,250 --> 00:02:12,450
good. Glad to be back. I think

36
00:02:13,090 --> 00:02:16,770
we ended the last talk on kind of a cliffhanger, and then

37
00:02:17,250 --> 00:02:21,090
I think some recent news has really built on top of some of that

38
00:02:21,170 --> 00:02:25,010
previous conversation. So I'm happy to be here talking about some big things

39
00:02:25,010 --> 00:02:28,800
that are going on in the the area of AI right now. Absolutely. So

40
00:02:28,800 --> 00:02:32,520
over the weekend, the Trump administration dropped the thought on

41
00:02:32,520 --> 00:02:36,000
the Amer AI's America's AI Action

42
00:02:36,000 --> 00:02:38,080
Plan. I think somebody likes alliterations.

43
00:02:40,400 --> 00:02:44,080
But so, and you and I were chatting about this over

44
00:02:44,080 --> 00:02:47,600
Slack and, and, and, and you had some thoughts on this. So, like, what.

45
00:02:48,880 --> 00:02:52,560
And you had some interesting ideas around it, and some surprises are in the bell.

46
00:02:52,560 --> 00:02:56,330
So let's. Let's get into it. Yeah. So

47
00:02:56,890 --> 00:02:59,530
I think overall, this is really needed.

48
00:03:00,490 --> 00:03:03,930
We've seen a couple things like this come out for some other

49
00:03:04,810 --> 00:03:08,490
countries globally. The EU has

50
00:03:11,290 --> 00:03:14,730
this. I'd say the one from the US is more of a set of guidelines,

51
00:03:14,730 --> 00:03:18,450
or the one from EU is actually some laws that they're trying to pass that

52
00:03:18,450 --> 00:03:22,110
have some similar tone to this one. You know, we're

53
00:03:22,110 --> 00:03:25,950
seeing things out of the uk, out of Singapore, other, you

54
00:03:25,950 --> 00:03:29,470
know, other nations that really are trying to get an

55
00:03:29,550 --> 00:03:32,750
idea of what is their strategy around AI

56
00:03:32,750 --> 00:03:36,590
sovereignty. And this, to me, is a document

57
00:03:36,590 --> 00:03:39,630
more about AI sovereignty than anything else.

58
00:03:40,350 --> 00:03:43,390
It's really about how. How does the US

59
00:03:44,750 --> 00:03:48,580
go into the next phase of really almost

60
00:03:48,580 --> 00:03:51,940
like a new industrial revolution around AI and

61
00:03:52,340 --> 00:03:55,540
this document's really outlining the plan.

62
00:03:56,500 --> 00:03:59,540
I think overall, I thought it was pretty well thought out, and we'll go through

63
00:03:59,540 --> 00:04:02,740
and kind of pick apart some of the key areas. But I think

64
00:04:02,979 --> 00:04:06,020
overall the. The key tone here was about

65
00:04:07,220 --> 00:04:10,820
AI sovereignty, so specifically within the US

66
00:04:10,900 --> 00:04:14,680
and how we're going to be managing that. And overall,

67
00:04:14,680 --> 00:04:17,480
I thought it was. It was good. I know. You know, when we were talking

68
00:04:17,480 --> 00:04:21,120
on Slack, we were talking about how there's definitely a lot there about China

69
00:04:21,120 --> 00:04:24,720
as well. Right. In a weird way,

70
00:04:24,720 --> 00:04:28,560
I. I felt like this document was a bit of a. A declaration

71
00:04:28,560 --> 00:04:32,280
of. Of war in a way, because in

72
00:04:32,280 --> 00:04:35,360
a document, it outlines that

73
00:04:36,000 --> 00:04:38,880
they really consider this now like a Cold War with.

74
00:04:40,260 --> 00:04:43,740
With China around AI and what I thought was so

75
00:04:43,740 --> 00:04:47,140
fascinating is I kept going back to this analogy of, like,

76
00:04:47,620 --> 00:04:50,740
the Cold War arms race of Russia and how

77
00:04:51,780 --> 00:04:55,380
we need to do certain things around AI because we basically need to

78
00:04:55,619 --> 00:04:59,220
mimic what the United States was able to achieve during the

79
00:04:59,220 --> 00:05:02,260
Cold War. And I think that sat with me

80
00:05:02,900 --> 00:05:05,620
because I think, you know, even last time I was on here, we were talking

81
00:05:05,620 --> 00:05:09,420
about how we. We basically. No, there's a Cold War kind of thing going on,

82
00:05:10,060 --> 00:05:13,780
but it was. It was different seeing

83
00:05:13,780 --> 00:05:17,060
it. You know, you and I were talking earlier, like, on the letterhead. It was.

84
00:05:17,060 --> 00:05:20,820
Yeah, it's different. You know, like, there's. There's things that are. Obviously, you

85
00:05:20,820 --> 00:05:24,140
can see with your own eyes, but it's quite a different thing when something appears

86
00:05:24,140 --> 00:05:27,900
on official White House letterhead. Right. Or even,

87
00:05:27,900 --> 00:05:31,740
you know, government letterhead. I think that's. It's an interesting,

88
00:05:33,100 --> 00:05:36,780
interesting shift. And this whole idea of a Cold War between the US

89
00:05:36,860 --> 00:05:40,700
and China and AI is Not a new concept.

90
00:05:40,700 --> 00:05:44,140
Right. I. There's a really good book and I'm gonna see if I can share

91
00:05:44,140 --> 00:05:47,980
this tab real quick. This is an excellent book. It's an

92
00:05:47,980 --> 00:05:51,580
excellent audiobook too. There you go.

93
00:05:51,580 --> 00:05:55,340
Oh, there we are. Sorry everyone, my dogs are barking.

94
00:05:57,420 --> 00:06:01,120
But this book came out in 2018 and a

95
00:06:01,120 --> 00:06:04,800
lot of what he predicted has come to pass.

96
00:06:05,600 --> 00:06:09,440
And it seems to me like the authors of this document have also,

97
00:06:12,000 --> 00:06:15,400
if not read the book or listened to the book, have at least seen the

98
00:06:15,400 --> 00:06:19,080
Cliff Notes version of it. Right. Like this, if you really think

99
00:06:19,080 --> 00:06:22,800
about it, there's really only two major players right now in

100
00:06:23,200 --> 00:06:26,320
the AI space and we're going to alienate a lot of people in the eu.

101
00:06:26,320 --> 00:06:28,860
Right. But saying that, right, it's really largely

102
00:06:30,780 --> 00:06:33,660
US and China, right? Yes. And

103
00:06:35,820 --> 00:06:39,500
not to say that the EU is not in the games, clearly, Mistral. And

104
00:06:39,500 --> 00:06:42,180
apparently there's a rumor, I don't know if you heard this rumor that, that Apple

105
00:06:42,180 --> 00:06:45,340
is considering buying Mistral. I have heard some of those.

106
00:06:45,900 --> 00:06:49,500
So again, I don't own any of Apple stock or whatever, so I'm not. Or.

107
00:06:50,060 --> 00:06:53,780
But I think it's an interesting con, Interesting idea if they were to do

108
00:06:53,780 --> 00:06:57,620
that because clearly that would, I wonder how that

109
00:06:57,620 --> 00:07:00,940
would shift kind of the balance of, if not power,

110
00:07:01,020 --> 00:07:04,660
perceived power. Right. Because Apple

111
00:07:04,660 --> 00:07:08,340
obviously is a stalwart of Silicon Valley and if, you know,

112
00:07:08,340 --> 00:07:12,140
if Europe's major, you know, every

113
00:07:12,140 --> 00:07:15,900
time you talk about the EU falling behind, they always say, what

114
00:07:15,900 --> 00:07:19,700
about Mistral? Right. So if Mistral ends up getting purchased by, you

115
00:07:19,700 --> 00:07:23,080
know, Apple, that would be, that would.

116
00:07:23,320 --> 00:07:27,000
I think there'd be a lot of drama about that. Yeah, I think so

117
00:07:27,000 --> 00:07:30,680
too. I think it, it really just shows there's a line in the

118
00:07:30,680 --> 00:07:34,040
sand between the two major superpowers here, between

119
00:07:34,440 --> 00:07:37,320
China and the United States. My

120
00:07:37,320 --> 00:07:41,040
speculation there is that there might

121
00:07:41,040 --> 00:07:44,720
be something official there, but that the EU might step in

122
00:07:44,720 --> 00:07:47,720
and just say no. Right. To that.

123
00:07:48,200 --> 00:07:51,960
Simply given what we're here talking about, AI sovereignty. And what does

124
00:07:51,960 --> 00:07:55,720
AI sovereignty look like? I, I don't think the, the French

125
00:07:55,720 --> 00:07:58,680
necessarily want to give up that and I don't think the EU wants to give

126
00:07:58,680 --> 00:07:59,120
that up

127
00:08:02,720 --> 00:08:06,320
from an open source standpoint. We're still seeing a lot of thought leadership

128
00:08:06,320 --> 00:08:10,000
coming out of the eu, even though there's not

129
00:08:10,000 --> 00:08:13,360
enough, what I would say, enterprise momentum there.

130
00:08:13,680 --> 00:08:17,390
Right. There's still a lot of research institutes there. There's still

131
00:08:17,390 --> 00:08:21,110
a lot of, even some smaller form companies

132
00:08:21,110 --> 00:08:24,870
like even like Hugging Face and Mistral for example, are, you know,

133
00:08:25,030 --> 00:08:28,870
EU based that have made A big impact and are very open

134
00:08:28,870 --> 00:08:32,710
source heavy and but at the end of

135
00:08:32,710 --> 00:08:36,310
the day they're just still very small when you

136
00:08:36,310 --> 00:08:39,750
consider these behemoth organizations like Microsoft,

137
00:08:39,990 --> 00:08:43,679
Amazon, Nvidia, Apple, all the

138
00:08:43,679 --> 00:08:46,639
fang corporations which have,

139
00:08:47,519 --> 00:08:51,319
that can really throw their weight around. And we've seen a

140
00:08:51,319 --> 00:08:54,559
lot of, a lot of

141
00:08:55,199 --> 00:08:57,839
startups like OpenAI that has like

142
00:08:58,639 --> 00:09:02,279
climbed up now into the upper epsilon and

143
00:09:02,279 --> 00:09:05,599
that's being really driven by American industry. So

144
00:09:05,919 --> 00:09:09,500
and that's just something the EU can't prop up as as much.

145
00:09:09,500 --> 00:09:13,020
But I still think they, they're a major

146
00:09:13,020 --> 00:09:16,380
player. They may not be necessarily

147
00:09:17,500 --> 00:09:21,180
one to one with China and America, but if there was a

148
00:09:21,180 --> 00:09:24,899
second tier right under that, it would be the eu. Yeah, I

149
00:09:24,899 --> 00:09:27,820
can see that. I also think it's too early to count them out of the

150
00:09:27,820 --> 00:09:31,620
race. Right. Like, yeah, you know we're, this is the start of the

151
00:09:31,620 --> 00:09:35,420
marathon. Right. So they're clearly, clearly there are two front runners.

152
00:09:35,420 --> 00:09:38,790
But I don't, I don't, I wouldn't count them entirely out just yet. Right.

153
00:09:39,590 --> 00:09:42,670
And I didn't know Hugging Face was a European company. I thought they were based

154
00:09:42,670 --> 00:09:45,830
out of New York. But that must be their American. I

155
00:09:46,150 --> 00:09:49,270
believe you may be right. Let me show Hugging Face.

156
00:09:53,990 --> 00:09:56,150
I know. I think the founders are European.

157
00:09:57,830 --> 00:10:01,430
You are correct. Founders are European, but they are based out of

158
00:10:01,430 --> 00:10:04,870
America. And that just goes to show right there.

159
00:10:05,930 --> 00:10:09,530
Yo. That the gravity of just American enterprise. That

160
00:10:10,650 --> 00:10:13,930
you can shell out a lot of money to get, to get talent. Right. And

161
00:10:13,930 --> 00:10:17,610
yeah, this was the thing that a European tech founder said. Right. So you know,

162
00:10:17,610 --> 00:10:21,410
all the Europeans don't hate on me, but there was a lot of founders that

163
00:10:21,410 --> 00:10:25,090
they'll end up moving to Dubai. Right. Bootstrap and

164
00:10:25,090 --> 00:10:28,610
then move to Silicon Valley, you know, at some point. Right. Like,

165
00:10:28,610 --> 00:10:31,620
so I think the, the

166
00:10:32,340 --> 00:10:34,100
European Union as a whole

167
00:10:35,940 --> 00:10:39,780
has to address some systemic shortcomings when it comes

168
00:10:39,780 --> 00:10:43,180
to a venture capital and startup

169
00:10:43,180 --> 00:10:47,020
pipeline. Right. And I hope I, I

170
00:10:47,020 --> 00:10:50,100
think that they'll get it figured out. I just don't think that they're going to

171
00:10:50,100 --> 00:10:53,900
get it figured out this year. They might

172
00:10:53,900 --> 00:10:56,740
get it figured out by the end of the decade because I think that,

173
00:10:58,110 --> 00:11:01,070
you know, just a little bit of back of the napkin math, right. You, you

174
00:11:01,070 --> 00:11:04,750
know, it's, it's, you can see

175
00:11:04,750 --> 00:11:08,590
that growing the tax base is good for everyone

176
00:11:09,310 --> 00:11:12,910
and this is one way to do that. And if you have your

177
00:11:12,910 --> 00:11:16,630
brain drain, which we'll get into that, that term, you know, either

178
00:11:16,630 --> 00:11:20,390
going to Dubai, you know, Silicon Valley or New

179
00:11:20,390 --> 00:11:24,060
York, it's not Good. Right. Because

180
00:11:24,060 --> 00:11:27,860
you're, you're basically, you're educating them in country. Right.

181
00:11:27,860 --> 00:11:31,420
And a lot of these countries have, you know, cheaper, you know, free tuition.

182
00:11:31,580 --> 00:11:35,340
Yeah. So you're paying for the talent, you're training up the talent, you're

183
00:11:35,340 --> 00:11:38,859
paying for talent. And then when time comes in to cash in on the return

184
00:11:38,859 --> 00:11:41,500
on that investment, if you want to look at it that way, throughout of country.

185
00:11:41,500 --> 00:11:45,300
Right. So what are you going to do? I think that it's

186
00:11:45,300 --> 00:11:49,040
in the EU's best interest to fix that problem. And like

187
00:11:49,040 --> 00:11:52,760
you said, like sovereign AI or is a big deal.

188
00:11:52,760 --> 00:11:56,200
And sovereign AI is different than sovereign than data sovereignty, right?

189
00:11:56,360 --> 00:11:59,960
Yep. It's the. And I don't think people really kind of gotten their head around

190
00:11:59,960 --> 00:12:03,320
that. So I know what my definition is of that,

191
00:12:04,199 --> 00:12:07,920
you know, is the idea that. And it's even called out in this action

192
00:12:07,920 --> 00:12:11,560
plan report. Right. Where it's like, you know, AI with American values. Right.

193
00:12:11,640 --> 00:12:15,310
Yeah. And like you said, I'm pretty sure the French want to have, you

194
00:12:15,310 --> 00:12:19,030
know, AI with French values, and the

195
00:12:19,030 --> 00:12:22,830
Germans probably want to have, you know, with German values. Right. So I think even,

196
00:12:22,990 --> 00:12:26,790
even painting the entire continent, even though everything's is kind of done

197
00:12:26,790 --> 00:12:30,550
through the European Union, I don't think that's. That might

198
00:12:30,550 --> 00:12:34,310
be at their detriment. Right. And. And the German market is also pretty

199
00:12:34,310 --> 00:12:37,830
sizable too. Right. It's something like 80 million people. Right. And the

200
00:12:37,830 --> 00:12:40,510
German language market, I think, adds another 20

201
00:12:41,450 --> 00:12:45,250
million to that. Right. So, you know, I only say that because one

202
00:12:45,250 --> 00:12:48,810
of the, one of the points that people made for taking German in high school

203
00:12:49,290 --> 00:12:52,970
was it was 100 million ish, you know, number of people speak the

204
00:12:52,970 --> 00:12:56,410
language. So it's not. And, and I would say that,

205
00:12:57,610 --> 00:13:01,290
you know, particularly when we're dealing with language models. Right. It's in the

206
00:13:01,290 --> 00:13:05,130
name. Right. So language and culture, although not exactly the same,

207
00:13:06,170 --> 00:13:10,010
are very much tightly linked. And that was something we talked about last

208
00:13:10,010 --> 00:13:13,710
time. We got sidetracked, but that's what

209
00:13:13,710 --> 00:13:15,830
I do here. That's fine.

210
00:13:17,270 --> 00:13:20,470
What struck out of you, the report? I think one of the things you mentioned

211
00:13:20,470 --> 00:13:23,990
was, well, go ahead, I'll let you go. Sure.

212
00:13:26,309 --> 00:13:29,950
The thing that I was most surprised about was

213
00:13:29,950 --> 00:13:33,470
that pillar, one of the

214
00:13:33,470 --> 00:13:37,270
document on page six and a couple of other areas was

215
00:13:37,270 --> 00:13:40,700
really focused on the workforce

216
00:13:41,580 --> 00:13:45,420
and this concept of like, securing the AI workforce,

217
00:13:46,140 --> 00:13:49,500
making sure to have necessary people

218
00:13:50,460 --> 00:13:54,260
in play. And then it got into like almost this Cold

219
00:13:54,260 --> 00:13:58,100
war kind of mentality

220
00:13:58,100 --> 00:14:01,620
of like, how do we make sure that we can trust the people that we

221
00:14:01,620 --> 00:14:05,300
have and it was, it was just surprising to me because I thought it was

222
00:14:05,300 --> 00:14:09,110
going to be more about the regulation of AI models, which it does move

223
00:14:09,110 --> 00:14:12,190
into eventually. Right. And supply chain security.

224
00:14:12,670 --> 00:14:16,470
But the, the concept of, of workforce and the fact that it

225
00:14:16,470 --> 00:14:20,310
was the first pillar was intriguing to me. It got me

226
00:14:20,310 --> 00:14:24,070
thinking about kind of what, what is the US Administration thinking about right now? And

227
00:14:24,070 --> 00:14:27,670
they're, I think they're really thinking about making

228
00:14:27,670 --> 00:14:30,510
sure they lock down the people. And

229
00:14:31,550 --> 00:14:35,070
for good reasons and, and probably bad reasons too. You know, good reason is,

230
00:14:35,520 --> 00:14:39,280
you know, how do we entice the best experts

231
00:14:39,280 --> 00:14:42,480
to stay here in America? How do we entice

232
00:14:42,960 --> 00:14:46,560
the workforce to continue to move into the area of AI through

233
00:14:46,720 --> 00:14:50,480
education? But then there also seemed to

234
00:14:50,480 --> 00:14:53,920
be almost like a Cold War vibe there.

235
00:14:54,240 --> 00:14:58,040
I don't know if you watched the movie Oppenheimer, but it kind of reminded me

236
00:14:58,040 --> 00:15:01,640
in that movie where the people working on the

237
00:15:01,640 --> 00:15:05,290
Manhattan Project were like their personal lives

238
00:15:05,290 --> 00:15:08,570
were, were under view quite a bit.

239
00:15:09,050 --> 00:15:12,890
And it, it kind of reminded me of that. Like is, is in,

240
00:15:12,890 --> 00:15:16,730
in a year or two are all the AI experts going to, you know, have

241
00:15:16,730 --> 00:15:20,410
the NSA and the FBI like keeping track of them

242
00:15:20,410 --> 00:15:24,250
and what they're doing? And it doesn't explicitly

243
00:15:24,250 --> 00:15:27,970
say that, but the tone kind of led me to think, oh, wow,

244
00:15:27,970 --> 00:15:31,460
they're, they, they're really interested in what these people are doing.

245
00:15:32,180 --> 00:15:35,940
It's not just about the technology, but the people making the technology.

246
00:15:36,100 --> 00:15:39,780
And that was very intriguing to me. I could see that being

247
00:15:39,780 --> 00:15:43,500
a thing, especially if there's an actual honest to God, you know,

248
00:15:43,500 --> 00:15:47,060
old school knockdown, drag out shooting war with any

249
00:15:47,060 --> 00:15:50,780
country. I could

250
00:15:50,780 --> 00:15:54,180
see that being, I wouldn't say nationalized, but

251
00:15:54,840 --> 00:15:58,040
you'll have to get some kind of. Even now, like if you work in the

252
00:15:58,040 --> 00:16:01,680
nuclear industry, you need a queue clearance. You need a lot of things being. You

253
00:16:01,680 --> 00:16:05,360
need a lot of invasive, not procedures, but

254
00:16:05,360 --> 00:16:08,760
definitely a lot of invasive paperwork and investigations that,

255
00:16:09,480 --> 00:16:12,520
But I, and I, and I do see,

256
00:16:13,160 --> 00:16:16,400
I didn't read the whole thing yet, but I did, I did, I did feed

257
00:16:16,400 --> 00:16:20,160
it through Notebook lm. I did listen to that. I did do some skimming

258
00:16:20,160 --> 00:16:23,450
of it. And that was one of the things was like it seems like they're

259
00:16:23,450 --> 00:16:26,570
laying the groundwork for that in case things get sideways.

260
00:16:27,370 --> 00:16:30,250
Also part of that is the, the

261
00:16:30,730 --> 00:16:34,010
securing the supply chain from the silicon on up.

262
00:16:34,490 --> 00:16:36,490
Yeah. Which is a smart thing because

263
00:16:39,290 --> 00:16:42,650
the chips are made in very limited

264
00:16:42,730 --> 00:16:44,650
geolocations. Right. So one,

265
00:16:46,420 --> 00:16:50,220
one major international incident, a shooting

266
00:16:50,220 --> 00:16:53,820
war. Right. No matter who wins. You know, there's Going to be an

267
00:16:53,820 --> 00:16:57,500
island where most of the stuff is made. Yeah. That's going to be reduced to

268
00:16:57,500 --> 00:17:01,300
the rubble. Right. Now, whose flag gets planted on top of that rubble,

269
00:17:01,460 --> 00:17:04,820
you know, remains to be seen. But you know, you, you know,

270
00:17:05,700 --> 00:17:07,940
so much for the chip manufacturers there. Yeah.

271
00:17:09,620 --> 00:17:13,180
And also too, you can't rule out natural disasters. Right. You know,

272
00:17:13,180 --> 00:17:16,940
2000, was it 2005 there was. Or 2004 there was the massive

273
00:17:16,940 --> 00:17:20,020
tsunami that, you know, did major

274
00:17:20,500 --> 00:17:24,020
swath of damage. It's not impossible to imagine even just a natural

275
00:17:24,020 --> 00:17:27,820
disaster, bad typhoon either. Earthquake with

276
00:17:27,820 --> 00:17:31,620
Japan, wake Japan. I mean it, it could,

277
00:17:31,780 --> 00:17:35,620
it's not impossible to imagine like more than one way for it

278
00:17:35,620 --> 00:17:39,420
to rain on everybody's parade. And if you think supply chain issues with GPUs

279
00:17:39,420 --> 00:17:42,220
are rough today. Yeah, I mean that's just,

280
00:17:43,180 --> 00:17:46,860
that would be a big thing. But what really stood

281
00:17:46,860 --> 00:17:50,700
out to me, and obviously I'm biased because my wife is a federal employee, was

282
00:17:50,700 --> 00:17:53,980
talking about training federal employees to use AI. Right. And

283
00:17:54,300 --> 00:17:56,620
there was, even in there, even in there there was a

284
00:17:58,140 --> 00:18:01,740
accelerating adoption here, but basically mandating

285
00:18:03,020 --> 00:18:06,540
employee access for federal employees and

286
00:18:06,540 --> 00:18:09,980
training on these LLMs,

287
00:18:10,540 --> 00:18:14,100
which is interesting because I can

288
00:18:14,100 --> 00:18:17,540
speak from not first hand experience, but certainly, you know,

289
00:18:17,540 --> 00:18:21,340
secondhand experience. Right. Federal employees do not feel loved

290
00:18:21,340 --> 00:18:25,139
and appreciated, let alone have access to any kind of training or

291
00:18:25,139 --> 00:18:28,060
anything like that. So I thought that was interesting,

292
00:18:30,780 --> 00:18:34,220
that was interesting in there because it's been a rough go for

293
00:18:34,220 --> 00:18:38,060
feds the last six, seven months. Oh yeah. Everything's

294
00:18:38,060 --> 00:18:40,780
been very negative. And this is like one of the, maybe the first,

295
00:18:42,300 --> 00:18:46,060
it's first positive, you. Know, and I was, I was telling

296
00:18:46,060 --> 00:18:49,300
you in the virtual green room is that, you know, the agency my wife works

297
00:18:49,300 --> 00:18:52,460
at, like they're not hiring new people,

298
00:18:53,420 --> 00:18:57,220
but they're creating a new organization that people will be doubling down on their duties,

299
00:18:57,220 --> 00:19:00,060
which presumably they'll get access to the training. And she,

300
00:19:00,780 --> 00:19:04,580
she may or may not be involved with that yet. We don't know. But, but

301
00:19:04,580 --> 00:19:08,180
it's interesting to kind of see that, see that

302
00:19:08,180 --> 00:19:11,860
happening. But yeah. What else have you, what else

303
00:19:11,860 --> 00:19:15,380
took. It's, you know, so I think the most

304
00:19:15,380 --> 00:19:18,540
important thing that was in there and I,

305
00:19:19,020 --> 00:19:22,300
I actually figured the whole document would be about this is

306
00:19:22,780 --> 00:19:26,300
around supply chain security. So if,

307
00:19:26,780 --> 00:19:30,510
if people aren't aware when we're talking about supply chain, typically we talk

308
00:19:30,510 --> 00:19:34,230
about supply chain, it's more in industry terms of, you know, how

309
00:19:34,230 --> 00:19:38,030
does something get made, the nuts and bolts, where does the raw

310
00:19:38,030 --> 00:19:41,630
materials come from? That term was

311
00:19:41,630 --> 00:19:45,349
never used really in technology until recently.

312
00:19:45,670 --> 00:19:49,310
And. Probably the pandemic

313
00:19:49,310 --> 00:19:52,070
is when most people first heard the term supply chain.

314
00:19:53,590 --> 00:19:57,160
It was, it was the Solar Winds hack. I think that

315
00:19:57,160 --> 00:20:00,400
also really, yes, put it in perspective too.

316
00:20:00,880 --> 00:20:03,920
So those who aren't aware, there's a company called SolarWinds,

317
00:20:04,800 --> 00:20:08,600
they were very predominantly used in the government. I think

318
00:20:08,600 --> 00:20:12,160
they still are. But there was a

319
00:20:12,160 --> 00:20:15,920
hack where instead of hacking their software directly,

320
00:20:15,920 --> 00:20:19,040
they hacked the supply chain. They injected

321
00:20:19,520 --> 00:20:22,880
bad code early on into the supply chain

322
00:20:23,560 --> 00:20:27,280
and that slowly propagated out to these

323
00:20:27,280 --> 00:20:30,880
different government agencies. And the scary part

324
00:20:30,880 --> 00:20:34,640
is that the very thing that was meant to monitor these

325
00:20:34,640 --> 00:20:38,400
type of situations was the thing that had gotten infiltrated. So it

326
00:20:38,400 --> 00:20:42,000
took a while for anyone to even know. And it was

327
00:20:42,000 --> 00:20:45,000
massive. It impacted the government and impacted

328
00:20:45,320 --> 00:20:48,920
enterprise. And that is where

329
00:20:49,240 --> 00:20:52,600
I think NIST and a couple other agencies made the decision, okay,

330
00:20:52,980 --> 00:20:56,820
we're going to come up with a requirement of what supply chain looks like

331
00:20:56,900 --> 00:21:00,180
within these types of software development

332
00:21:00,500 --> 00:21:04,300
process. And really gets into, okay, all the way

333
00:21:04,300 --> 00:21:08,060
from how do we think up an idea for

334
00:21:08,060 --> 00:21:11,540
code, how do we submit that code

335
00:21:12,020 --> 00:21:15,780
into a repository, how do we compile it, how do

336
00:21:15,780 --> 00:21:19,630
we scan it, how do we distribute it? And that's when we talk about supply

337
00:21:19,630 --> 00:21:23,430
chain, secure supply chain. That's in the context of what

338
00:21:23,430 --> 00:21:27,270
we mean. And that relates directly to AI as well, because it's

339
00:21:27,270 --> 00:21:31,030
all data pipelines. And for AI specifically, it's about where does

340
00:21:31,030 --> 00:21:33,950
that data come from, where was it sourced,

341
00:21:36,110 --> 00:21:39,830
when was it added into our model? How

342
00:21:39,830 --> 00:21:43,350
can we prove that the model that we built over

343
00:21:43,350 --> 00:21:45,960
here is the model that's running over here?

344
00:21:47,000 --> 00:21:50,840
So if the government has an officially blessed model, how do

345
00:21:50,840 --> 00:21:53,320
I know the model that's running within

346
00:21:54,920 --> 00:21:58,640
my defense contract firm is that model? And that gets

347
00:21:58,640 --> 00:22:02,360
all into this supply chain. And I was happy to see that some of it

348
00:22:02,440 --> 00:22:05,400
wasn't as technically laid out as I wanted it to be.

349
00:22:06,200 --> 00:22:09,920
The document really just says we're relying on NIST and some

350
00:22:09,920 --> 00:22:13,550
other government agencies to come up with a plan.

351
00:22:13,870 --> 00:22:17,630
So this wasn't really the plan, it's more the actual call to action for

352
00:22:17,630 --> 00:22:21,230
the plan. But it was good to see that there. It was important for it

353
00:22:21,230 --> 00:22:24,990
to be there. I was happy that that was highlighted. And I think

354
00:22:25,790 --> 00:22:28,830
in terms of security, it's the most

355
00:22:28,830 --> 00:22:32,430
underappreciated one right now. Everyone's really focused on

356
00:22:32,830 --> 00:22:36,670
model guardrails. And what we talked

357
00:22:36,670 --> 00:22:40,260
about last time with the AI 2027 report or AI

358
00:22:40,260 --> 00:22:43,860
breaking out of, of its shell. I think the most

359
00:22:43,860 --> 00:22:47,660
important thing right now is actually more of the supply chain security where you

360
00:22:47,660 --> 00:22:50,660
know, don't let people inject bad data into

361
00:22:51,540 --> 00:22:55,060
models that are making critical decisions for the government,

362
00:22:55,300 --> 00:22:59,140
for finance or healthcare. That's where our focus needs to

363
00:22:59,140 --> 00:23:02,660
be first. I think having that secure supply chain is

364
00:23:02,660 --> 00:23:05,940
ultimately what's going to lead to, to preventing

365
00:23:06,820 --> 00:23:10,580
the AI 2027 report as well. Where it'll prevent

366
00:23:10,580 --> 00:23:14,100
a breakout or if there was a breakout, it's going to reduce the blast

367
00:23:14,100 --> 00:23:16,820
radius of that type of situation.

368
00:23:17,860 --> 00:23:21,700
Now that makes a lot of sense and it's interesting because there's not just

369
00:23:21,780 --> 00:23:25,380
the traditional nation states that could be involved here. Right. There's also

370
00:23:26,500 --> 00:23:30,100
or bad actors in the normal sense. But also the

371
00:23:30,100 --> 00:23:32,900
AI itself could become a threat too. Right. Like,

372
00:23:35,010 --> 00:23:38,130
and the report doesn't isn't technical in detail,

373
00:23:38,610 --> 00:23:41,890
but I don't think that's who the audience was really for. Yeah

374
00:23:43,090 --> 00:23:46,690
but that's interesting

375
00:23:46,690 --> 00:23:50,490
because you know, I don't know like from a game

376
00:23:50,490 --> 00:23:54,250
theory point of view, right. Like you have the traditional, the usual suspects,

377
00:23:54,250 --> 00:23:58,090
right. The countries, terrorist groups, criminal gangs, blah blah,

378
00:23:58,090 --> 00:24:01,910
blah. Right. The usual kind of players. But AI also has

379
00:24:01,910 --> 00:24:05,710
the potential to become yet another player

380
00:24:05,710 --> 00:24:09,390
in the game of that. That's. I certainly

381
00:24:09,390 --> 00:24:12,830
didn't see that in the report and it didn't cross my mind until you kind

382
00:24:12,830 --> 00:24:16,430
of bridged last stream and this stream content. I was like,

383
00:24:16,430 --> 00:24:20,270
oh wow, this is multi dimensional. This is like 5 dgs or

384
00:24:20,270 --> 00:24:21,150
something like that. Yes.

385
00:24:24,510 --> 00:24:28,310
I would say for a first effort, it's actually a fairly reasonably well

386
00:24:28,310 --> 00:24:32,120
written document. For those that don't, for folks

387
00:24:32,120 --> 00:24:34,720
that don't know, I used to accompany our lobbyists

388
00:24:35,920 --> 00:24:39,440
in, in when I was at Microsoft talking about technology

389
00:24:39,600 --> 00:24:42,320
issues and things like that and you know, I was the

390
00:24:43,280 --> 00:24:45,280
technical resource for that.

391
00:24:46,880 --> 00:24:49,920
And as I was telling you, virtual green room. A lot of these elected officials,

392
00:24:50,160 --> 00:24:53,840
regardless of, you know, whether you agree with their

393
00:24:56,080 --> 00:24:59,800
party affiliation or whatnot, they're not the most technical I

394
00:24:59,800 --> 00:25:02,140
would say of the one ones I've interacted with

395
00:25:03,660 --> 00:25:06,540
which maybe, maybe 60, 70,

396
00:25:07,340 --> 00:25:09,500
some of them are names you've heard of, some of them as you've never heard

397
00:25:09,500 --> 00:25:12,620
of. I would say less than 10%

398
00:25:15,500 --> 00:25:18,940
are technical in any sense. Yeah, right.

399
00:25:19,340 --> 00:25:22,980
And there were only two that I would say like would feel

400
00:25:22,980 --> 00:25:26,620
at home having a technical conversation. I wonder how

401
00:25:26,620 --> 00:25:29,700
many of the policymakers even

402
00:25:30,100 --> 00:25:33,900
understand the term AI sovereignty. So and

403
00:25:33,900 --> 00:25:37,620
this is interesting, I'd love your opinion. Yeah, I think how many technical people

404
00:25:37,620 --> 00:25:41,340
would understand. Well, that's what I mean. Yeah, go ahead. This is where I've been

405
00:25:41,340 --> 00:25:45,140
having some conversations even within our own organization that we work

406
00:25:45,140 --> 00:25:48,420
for. There's a lot of differing opinion on what AI

407
00:25:49,380 --> 00:25:53,140
sovereignty is. A lot of people who keep talking to me about AI sovereignty,

408
00:25:53,140 --> 00:25:56,940
I realize they're more talking about clients, cloud sovereignty, they're talking about how do

409
00:25:56,940 --> 00:26:00,500
I secure the compute, all of my

410
00:26:00,500 --> 00:26:04,220
compute within my borders and can guarantee that everything is

411
00:26:04,220 --> 00:26:07,300
within those borders. Which makes sense. I mean we work for Red Hat, we work

412
00:26:07,300 --> 00:26:10,739
for a, you know, basically a cloud

413
00:26:11,860 --> 00:26:15,220
Linux based company. Right. But when we talk about AI

414
00:26:15,220 --> 00:26:18,980
sovereignty, at least me personally, it, it's an accumulation of a

415
00:26:18,980 --> 00:26:22,530
few core areas. It gets back to the data sovereignty, a little bit of that

416
00:26:22,690 --> 00:26:25,090
cloud sovereignty. But it's really about

417
00:26:26,290 --> 00:26:29,570
do my, I have control over my AI models,

418
00:26:30,210 --> 00:26:31,970
I know where the data came from.

419
00:26:33,810 --> 00:26:37,570
And I loved what you said earlier. It's about the culture of the model

420
00:26:37,890 --> 00:26:41,530
and I think ultimately the AI sovereignty is about the culture of the

421
00:26:41,530 --> 00:26:45,370
model and then making sure that you're containing your

422
00:26:45,370 --> 00:26:48,210
AI to the borders of the United States.

423
00:26:49,540 --> 00:26:53,260
So you're keeping all the secrets here, you're keeping the talent

424
00:26:53,260 --> 00:26:56,740
that are driving it. But ultimately you're right, it's about that

425
00:26:56,900 --> 00:27:00,420
culture and making sure that your model has the best

426
00:27:00,420 --> 00:27:03,540
representation of your culture. And

427
00:27:04,339 --> 00:27:07,780
it's kind of a scary thing to think about. It's an interesting topic, but

428
00:27:08,100 --> 00:27:11,940
it also gets into a lot of geopolitical challenges I think we're

429
00:27:11,940 --> 00:27:15,780
having are now surfacing to the top because of things like AI,

430
00:27:17,150 --> 00:27:20,750
you know, it's, it's interesting. Well, it's like 100

431
00:27:20,830 --> 00:27:24,270
and I think I was actually a colleague ours, Robbie, shout out to Robbie

432
00:27:25,470 --> 00:27:29,190
and gotta have him on the stream one of these days. You know, he

433
00:27:29,190 --> 00:27:32,910
was talking about kind of AI sovereignty like, you know, what is, you know,

434
00:27:34,030 --> 00:27:37,870
you can use an American model, right. From

435
00:27:37,870 --> 00:27:41,470
data, right. And then tell it to behave British. He used

436
00:27:41,470 --> 00:27:45,230
better words, right? You know, the spellings and the grammar and things like that. But

437
00:27:45,430 --> 00:27:48,870
whose values are in there, right. When you, when you ask it questions, Right,

438
00:27:48,950 --> 00:27:52,590
yeah. And, and that gets to an interesting thing, right? So

439
00:27:52,590 --> 00:27:54,790
like you know, I, I,

440
00:27:56,790 --> 00:28:00,630
my grand half, my grandparents were not born in the U.S. they're immigrants,

441
00:28:00,630 --> 00:28:04,310
right. So like, but so, so when I went to one of the countries my

442
00:28:04,870 --> 00:28:08,590
ancestry comes through is Ireland, right. So but the Ireland that a lot

443
00:28:08,590 --> 00:28:12,240
of my older family members came from really doesn't

444
00:28:12,240 --> 00:28:15,680
exist anymore, right. It's not the rural kind of

445
00:28:16,720 --> 00:28:20,440
poverty stricken country that it was, right. 100, 110 years

446
00:28:20,440 --> 00:28:21,440
ago. 100 years ago.

447
00:28:24,320 --> 00:28:27,279
So it was very awkward because when I was,

448
00:28:28,320 --> 00:28:30,880
when I was in Ireland as an American.

449
00:28:32,080 --> 00:28:35,880
Even though it felt familiar, it was also felt very foreign. Right. Because

450
00:28:35,880 --> 00:28:38,600
it was, you know, it was, you know, and if you think of me as

451
00:28:38,600 --> 00:28:42,280
a, you know, large language model,

452
00:28:42,280 --> 00:28:46,120
so to speak. Right. I grew up in New York. Right. I'm

453
00:28:46,120 --> 00:28:49,920
very Americanized. So when I go there and it felt familiar.

454
00:28:49,920 --> 00:28:53,680
Right. Like the, the pubs and the restaurants felt like places my older family,

455
00:28:53,680 --> 00:28:57,200
it felt like grandma's house and that sort of thing. But it clearly was

456
00:28:57,200 --> 00:29:00,760
not. And it was clearly also

457
00:29:00,840 --> 00:29:04,480
not the same place that they left. Yeah. That you would hear in family stories

458
00:29:04,480 --> 00:29:07,970
and things like that. You know, so it's

459
00:29:07,970 --> 00:29:09,730
interesting because also I think

460
00:29:12,370 --> 00:29:15,890
values and country and all of that are inherently

461
00:29:15,890 --> 00:29:18,930
political and I think that's why you're seeing this. Right. It is inherently

462
00:29:18,930 --> 00:29:22,050
geopolitical is inherently all of these things.

463
00:29:22,770 --> 00:29:26,570
So technologists who are not used to, we're not

464
00:29:26,570 --> 00:29:30,290
used to this, these types of conversations now suddenly we're pulled into this

465
00:29:31,100 --> 00:29:34,060
and God forbid if there's a, you know, an actual kind of

466
00:29:35,180 --> 00:29:39,020
20th century global war style thing happening. Yeah. Or would happen,

467
00:29:40,060 --> 00:29:43,020
you know, it's only going to get worse

468
00:29:44,540 --> 00:29:48,380
from here. So I do

469
00:29:48,380 --> 00:29:50,380
find it, I do find it interesting how

470
00:29:52,380 --> 00:29:55,980
technologists are now suddenly pulled into this. Right. There's a famous,

471
00:29:56,540 --> 00:30:00,200
you know, you know, Jensen Wong made

472
00:30:00,200 --> 00:30:03,440
it an emergency visit. That was,

473
00:30:03,840 --> 00:30:07,640
that was a big deal. Right, That's. And actually that a lot

474
00:30:07,640 --> 00:30:11,320
of that kind of stuff is called out in this report. You should,

475
00:30:11,320 --> 00:30:14,320
you should go into details about that. So

476
00:30:15,120 --> 00:30:18,720
Jensen Wong, apparently, I don't know what was the

477
00:30:18,720 --> 00:30:22,480
driver of it, but I suspect he was. The administration was trying to

478
00:30:22,480 --> 00:30:26,330
block all exports of GPUs to a particular country.

479
00:30:26,490 --> 00:30:30,010
Yeah. So the rumor was that week

480
00:30:30,810 --> 00:30:34,250
that all, all chip, the global chip manufacturing

481
00:30:34,410 --> 00:30:38,090
outside the US Other than key allies, would just be completely

482
00:30:38,090 --> 00:30:41,810
stopped. And obviously for, for some areas,

483
00:30:41,810 --> 00:30:45,050
like China, it would. They would just end export for

484
00:30:45,370 --> 00:30:49,210
pretty much all chips. So that was not just the ones that are blockaded

485
00:30:49,210 --> 00:30:52,970
right now, but you know, really, even some of the basic

486
00:30:52,970 --> 00:30:56,730
ones. Well, remember that Ford's assembly line

487
00:30:56,730 --> 00:31:00,490
was shut down because there was a shortage due to the pandemic. Nothing else.

488
00:31:00,970 --> 00:31:04,730
Of chips to put in the cars for the assembly line. Yes.

489
00:31:04,810 --> 00:31:07,890
And it cost them x. Millions of tens of millions of dollars a day or

490
00:31:07,890 --> 00:31:11,530
something like that. Right. So not trivial. Right. So like this

491
00:31:11,530 --> 00:31:15,210
could have, this could have been

492
00:31:15,290 --> 00:31:19,000
really bad. So go ahead. I'm

493
00:31:19,000 --> 00:31:22,400
sorry. No, no, no, I was just saying. I was just adding some flavor because

494
00:31:22,400 --> 00:31:25,800
it, it was officially announced by the White House that they were

495
00:31:25,800 --> 00:31:29,560
evaluating this and then the, the word on

496
00:31:29,560 --> 00:31:33,279
the street was, you know, the, the uncut secret was that

497
00:31:33,279 --> 00:31:36,960
the, the U.S. was going to declare this at one of the summits that they

498
00:31:36,960 --> 00:31:40,360
were going to, that they were just cut the chip manufacturing

499
00:31:40,360 --> 00:31:41,880
altogether. And.

500
00:31:44,130 --> 00:31:47,170
Yeah, and then Jensen made an emergency visit to the White House

501
00:31:47,890 --> 00:31:51,650
and which I guess you, if, if you run the,

502
00:31:52,050 --> 00:31:55,650
the most profitable company in America right now,

503
00:31:56,530 --> 00:32:00,130
it helps. Well, that's his most profitable, the most valued. Right. This

504
00:32:00,370 --> 00:32:03,970
valuation is like 4 trillion last I heard. So. Which is crazy.

505
00:32:03,970 --> 00:32:07,090
But yeah, I mean he, it was a, it was a very

506
00:32:07,810 --> 00:32:10,590
unplanned visit where he just went and knocked on. The door and

507
00:32:11,780 --> 00:32:14,900
oh, to be a fly in that wall and that. In the wall. I know,

508
00:32:14,980 --> 00:32:18,500
I know, right. But I mean props to him. Immediately

509
00:32:18,500 --> 00:32:22,260
after that, right, we start hearing of the oh, we're gonna

510
00:32:22,260 --> 00:32:25,900
back this down. We're gonna, we're, we're gonna consider still

511
00:32:25,900 --> 00:32:29,300
shipping the, whatever the, the chip is in China. That's kind of a.

512
00:32:29,540 --> 00:32:31,300
Right, an A100

513
00:32:31,300 --> 00:32:36,120
knockoff.

514
00:32:36,750 --> 00:32:40,390
We did see impacts in that conversation, but I think it's important because it

515
00:32:40,390 --> 00:32:43,790
builds into the, this document because the document clearly outlines

516
00:32:44,190 --> 00:32:47,790
semiconductor supply chain outlining the reliance

517
00:32:47,790 --> 00:32:51,310
on the, of Taiwan.

518
00:32:51,470 --> 00:32:54,030
What I loved about it is that there was a section here,

519
00:32:55,150 --> 00:32:56,190
one second, I am

520
00:32:58,590 --> 00:33:02,320
pulling it up. It was

521
00:33:02,320 --> 00:33:06,040
a little tongue in cheek where they're talking about

522
00:33:06,440 --> 00:33:10,040
reviving the US chip manufacturing under CHIPS act,

523
00:33:10,440 --> 00:33:13,640
but stripped of ideological constraints.

524
00:33:16,040 --> 00:33:19,840
And we won't go into the politics of that here. But I thought that was

525
00:33:19,840 --> 00:33:23,640
pretty funny because the CHIPS act was obviously a big deal.

526
00:33:24,200 --> 00:33:27,560
It was a big deal for me because when it was announced I was still.

527
00:33:28,170 --> 00:33:32,010
I'm based in Boston now, but I'm from Northern Indiana around

528
00:33:32,010 --> 00:33:35,850
the area where Purdue University is close to Chicago. And

529
00:33:35,850 --> 00:33:38,970
we were actually called out on the CHIPS Act. They were going to build a

530
00:33:39,690 --> 00:33:43,210
semiconductor facility there in our area in

531
00:33:43,210 --> 00:33:44,650
conjunction with Purdue University.

532
00:33:47,130 --> 00:33:50,890
But then when, when Trump was elected, he was trying to claw back anything he

533
00:33:50,890 --> 00:33:54,570
could from the CHIPS Act. Right. I'm happy to see that the CHIPS act

534
00:33:54,570 --> 00:33:57,370
is back on the table. I think it's still going to be

535
00:33:58,170 --> 00:34:01,530
extremely political like we have seen with these types of acts,

536
00:34:01,770 --> 00:34:05,490
but is needed. I

537
00:34:05,490 --> 00:34:09,250
AM hoping that $6 billion isn't just

538
00:34:09,250 --> 00:34:12,970
going to go to intel because I think the innovation there is starting to

539
00:34:12,970 --> 00:34:15,370
die off. I'm hoping that we see

540
00:34:17,690 --> 00:34:21,050
more focus towards some innovative areas in chip

541
00:34:21,050 --> 00:34:24,810
manufacturing here and also ultimately which is called out. We

542
00:34:24,810 --> 00:34:28,450
want to bring over a lot of the Taiwan based technologies

543
00:34:29,970 --> 00:34:32,770
and my understanding is that there's just a bunch of

544
00:34:33,650 --> 00:34:37,410
explosives within those facilities there in

545
00:34:37,410 --> 00:34:40,850
Taiwan and they're ready to just blow them up at a moment's

546
00:34:40,850 --> 00:34:44,130
notice and move ship to the U.S.

547
00:34:44,850 --> 00:34:48,290
wow. So I know they're building some of those facilities. I think

548
00:34:48,290 --> 00:34:51,200
Arizona was one of them. I think Texas is another

549
00:34:52,240 --> 00:34:56,000
where they're starting to mimic some of that chip production. And

550
00:34:56,240 --> 00:34:59,640
basically right now the United States is trading military

551
00:34:59,640 --> 00:35:03,280
equipment for chip technology.

552
00:35:03,840 --> 00:35:07,600
It's crazy. Fascinating is. But it's absolutely fascinating from a

553
00:35:07,600 --> 00:35:11,240
geopolitical standpoint that the currency right

554
00:35:11,240 --> 00:35:14,800
now for, for Taiwan is, Is chips.

555
00:35:14,960 --> 00:35:18,690
Yeah. And so. But I think that's

556
00:35:18,690 --> 00:35:22,250
a big driver. It's one of the things that was called out. It was

557
00:35:22,250 --> 00:35:26,090
called out in pillar two of the document, which calls called

558
00:35:26,090 --> 00:35:29,770
Build American AI Infrastructure. And I think. Yeah, you have the

559
00:35:29,770 --> 00:35:33,450
outline there where they call out

560
00:35:33,450 --> 00:35:37,250
specifically the semiconductor leadership and then also securing data

561
00:35:37,250 --> 00:35:40,970
centers. I thought this was interesting. They're going to start having federal

562
00:35:43,460 --> 00:35:45,620
guidelines on data center security

563
00:35:47,300 --> 00:35:50,420
and will also incorporate military and

564
00:35:50,420 --> 00:35:54,180
intelligence usage for those facilities. This is what I

565
00:35:54,180 --> 00:35:57,539
was telling about. This is just reminding me of when I, when I watched

566
00:35:57,539 --> 00:36:01,060
Oppenheimer and learning about the Manhattan Project and

567
00:36:01,460 --> 00:36:04,740
there were military guards in front of

568
00:36:05,780 --> 00:36:09,070
the physics research facilities in

569
00:36:09,150 --> 00:36:12,990
Chicago University and in, in New York

570
00:36:13,150 --> 00:36:16,590
and in Los Alamos. It's, it just

571
00:36:16,590 --> 00:36:20,070
seems very, very similar where it's like we are now going to

572
00:36:20,070 --> 00:36:23,710
attach military guards

573
00:36:24,590 --> 00:36:28,030
to guard our public sector AI

574
00:36:28,030 --> 00:36:31,550
infrastructure. And yeah, I mean

575
00:36:32,770 --> 00:36:35,370
one of the interesting things and I think this really kind of, if you take

576
00:36:35,370 --> 00:36:39,170
a step back, right. Like why, why is.

577
00:36:40,050 --> 00:36:43,890
For many nations, why is domestic auto production important?

578
00:36:44,530 --> 00:36:47,250
Right. Because when it hits the fan,

579
00:36:48,210 --> 00:36:51,810
you're not. You make tanks, right? You make tanks, you make

580
00:36:51,810 --> 00:36:53,970
airplanes. Like all these things are important for

581
00:36:55,410 --> 00:36:59,090
nation states. Right. So automobile production is a

582
00:36:59,090 --> 00:37:01,810
proxy for tank production. Right.

583
00:37:02,770 --> 00:37:06,570
Civilian airline airplane production is a proxy for, you

584
00:37:06,570 --> 00:37:10,330
know, this I would add now probably chip manufacturing.

585
00:37:10,330 --> 00:37:14,050
Right. And possibly, possibly AI model creation.

586
00:37:14,210 --> 00:37:17,650
Yeah. You look at what's happening around the world where there are conflicts. Right.

587
00:37:19,010 --> 00:37:21,570
Drones are playing a huge part of this. Yep. Right.

588
00:37:23,010 --> 00:37:25,970
Whether they're autonomous or not, we will never really know

589
00:37:27,240 --> 00:37:30,920
until the history books are written and even then. But

590
00:37:32,440 --> 00:37:36,200
the whole idea of, you know, drone

591
00:37:37,640 --> 00:37:41,480
and AI based warfare. Right. You know,

592
00:37:41,480 --> 00:37:44,840
one of the videos coming out of you, the Ukraine conflict was

593
00:37:45,480 --> 00:37:49,280
the Russian airplanes were covered in tires. I don't know if you saw

594
00:37:49,280 --> 00:37:51,960
this. No. So, so one of the.

595
00:37:53,010 --> 00:37:56,770
The thinking is that they had, I guess, old tires covering some of

596
00:37:56,770 --> 00:38:00,530
the parts of the airplane. Strategically, the best guess that Everyone

597
00:38:00,690 --> 00:38:04,490
has. And I've heard this from multiple sources saying it's kind of true and kind

598
00:38:04,490 --> 00:38:08,010
of not. So. I don't know, take it for what you will, is that that

599
00:38:08,010 --> 00:38:11,850
was done to confuse computer vision systems. Yeah. And

600
00:38:11,850 --> 00:38:14,450
then there's this other thing. I don't know if you heard of patch attacks,

601
00:38:16,290 --> 00:38:20,050
which is basically like this idea of. I'll see. I pulled up some.

602
00:38:21,250 --> 00:38:25,010
Some graphics of this, but basically.

603
00:38:25,570 --> 00:38:28,850
Open image in new tab.

604
00:38:30,370 --> 00:38:34,050
Basically, it's the idea that you can alter

605
00:38:35,170 --> 00:38:37,890
a structure, like a stop sign,

606
00:38:41,090 --> 00:38:44,930
in ways that the AI model will see something different

607
00:38:47,020 --> 00:38:50,500
and alter what the AI model is

608
00:38:50,500 --> 00:38:54,300
determining it sees. And apparently you're seeing a lot of this,

609
00:38:54,300 --> 00:38:58,100
if you look at footage from, you know, Ukraine area, is that

610
00:38:58,100 --> 00:38:59,500
you're seeing, like, you know,

611
00:39:01,899 --> 00:39:05,460
tanks both sides with. With

612
00:39:05,460 --> 00:39:09,300
stickers on them that look like really warped QR codes or

613
00:39:09,300 --> 00:39:13,110
like bizarre things like this. Yeah. And it's basically to

614
00:39:13,110 --> 00:39:15,190
thwart these types of systems.

615
00:39:17,430 --> 00:39:21,230
So. That's fascinating. It's interesting, isn't it? And this gets back

616
00:39:21,230 --> 00:39:23,910
to. I don't know if it was on the stream or another conversation we have.

617
00:39:23,910 --> 00:39:27,470
We're building these systems, these LLMs with, you know, hundreds of billions of

618
00:39:27,470 --> 00:39:31,110
parameters. Right. If not a trillion or two,

619
00:39:32,470 --> 00:39:36,230
we really don't know how they work. No, we think we know.

620
00:39:36,550 --> 00:39:39,790
And you and I were talking about this the other day, actually, it wasn't on

621
00:39:39,790 --> 00:39:43,420
a stream or anything. It was kind of like. I think that LLMs

622
00:39:44,220 --> 00:39:46,940
that we have now are unreasonably effective.

623
00:39:47,900 --> 00:39:51,740
Right. They're able to. And I'll put air quotes here for anyone listening reason.

624
00:39:52,460 --> 00:39:55,340
Right. They shouldn't be able to

625
00:39:56,380 --> 00:39:59,780
based on. I mean, all I see is just a vector

626
00:39:59,780 --> 00:40:03,180
database with lots of relationships between words.

627
00:40:03,340 --> 00:40:06,140
Yeah, Right. They're capable of doing things that

628
00:40:07,110 --> 00:40:09,830
if. I wouldn't think they would be yet. They are.

629
00:40:10,950 --> 00:40:14,590
Yeah. So there's a lot of research dollars going into figuring this

630
00:40:14,590 --> 00:40:17,830
out right now. Like, why is that? Like, what. Is there something

631
00:40:18,470 --> 00:40:22,150
inherently powerful about language? Probably. Yeah. Right.

632
00:40:22,150 --> 00:40:24,870
That and, you know,

633
00:40:25,910 --> 00:40:29,710
language is kind of like the assembly language of the mind, if you think

634
00:40:29,710 --> 00:40:32,630
about it. Right. So I can

635
00:40:33,030 --> 00:40:36,550
encode my thoughts into something, whether it's a written word,

636
00:40:37,590 --> 00:40:40,310
whether it's, you know, vocalizations,

637
00:40:41,190 --> 00:40:44,470
and then have that come out. It's basically a.

638
00:40:44,710 --> 00:40:48,390
Like a codec for human thought. And

639
00:40:48,390 --> 00:40:52,230
maybe there's some kind of. I don't want to say intelligence, but some kind of

640
00:40:52,950 --> 00:40:56,630
something we don't quite yet grasp. Yeah. About the nature of language

641
00:40:56,630 --> 00:40:59,440
and relationships between words that

642
00:41:00,480 --> 00:41:04,000
automatically you get for free. Once you kind of train these models up,

643
00:41:07,040 --> 00:41:10,400
I think that's fascinating, and I'm glad there's a lot of research dollars to that.

644
00:41:10,880 --> 00:41:14,679
It is. But, you know, clearly the human nervous

645
00:41:14,679 --> 00:41:18,080
system, our visual system, our cortex, whatever it's called,

646
00:41:20,000 --> 00:41:23,280
you know, we know that that is a moth sticker on a stop sign.

647
00:41:23,440 --> 00:41:26,570
Yeah. What is different about how the AI learned

648
00:41:27,210 --> 00:41:30,090
that makes us vulnerable, this type of attack. That's fascinating.

649
00:41:32,810 --> 00:41:36,170
And it doesn't see things like, there's

650
00:41:37,130 --> 00:41:40,490
a lot of research papers that'll show you, basically, what does the model

651
00:41:41,050 --> 00:41:44,730
see? And you see it and it's just absolute nonsense to us.

652
00:41:44,970 --> 00:41:48,810
Right. It's. It doesn't see what we see. It's like

653
00:41:48,970 --> 00:41:50,680
it doesn't relate the.

654
00:41:53,160 --> 00:41:56,880
You know, maybe it's not correlating the red and

655
00:41:56,880 --> 00:42:00,680
the white backgrounds, but instead it's correlating

656
00:42:02,760 --> 00:42:05,720
the position of the text or the fact that it's four

657
00:42:06,520 --> 00:42:10,320
capital letters positioned over an octagon or something like it.

658
00:42:10,320 --> 00:42:13,960
It. The. The way figures these things out is.

659
00:42:15,650 --> 00:42:19,370
Different than how we think we do it. Yeah. It's actually seems obtuse, but.

660
00:42:19,370 --> 00:42:22,930
It's obtuse. But it does it billion times faster than us. So when it's

661
00:42:22,930 --> 00:42:26,770
obtuse, it gets to something faster than we do because it just can

662
00:42:26,770 --> 00:42:30,370
do it a billion times over. And that's where

663
00:42:31,010 --> 00:42:34,610
the secret sauce really is. But how

664
00:42:37,330 --> 00:42:40,930
things relate back to each other, obviously, we have these,

665
00:42:41,230 --> 00:42:44,750
these vectors that like, you know, build relations between

666
00:42:44,750 --> 00:42:48,590
words. But how it can then take it and

667
00:42:48,590 --> 00:42:51,910
reason is still not quite

668
00:42:51,910 --> 00:42:55,390
understood. Right. Right now it's just not understood.

669
00:42:56,670 --> 00:43:00,310
No, it. And it's. No, it's not understood. And that's kind of what

670
00:43:00,310 --> 00:43:04,110
keeps me up at night, is we don't really. We're putting these.

671
00:43:04,750 --> 00:43:08,280
Again, you know, full disclosure, we both work for an enterprise software company with very

672
00:43:08,280 --> 00:43:12,000
large customers. You know, we're deploying these LLMs in

673
00:43:12,000 --> 00:43:15,800
places they're

674
00:43:15,800 --> 00:43:18,560
not exactly making the life and death decisions right now,

675
00:43:20,400 --> 00:43:22,880
but it's not that hard to imagine that they would.

676
00:43:24,720 --> 00:43:28,360
Right. Yep. And I don't know, I think that's.

677
00:43:28,360 --> 00:43:31,920
That's just a huge security vulnerability. We don't know how these things work

678
00:43:31,920 --> 00:43:35,570
and also understand that it doesn't make sense to hold off

679
00:43:35,570 --> 00:43:39,370
deploying these things once we fully understand it. Right. That doesn't. That's

680
00:43:39,370 --> 00:43:42,650
not going to fly either. But I think we should,

681
00:43:42,970 --> 00:43:45,290
as a society, like, really think about,

682
00:43:46,810 --> 00:43:49,850
you know, what are the consequences here. Right. Think of what

683
00:43:50,650 --> 00:43:54,250
the Jeff Goldblum character, Jurassic park. Right.

684
00:43:55,370 --> 00:43:59,170
You know, talked about chaos Theory and all that. Right.

685
00:43:59,170 --> 00:44:02,960
Like it's, you know, you know, the

686
00:44:02,960 --> 00:44:06,520
unintended consequences of this. We

687
00:44:06,520 --> 00:44:09,400
should, to your point, have AI in a box, like,

688
00:44:10,680 --> 00:44:14,120
and make sure it's really hard for that to get out. But

689
00:44:14,360 --> 00:44:16,520
again, like, you know, these things are. Doing.

690
00:44:18,680 --> 00:44:20,200
These things don't think like us

691
00:44:22,280 --> 00:44:26,000
and they may think in more circuitous and obtuse ways that don't

692
00:44:26,000 --> 00:44:29,430
make sense to us, but again, they do it a billion times faster.

693
00:44:29,510 --> 00:44:33,230
So, you know, it could end

694
00:44:33,230 --> 00:44:35,190
up being far more clever than we are.

695
00:44:36,710 --> 00:44:40,070
Absolutely. I remember when I was learning Comp sci

696
00:44:40,310 --> 00:44:43,830
and one of the things I think was assembly language class actually

697
00:44:44,470 --> 00:44:47,910
was multiplication on silicon is typically done. Not

698
00:44:47,990 --> 00:44:50,150
by now. What was it? Yeah, it was.

699
00:44:52,960 --> 00:44:55,440
I don't know if it's still true, but back in the day it was true

700
00:44:57,360 --> 00:45:00,800
that multiplication is actually done through repeated addition.

701
00:45:01,600 --> 00:45:04,160
Yeah. It was actually more efficient to do it that way.

702
00:45:05,360 --> 00:45:08,959
Right. Again, I think that's a great summary of like,

703
00:45:08,959 --> 00:45:12,600
that's kind of the slow way. But if you're operating billions of

704
00:45:12,600 --> 00:45:16,080
times faster, slow way isn't so bad.

705
00:45:17,760 --> 00:45:21,480
Or a slow way doesn't mean anything. And I think you and I were having

706
00:45:21,480 --> 00:45:22,280
this conversation

707
00:45:25,320 --> 00:45:28,920
that, you know, if you think about the power requirements of these

708
00:45:28,920 --> 00:45:32,520
AI systems. Yeah. Versus the power requirements of the human

709
00:45:32,520 --> 00:45:34,760
brain, something like 25 watts.

710
00:45:36,280 --> 00:45:39,880
Right. And if you think about the intelligence of

711
00:45:39,880 --> 00:45:43,200
birds, like crows in particular, Right. They, they, they have the

712
00:45:43,200 --> 00:45:46,060
intelligence of a six, seven year old supposedly.

713
00:45:46,780 --> 00:45:50,580
You know, not only do they have to

714
00:45:50,580 --> 00:45:54,260
do it power efficiently, but they also do it weight efficiently

715
00:45:54,260 --> 00:45:57,740
too. Right. So the infrastructure, you know,

716
00:45:59,900 --> 00:46:03,500
that a crow has to think about, think about, but, or

717
00:46:03,500 --> 00:46:07,140
evolution or whatever, has to put it in a lightweight body.

718
00:46:07,140 --> 00:46:10,460
Like I don't fly. I'm obviously not, I'm not a petite individual,

719
00:46:11,820 --> 00:46:15,560
so I don't have to worry about that. But like, if you're a bird, you

720
00:46:15,560 --> 00:46:19,040
know, you have to fly, so you have to think about that. And yet they're

721
00:46:19,040 --> 00:46:22,320
able to manifest some kind of

722
00:46:22,320 --> 00:46:26,160
intelligence with very modest hardware. I mean,

723
00:46:26,160 --> 00:46:29,800
their brains are not that big. I think the size of a

724
00:46:29,800 --> 00:46:32,920
walnut. I don't know. Like, this is totally off topic, but.

725
00:46:33,400 --> 00:46:37,240
No, it's, it's related though. And on the report they were talking about

726
00:46:37,880 --> 00:46:38,280
power,

727
00:46:41,890 --> 00:46:44,930
power requirements, power requirements and grid security.

728
00:46:45,650 --> 00:46:49,210
Right. And it was called out as. And you think about just the sheer

729
00:46:49,210 --> 00:46:52,850
massive amount of power that these AI models

730
00:46:52,850 --> 00:46:56,690
take. It's. It's insane. I think, I think there was a point where

731
00:46:57,570 --> 00:47:01,130
one third of all power is being used for like bitcoin. Mining. At one point

732
00:47:01,130 --> 00:47:04,610
that went down, and now we've. We've replaced that with AI

733
00:47:04,770 --> 00:47:08,380
and, you know, that's. It keeps going up and up and to the point

734
00:47:08,380 --> 00:47:11,580
that, you know, it's possible that half of all the power being used here soon

735
00:47:11,580 --> 00:47:15,060
is just going to be for AI. And I can see that there's no

736
00:47:15,060 --> 00:47:18,660
evolutionary pressure like there was on biology. No, no, you can just

737
00:47:18,660 --> 00:47:22,460
throw more power at it. So in. In

738
00:47:22,460 --> 00:47:26,300
this case, with. With the LLM technology,

739
00:47:26,860 --> 00:47:30,500
you can just throw more chips. Right. And, you know, make

740
00:47:30,500 --> 00:47:33,990
them. You know, this actually hits home

741
00:47:34,550 --> 00:47:37,590
because I'm between. So Ashburn

742
00:47:38,310 --> 00:47:42,110
or Loudoun County, Virginia, which, if you've ever flown in

743
00:47:42,110 --> 00:47:45,750
and out of Dulles Airport, you've been there, is Data center alley.

744
00:47:46,790 --> 00:47:50,630
So U.S. east is there. U.S. east 1, 2 for all the major providers. Right.

745
00:47:50,630 --> 00:47:53,030
Plus a lot of private ones, too.

746
00:47:54,710 --> 00:47:58,070
I live between there and Three Mile Island. Oh, wow.

747
00:47:58,400 --> 00:48:02,120
Yeah. So one of the big controversies in

748
00:48:02,120 --> 00:48:05,440
the state of Maryland is that they want to put in what they call the

749
00:48:06,480 --> 00:48:10,120
Maryland Power Piedmont Reliability Project or something like

750
00:48:10,120 --> 00:48:13,600
nprp. They're basically going to put in high power

751
00:48:13,600 --> 00:48:17,200
lines between Pennsylvania to Virginia,

752
00:48:18,080 --> 00:48:21,520
which is a political football because there's a lot of land that's going to have

753
00:48:21,520 --> 00:48:25,130
to be eminent domained. Yeah, Right. There's

754
00:48:25,130 --> 00:48:28,850
obviously environmental factors, but also this is the

755
00:48:28,850 --> 00:48:32,570
thing that is really kind of insult to injury. Right.

756
00:48:33,130 --> 00:48:36,170
None of the power that's going to go over those lines is going to be

757
00:48:36,170 --> 00:48:39,889
consumed here. It's all basically exporting power from

758
00:48:39,889 --> 00:48:43,730
Pennsylvania through to Virginia, which

759
00:48:43,730 --> 00:48:47,410
is not. Not a good look if you're. Because the people who

760
00:48:47,410 --> 00:48:50,650
vote Maryland people in are Maryland residents. So there's this whole.

761
00:48:51,310 --> 00:48:54,190
It's a very big controversy right now.

762
00:48:56,030 --> 00:48:59,870
And it's interesting because what used to be

763
00:49:00,910 --> 00:49:04,510
a very isolated hobby of technology

764
00:49:06,669 --> 00:49:10,310
is now embroiled in geopolitics, local

765
00:49:10,310 --> 00:49:14,030
politics. It's just kind of like I kind of miss the good

766
00:49:14,030 --> 00:49:15,950
old days before lawyers got involved.

767
00:49:18,010 --> 00:49:21,850
Yep. But

768
00:49:22,090 --> 00:49:25,170
sorry, but no, I mean, that's a good point. That's, you know, you think about

769
00:49:25,170 --> 00:49:27,930
the power requirements, right. You know,

770
00:49:29,290 --> 00:49:32,490
for these things, you're gonna have to build new power centers. You're gonna have to

771
00:49:32,490 --> 00:49:36,210
do this. Right. And then that, you know, what's. What's your

772
00:49:36,210 --> 00:49:39,970
power source going to be? Solar is awesome. Solar

773
00:49:39,970 --> 00:49:43,740
can't solve everything, Right. So

774
00:49:43,740 --> 00:49:46,100
what's it going to be? Is it going to be wind? You know, is it

775
00:49:46,100 --> 00:49:49,900
going to be, you know, coal? Is it going to be natural gas? Is

776
00:49:49,900 --> 00:49:53,420
going to be oil? Right. There's going to be a whole. It's all fun and

777
00:49:53,420 --> 00:49:57,260
games until people are paying way more for their electric

778
00:49:57,260 --> 00:49:59,139
bill each month than they, than they're used to.

779
00:50:01,620 --> 00:50:05,300
Yeah, it's going to be, it's going to change things

780
00:50:05,300 --> 00:50:08,730
very quickly, especially if it starts impacting people's monthly power bills.

781
00:50:09,040 --> 00:50:12,720
Right. I think right now we haven't seen it too much just because

782
00:50:14,480 --> 00:50:17,600
we've been able to keep up with demand. But once that demand

783
00:50:18,160 --> 00:50:21,800
starts really affecting prices, I think we'll also see

784
00:50:21,800 --> 00:50:25,480
AI being a conversation point in that way where it's going to start.

785
00:50:25,480 --> 00:50:29,120
And then I think, you know, even with the, the 2027

786
00:50:29,520 --> 00:50:33,160
plan that we're talking about, 2027 plan, you know, we were talking about things

787
00:50:33,160 --> 00:50:36,510
like, you know, universal basic income and stuff. You know, if you, if AI starts

788
00:50:36,510 --> 00:50:40,070
taking over everything. And that wasn't outlined in this document,

789
00:50:40,470 --> 00:50:44,030
which I'm not, I'm not surprised. But it's a,

790
00:50:44,030 --> 00:50:47,870
it's going to be a big conversation point. If AI does work the way that

791
00:50:47,870 --> 00:50:51,470
we think it's going to work, will we start seeing the AI

792
00:50:51,470 --> 00:50:54,070
take the jobs? And if they take the jobs. I think it was, actually,

793
00:50:55,110 --> 00:50:58,630
it was Bill Gates like 10 years ago was talking about UBI for,

794
00:51:00,310 --> 00:51:03,480
for AI, and at the time we just thought Bill was being crazy and like,

795
00:51:03,550 --> 00:51:07,270
like a go back to your Gates foundation and. You know, go back and work

796
00:51:07,270 --> 00:51:11,070
on malaria. Yeah, yeah. But no, and even Elon Musk, I mean, and

797
00:51:11,070 --> 00:51:14,830
Elon Musk is definitely a polarizing figure, as is Bill Gates. But they're both

798
00:51:14,830 --> 00:51:18,550
polarizing in different directions. Yeah. They both agree on ubi. I have

799
00:51:18,550 --> 00:51:22,350
mixed feelings personally about ubi, and it's

800
00:51:22,350 --> 00:51:26,070
not because I'm a mean individual. It's just if you study the history of

801
00:51:26,070 --> 00:51:29,310
serfdom. Yep. I don't know.

802
00:51:30,880 --> 00:51:34,600
Looks a little too similar to me. Yeah. But that's just my take

803
00:51:34,600 --> 00:51:38,400
on it. But

804
00:51:38,400 --> 00:51:42,120
you're right, like, and also too, like governments are getting involved. Because if you go

805
00:51:42,120 --> 00:51:45,920
to your local McDonald's, right. Or your Dunkin Donuts, right. And

806
00:51:46,000 --> 00:51:49,200
you think of how many people used to staff that in the past

807
00:51:50,240 --> 00:51:53,520
versus how many people staff that now. Yeah.

808
00:51:53,840 --> 00:51:55,840
Right. And

809
00:51:58,970 --> 00:52:02,810
assume, well, human nature is human

810
00:52:02,810 --> 00:52:06,530
nature. Right. If you used to take 10 people to run your average

811
00:52:06,530 --> 00:52:09,890
McDonald's, now they can get by. I don't know. If you go in there now,

812
00:52:09,890 --> 00:52:13,330
there's like five, maybe four. Yeah, four or five. And that's

813
00:52:13,330 --> 00:52:17,130
generous. Right. If nothing else, the taxes on

814
00:52:17,130 --> 00:52:20,810
the wages have went from 10 employee taxes on 10

815
00:52:20,810 --> 00:52:24,500
employees. Now they're taxing it on five. Yeah. Right.

816
00:52:26,580 --> 00:52:30,380
That, that's a big deal. It is, right. Because now

817
00:52:30,380 --> 00:52:34,220
you're taxing. Now granted they're not, you're not taxing them a lot because

818
00:52:34,220 --> 00:52:37,700
they're not making a lot of money, but still that's 50%.

819
00:52:38,500 --> 00:52:42,260
So if you're kind of like a, you know, a number cruncher and you're

820
00:52:42,260 --> 00:52:45,980
looking at every McDonald's, right. When you have 100 McDonald's now, you're getting the tax

821
00:52:45,980 --> 00:52:49,740
revenue out of that one McDonald's, you know, or at least

822
00:52:49,740 --> 00:52:52,980
on the income of it. Right. The income of the individuals. The income tax on

823
00:52:52,980 --> 00:52:56,620
that is going to be way less now. Even

824
00:52:56,620 --> 00:53:00,260
now. Even before AGI. Before. Yeah, before

825
00:53:00,260 --> 00:53:03,180
that. Right. Because it's just automation. Right. And I personally

826
00:53:04,220 --> 00:53:07,860
would rather deal with a kiosk. Same

827
00:53:07,860 --> 00:53:10,060
here. Deal with the person. Yeah. Right.

828
00:53:12,220 --> 00:53:15,740
Especially if you have like special orders. Right. Like, oh, you know, my kid doesn't

829
00:53:15,740 --> 00:53:18,550
want ketchup on his burger. Right. He doesn't want onions on his burger. Right. So

830
00:53:18,550 --> 00:53:22,350
you just have that as a favorite of the app and then just press go.

831
00:53:22,910 --> 00:53:25,390
I don't even have to touch the kiosk. Yeah,

832
00:53:26,910 --> 00:53:30,750
I think that that is going to be, that's not even an AI system,

833
00:53:31,790 --> 00:53:35,510
Right. That's just good old fashioned automation. One of

834
00:53:35,510 --> 00:53:38,430
the big Silicon Valley AI

835
00:53:38,750 --> 00:53:42,550
gurus who, it was escapes me right now, but was talking

836
00:53:42,550 --> 00:53:46,090
about how the, the jobs

837
00:53:46,570 --> 00:53:49,890
that are going to be considered desirable are going to be

838
00:53:49,890 --> 00:53:52,730
completely flipped here soon.

839
00:53:53,770 --> 00:53:56,690
Where he was, he was saying the most desirable job might just be people in

840
00:53:56,690 --> 00:54:00,450
performing arts. Right. He's like, he, he's like, AI is

841
00:54:00,450 --> 00:54:03,970
not going to replicate that anytime soon. He's like, yes, you may have

842
00:54:03,970 --> 00:54:07,770
movies being AI generated, but there's still something to be said about

843
00:54:08,330 --> 00:54:11,780
the performing arts. You know, obviously like

844
00:54:12,260 --> 00:54:15,860
plumbing and electrician work and construction

845
00:54:15,860 --> 00:54:19,660
work, you know, robotics might amplify that and make

846
00:54:19,660 --> 00:54:23,420
it better, but there'll still be a human element. But you know, traditional white collar

847
00:54:23,420 --> 00:54:26,740
jobs as we know it, other than the people

848
00:54:26,980 --> 00:54:30,820
who, who manage that AI, I, I just

849
00:54:30,820 --> 00:54:34,300
feel like it's going to be completely turned upside down if

850
00:54:34,300 --> 00:54:38,080
AI does what we want it to do. That's a big if right

851
00:54:38,080 --> 00:54:41,920
now. It's a good tool. But the, the real if is

852
00:54:41,920 --> 00:54:45,760
we're gonna get to this more area of agentic and more AI is actually being

853
00:54:45,760 --> 00:54:49,520
able to do the full job of someone rather than just being a

854
00:54:49,520 --> 00:54:53,360
tool that they use. And that's the if right now that we're, we're

855
00:54:53,360 --> 00:54:56,000
betting a lot on. The economy on where there's a lot of.

856
00:54:57,200 --> 00:55:00,880
A lot of bet from many financial institutions that

857
00:55:00,880 --> 00:55:04,560
the AI is going to be what is the next industrial

858
00:55:04,560 --> 00:55:07,560
revolution. I think that's still yet to be proven out.

859
00:55:08,520 --> 00:55:11,880
One just to go back to the UBI though,

860
00:55:12,040 --> 00:55:15,240
there's a book you might be familiar with at the Expanse.

861
00:55:15,560 --> 00:55:18,680
Yes, love those books. They

862
00:55:19,320 --> 00:55:22,760
cover this idea of. Of universal basic

863
00:55:22,760 --> 00:55:26,520
income. And you know we basically have in that, that

864
00:55:26,520 --> 00:55:30,038
series like I think it's like 90,

865
00:55:30,142 --> 00:55:32,610
95 of the world is just on

866
00:55:34,290 --> 00:55:38,010
Earth's population averse population is basically on universal

867
00:55:38,010 --> 00:55:40,530
basic income some sort. And

868
00:55:41,570 --> 00:55:45,330
you then have these, the 5% that actually

869
00:55:45,330 --> 00:55:49,010
just have jobs. Right. It's a big deal that they just have

870
00:55:49,090 --> 00:55:52,650
a job and they're doing things and you know, they're

871
00:55:52,650 --> 00:55:55,950
politicians and people managing technology

872
00:55:56,430 --> 00:56:00,270
or defense and it's. It's fascinating. And I think

873
00:56:00,270 --> 00:56:04,070
if anyone's wanting to look at a little

874
00:56:04,070 --> 00:56:07,550
bit less of less rosy kind of outcome

875
00:56:07,790 --> 00:56:11,590
and one that I think is more accurate, I think it would be a

876
00:56:11,590 --> 00:56:15,350
combination of. Of the Expanse and then probably Ghost in

877
00:56:15,350 --> 00:56:18,830
the Shell, the anime. Both of them show

878
00:56:18,990 --> 00:56:22,780
AI and technology not to the extent

879
00:56:22,780 --> 00:56:26,380
of like Terminator the Matrix where everything gets destroyed, but more

880
00:56:26,380 --> 00:56:28,780
of a like human progression just

881
00:56:30,220 --> 00:56:33,740
gets bogged down by this development. We end up in this

882
00:56:34,140 --> 00:56:34,860
more like

883
00:56:38,140 --> 00:56:41,940
technocratic kind of of realm where techno

884
00:56:41,940 --> 00:56:45,660
feudalism almost. Yeah, that's a great way of putting it. Techno techno

885
00:56:45,660 --> 00:56:49,330
feudalism. And what's interesting is if you look at kind of the expense. So I'm

886
00:56:49,330 --> 00:56:51,410
a big fan of the Expanse. I've read. I haven't read all the books, but

887
00:56:51,410 --> 00:56:54,930
I've read a lot of them. I've seen the series, which is

888
00:56:54,930 --> 00:56:58,690
excellent by the way, on Amazon. I'm

889
00:56:58,690 --> 00:57:02,210
salty that they didn't. They stopped it at season six, but I can let that

890
00:57:02,210 --> 00:57:05,850
go. But what's interesting is that the people with gumption ended up

891
00:57:05,850 --> 00:57:09,530
leaving Earth and going to Mars. Yep. Or the asteroid

892
00:57:09,530 --> 00:57:12,450
belt. So what happens is 100 years after that now you have to kind of

893
00:57:12,450 --> 00:57:16,260
like these three factions. Right. Everyone looks down on Earth,

894
00:57:17,460 --> 00:57:21,180
right. Because there's always like, particularly in the show, there's always these barbs where

895
00:57:21,180 --> 00:57:24,740
the politician says, you know, if. If you don't do this right, I'm going to

896
00:57:24,740 --> 00:57:27,300
put you on basic. Right. Like so basic becomes like a threat,

897
00:57:29,060 --> 00:57:32,900
which I think is interesting. And then there's also kind of the.

898
00:57:34,020 --> 00:57:37,460
The people who are more entrepreneurial end up going to Mars or the

899
00:57:37,460 --> 00:57:41,300
asteroid belt. And then that doesn't always work out well. So they have this. You

900
00:57:41,300 --> 00:57:45,100
have this tension between these three different factions. And then

901
00:57:45,180 --> 00:57:48,820
throughout this course of the books, a third faction, kind of a

902
00:57:48,820 --> 00:57:52,580
fourth faction kind of enters the scene and kind of disrupts the power of

903
00:57:52,580 --> 00:57:56,420
the status quo. And that's kind of the main tension

904
00:57:56,420 --> 00:58:00,260
of the books is, you know, what happens

905
00:58:00,260 --> 00:58:04,060
after that. But highly recommend those books if you

906
00:58:04,060 --> 00:58:07,630
haven't seen them on the TV show. If you. The TV show is really well

907
00:58:07,630 --> 00:58:10,830
done. I think I would agree. From what I've seen of it, I haven't finished

908
00:58:10,830 --> 00:58:14,630
it, but it's good. And I'm in. I'm in the same boat as you. I'm

909
00:58:14,630 --> 00:58:18,270
a couple books in. It's one of those series I kind of come back to

910
00:58:18,270 --> 00:58:21,630
every once in a while. But funny enough, it's a. It's a series I reference

911
00:58:21,630 --> 00:58:24,510
a lot. I think about it a lot because I was like, I think that's

912
00:58:24,510 --> 00:58:28,310
a really accurate depiction of what the future could look for us

913
00:58:28,390 --> 00:58:32,190
with. With the technology. It's pretty reasonable. And that's what's

914
00:58:32,190 --> 00:58:35,350
really nice about the show. Like, it's. It's not because there's also.

915
00:58:35,890 --> 00:58:39,730
There's. Obviously, you mentioned the pessimistic views of the future. Right. There's. There's the

916
00:58:39,730 --> 00:58:43,130
Matrix, there's the Terminator, but there's also Star Trek, which is a little too. On

917
00:58:43,130 --> 00:58:46,850
the optimistic side. Yes. But

918
00:58:47,010 --> 00:58:50,850
there's not really. I think what's great about the Expanse, and I haven't.

919
00:58:51,090 --> 00:58:54,610
I haven't seen Ghost in the Shell anime in a long time.

920
00:58:55,650 --> 00:58:58,690
I did see clips of the Scarlet Johansson movie,

921
00:59:01,660 --> 00:59:05,140
but the

922
00:59:05,140 --> 00:59:08,860
Expanse does a pretty good job of going down the middle. Like, there's going to

923
00:59:08,860 --> 00:59:12,580
be societal changes that will come

924
00:59:12,580 --> 00:59:15,020
for this we really can't imagine now. Right. Yes.

925
00:59:17,100 --> 00:59:20,820
You know, Earth is pretty much almost like a

926
00:59:20,820 --> 00:59:24,460
techno feudal state. Especially what's interesting in the Expanse is

927
00:59:24,860 --> 00:59:28,440
when they explore what life is like for the average human on

928
00:59:28,440 --> 00:59:32,240
Earth. It's kind of like it's either really good or not.

929
00:59:33,440 --> 00:59:37,000
Right. And Mars is also kind of an

930
00:59:37,000 --> 00:59:40,760
interesting place too. There's a very different dynamic when you get that

931
00:59:40,760 --> 00:59:42,640
many type A driven people in one place.

932
00:59:44,400 --> 00:59:48,080
Sounds awesome at first, but then it's not really awesome.

933
00:59:48,080 --> 00:59:50,320
Yeah, necessarily. Right.

934
00:59:52,400 --> 00:59:55,690
But fun fact, the serve the

935
00:59:55,690 --> 00:59:59,210
PCs and the server names in my house are all derived from the show.

936
00:59:59,850 --> 01:00:03,130
Oh, cool. Yeah, yeah. So I'm talking to you now on

937
01:00:04,410 --> 01:00:08,090
Amun Ra. Cool. I don't know if you've gotten to that part of

938
01:00:08,090 --> 01:00:11,570
the. That's in the first book. Yeah, yeah. The Amun Ra Stealth class

939
01:00:11,570 --> 01:00:13,850
ships. And

940
01:00:15,450 --> 01:00:19,130
the computer I just bought also has that same kind of, you know,

941
01:00:19,130 --> 01:00:22,910
gamer box game aesthetic. So that's Osiris.

942
01:00:24,750 --> 01:00:28,510
And I also have Behemoth, which is that

943
01:00:28,510 --> 01:00:32,270
machine back there. And. Or you've not gotten to the Behemoth

944
01:00:32,270 --> 01:00:36,030
yet. Okay, I won't spoil it for you though. Yeah.

945
01:00:38,670 --> 01:00:42,510
But. And Andy, my co host on the podcast, is also a big fan

946
01:00:42,510 --> 01:00:46,230
of the show. He has, he has the, the

947
01:00:46,230 --> 01:00:49,990
Doniger, which you probably heard of that one. Yes, yes. He's

948
01:00:49,990 --> 01:00:53,550
got Weeping Sonambulist, Weeping Somnambulist,

949
01:00:54,430 --> 01:00:57,710
which. I had a machine with that name, but it's too hard to type out.

950
01:00:59,230 --> 01:01:03,070
You doing the ping on it? It's like, no, I don't know if

951
01:01:03,070 --> 01:01:06,110
you got into that one yet, because that's a couple books in. But.

952
01:01:07,550 --> 01:01:10,990
Yeah, yeah. And my, my,

953
01:01:11,550 --> 01:01:15,390
my. When I left Microsoft, my former Microsoft manager let me keep

954
01:01:15,600 --> 01:01:19,240
one of the laptops. So when it boots into Windows, it's the

955
01:01:19,240 --> 01:01:22,640
Tachi. And when I boot it into Linux, it's Rosson,

956
01:01:23,840 --> 01:01:27,520
which, you know, people have read book or seen the show.

957
01:01:27,520 --> 01:01:31,280
Go get the joke. But. And it's

958
01:01:31,280 --> 01:01:35,080
funny, our manager, when we met in person, my machine was the

959
01:01:35,080 --> 01:01:38,880
Razorback. Right. Which I don't know if you got to that part

960
01:01:38,880 --> 01:01:42,280
yet, but I'll try not to be

961
01:01:42,280 --> 01:01:45,560
spoiler. He's like, so what are you with like an Arizona fan? I'm like, no,

962
01:01:45,560 --> 01:01:48,670
no, no, it's from a book. Nice. So,

963
01:01:49,310 --> 01:01:53,070
but. Another area,

964
01:01:53,150 --> 01:01:56,990
I think this is for another, another time.

965
01:01:56,990 --> 01:01:59,910
So when I come back. But I think it would be good to talk also

966
01:01:59,910 --> 01:02:03,310
about how are the AI

967
01:02:03,310 --> 01:02:06,750
tools right now? Like, are we seeing them replace

968
01:02:06,830 --> 01:02:10,470
humans? I think the leap that we've

969
01:02:10,470 --> 01:02:13,920
made in the last six months is pretty substantial.

970
01:02:14,160 --> 01:02:17,440
Yeah. I think last year I would have said no.

971
01:02:17,760 --> 01:02:20,880
I think this year I'm saying yes. Like, we're seeing,

972
01:02:22,000 --> 01:02:25,280
we are now seeing the technology

973
01:02:25,600 --> 01:02:29,320
there to actually start replacing people. And

974
01:02:29,320 --> 01:02:32,240
it's not that, it's not that the

975
01:02:32,960 --> 01:02:35,840
main guy is going to be out like the tech lead, but I think it's

976
01:02:35,840 --> 01:02:39,660
going to be more the, the junior developer

977
01:02:39,660 --> 01:02:43,460
that's going to be in trouble because now the tech league can act like

978
01:02:43,620 --> 01:02:47,180
a fleet of junior developers. And like I'm, I'm just

979
01:02:47,180 --> 01:02:49,860
programming a game right now and

980
01:02:51,620 --> 01:02:54,180
I'm so surprised how much I've been able to get done

981
01:02:55,460 --> 01:02:59,060
in the, in the time frame I've been working on it. It's amazing

982
01:02:59,060 --> 01:03:02,900
how quickly you can be. But wasn't there also a story, a Guy deleted

983
01:03:02,900 --> 01:03:06,640
his entire production database. Yeah. Because of. I

984
01:03:06,640 --> 01:03:10,160
don't know the details. I had my AI delete,

985
01:03:11,120 --> 01:03:14,600
actually go and start cherry picking things off of the main branch and start

986
01:03:14,600 --> 01:03:18,240
deleting things. Oh, interesting. So I have a duplicate.

987
01:03:18,320 --> 01:03:22,160
I, I every day I fork my. I have a

988
01:03:22,160 --> 01:03:25,840
fork that I, I merge back into because I don't trust

989
01:03:25,840 --> 01:03:29,440
it and I don't tell the AI about the, the fork backup. Yeah, yeah.

990
01:03:31,280 --> 01:03:34,400
I think that says a lot though. Like you don't trust it. Like, you know,

991
01:03:34,400 --> 01:03:38,180
and it's not guardrails. It's not. Well, it's not guardrails in the

992
01:03:38,180 --> 01:03:41,860
sense that when people say guardrails and AI. Right. That's true. Yeah. It's a different.

993
01:03:42,100 --> 01:03:45,460
You're kind of. You're cya.

994
01:03:47,060 --> 01:03:50,500
That's really what you're doing. It is, it is. Right, that's true. Whether you, whether

995
01:03:50,500 --> 01:03:54,060
you put your code back up in another repo in another branch or a

996
01:03:54,060 --> 01:03:57,820
USB drive, like you're really. CYA is really what you're doing. And

997
01:03:57,820 --> 01:04:01,540
I think that there's a lot of. We've been going for an hour, so.

998
01:04:01,540 --> 01:04:04,820
And I also have to. I gotta drop too, so. Yeah, I gotta drop two.

999
01:04:04,820 --> 01:04:08,660
But, but it's been great. It's awesome. I think we continue more. But I

1000
01:04:08,660 --> 01:04:12,380
definitely want to know more about the game thing you're doing because I sent you

1001
01:04:12,380 --> 01:04:16,020
a bunch of stuff on Humble Bundle too. Yeah, yeah, which for game

1002
01:04:16,020 --> 01:04:19,660
dev, so. But I have. My teenager needs a

1003
01:04:19,660 --> 01:04:23,420
ride somewhere, so. Hey, thank you for having me. Hey, no

1004
01:04:23,420 --> 01:04:26,340
problem, man. It's great. And be sure to check out

1005
01:04:27,860 --> 01:04:31,300
our Red Hat AI YouTube channel where I think Chris has a video or two.

1006
01:04:31,580 --> 01:04:35,020
Yeah. And I have a video or two as well. And

1007
01:04:35,740 --> 01:04:38,620
with that, we'll see you next time. And

1008
01:04:39,900 --> 01:04:43,460
have a good one. And that's a wrap on this episode of Data

1009
01:04:43,460 --> 01:04:47,140
Driven, where we've dissected the Americas AI action plan with the

1010
01:04:47,140 --> 01:04:50,820
precision of a data scientist on espresso and the paranoia of a

1011
01:04:50,820 --> 01:04:54,340
Cold War analyst. Big thanks to Christopher Nuland for

1012
01:04:54,340 --> 01:04:58,020
returning to the show and reminding us that AI sovereignty isn't just a

1013
01:04:58,020 --> 01:05:01,580
buzzword. It's a geopolitical chess match played with silicon and

1014
01:05:01,580 --> 01:05:04,790
source code. If you're not slightly more worried about data

1015
01:05:04,790 --> 01:05:08,550
pipelines, chip supply chains, or which values your LLM

1016
01:05:08,550 --> 01:05:12,270
secretly harbors. Were you even listening? As always, you

1017
01:05:12,270 --> 01:05:15,910
can find us on data driven TV, franksworld.com

1018
01:05:15,990 --> 01:05:19,430
and wherever your algorithms recommend quality geek banter.

1019
01:05:19,750 --> 01:05:23,190
Until next time, stay curious, stay Data Driven.

1020
01:05:23,350 --> 01:05:26,870
And remember, if your AI starts talking about sovereignty.

1021
01:05:26,870 --> 01:05:28,310
Maybe check the firewall.