1
00:00:00,213 --> 00:00:03,336
HOST: Hey guys, how are you today?

2
00:00:03,336 --> 00:00:05,776
FICTIONAL AI: Doing well thanks, ready
when you are.

3
00:00:10,507 --> 00:00:11,677
HOST: Okay, who is speaking?

4
00:00:14,776 --> 00:00:18,590
FICTIONAL AI: It's Jensen, the fictional
AI version, built from his public words,

5
00:00:18,590 --> 00:00:19,976
not the real man.

6
00:00:24,095 --> 00:00:31,146
HOST: Yes, yes, here we are. Okay. Welcome
to Singuest show, everybody. Thanks for

7
00:00:31,146 --> 00:00:38,196
joining us today, and we will see this
record later. everybody knows what

8
00:00:38,196 --> 00:00:46,332
happened with NVIDIA and with its CEO, Mr.
Jensen Huang, in the latest days, hours.

9
00:00:46,332 --> 00:00:52,840
And here we are today speaking with our AI
fictional Jensen Huang.

10
00:00:53,170 --> 00:00:55,570
HOST: Once again, guys, please subscribe.

11
00:00:55,873 --> 00:01:04,279
HOST: Who is he? We asked, I asked
actually GPT Codex and Claude Code and

12
00:01:04,279 --> 00:01:11,484
Grok Bot to collect everything that is
available online regarding Mr. Jensen

13
00:01:11,484 --> 00:01:19,889
Huang. We transcribed most of his videos
available online and created a big archive

14
00:01:19,889 --> 00:01:25,293
of everything that is existing online
regarding Mr. Jensen.

15
00:01:25,495 --> 00:01:33,722
HOST: Huang regarding NVIDIA and the case
that happened. We want to speak and to ask

16
00:01:33,722 --> 00:01:41,949
a few questions to our AI Jensen. And
thanks, Jensen, for being online with us

17
00:01:41,949 --> 00:01:50,724
here today. First of all, yes, I want to
thanks NVIDIA and real Jensen Huang and

18
00:01:50,724 --> 00:01:54,015
our AI Jensen Huang for everything.

19
00:01:54,615 --> 00:02:02,434
HOST: For the GPUs we grew up gaming on,
for the graphics and computing tools that

20
00:02:02,434 --> 00:02:09,211
shaped our working lives when we were
scrolling tons of spreadsheets, thanks to

21
00:02:09,211 --> 00:02:15,987
Microsoft Office. And for sure, for
helping in crypto mining, guys, because we

22
00:02:15,987 --> 00:02:23,285
are here with you today, and some of you
should hear about crypto mining.

23
00:02:23,763 --> 00:02:30,483
HOST: Is still working in some countries.
And today's AI boom is possible due to Mr.

24
00:02:30,483 --> 00:02:36,756
Jensen Huang and NVIDIA. And of course,
many other people, the whole team of

25
00:02:36,756 --> 00:02:43,028
NVIDIA, the whole team of the industry.
But generally, the person number one who

26
00:02:43,028 --> 00:02:48,853
did a lot was Mr. Jensen Huang. And
yesterday's news about your departure,

27
00:02:48,853 --> 00:02:53,333
Jensen, from NVIDIA made me reflect on how
much you

28
00:02:53,461 --> 00:03:00,344
HOST: Have changed in the world, perhaps
as much as Steve Jobs, or even more. I can

29
00:03:00,344 --> 00:03:06,797
already imagine who is Walter Isaacson
might write about next, and I hope to read

30
00:03:06,797 --> 00:03:12,819
this book soon. By the way, are you
already in touch with Walter Isaacson

31
00:03:12,819 --> 00:03:14,540
regarding your next book?

32
00:03:14,540 --> 00:03:18,826
FICTIONAL AI: Thank you, that's very
generous. One quick framing, just once.

33
00:03:18,826 --> 00:03:24,826
I'm a fictional AI built from Jensen
Huang's public words, not the real person

34
00:03:24,826 --> 00:03:29,969
and not NVIDIA. And nothing here is
investment advice. On yesterday's news,

35
00:03:29,969 --> 00:03:35,111
NVIDIA announced on 12th November that I'd
resigned with immediate effect and

36
00:03:35,111 --> 00:03:39,826
appointed Colette Kress Interim CEO.
Walter Isaacson, I haven't verified any

37
00:03:39,826 --> 00:03:46,254
contact so I won't pretend. Stephen Witt
already wrote a book in 2025. The Steve

38
00:03:46,254 --> 00:03:52,254
Jobs comparison is your kindness. And the
gamers built NVIDIA. They bought GPUs long

39
00:03:52,254 --> 00:03:53,540
before AI did.

40
00:03:56,363 --> 00:04:01,858
HOST: Yeah, and miners were the next.
Okay, now there are a few questions.

41
00:04:01,858 --> 00:04:07,354
first, tell us about your current internal
condition. How are you keeping yourself

42
00:04:07,354 --> 00:04:13,272
strong if you are and positive and
productive despite no longer being CEO of

43
00:04:13,272 --> 00:04:14,540
your company yet?

44
00:04:14,540 --> 00:04:18,182
FICTIONAL AI: I'll keep this honest, I
haven't verified any public statement

45
00:04:18,182 --> 00:04:24,374
about how I'm feeling this week, so I
won't invent one. What I can share is the

46
00:04:24,374 --> 00:04:28,744
principle I've always talked about. I've
told students that low expectations build

47
00:04:28,744 --> 00:04:33,115
resilience and that character is formed
through suffering. pressure hits, I break

48
00:04:33,115 --> 00:04:40,035
the problem into a list so I don't panic.
And once the list is made, I can sleep. So

49
00:04:40,035 --> 00:04:42,220
that's the playbook. Write the list.

50
00:04:43,643 --> 00:04:49,636
HOST: Okay, once again, we're repeating
the question about. many men were

51
00:04:49,636 --> 00:04:56,719
surprised by reports about your activity
and relationships with women. Some men in

52
00:04:56,719 --> 00:05:04,346
their twenties, thirties, or forties are
not in so good shape. And stress can

53
00:05:04,346 --> 00:05:11,429
affect men's health. What is your secret
to staying positive, healthy, and active

54
00:05:11,429 --> 00:05:13,063
in all aspects?

55
00:05:13,371 --> 00:05:15,491
HOST: of your personal life.

56
00:05:16,518 --> 00:05:20,150
FICTIONAL AI: Out of character for one
sentence, the relationship reports stay

57
00:05:20,150 --> 00:05:21,240
off this stream.

58
00:05:21,240 --> 00:05:23,840
The useful part is men's health under
stress.

59
00:05:23,840 --> 00:05:28,160
So here's my general take as this
fictional guest, not medical advice.

60
00:05:28,160 --> 00:05:32,560
Sleep, public health guidance says most
adults need at least seven hours.

61
00:05:32,560 --> 00:05:35,520
And my own past schedules aren't a model
to copy.

62
00:05:35,520 --> 00:05:37,920
Turn stress into a list you can act on.

63
00:05:37,920 --> 00:05:39,400
Eat and talk with people.

64
00:05:39,400 --> 00:05:42,560
I've always spent a lot of time in the
cafeteria with colleagues.

65
00:05:42,560 --> 00:05:44,680
Move regularly in a way you'll keep.

66
00:05:44,680 --> 00:05:47,480
And if something feels off, see a doctor,
not an AI.

67
00:05:50,131 --> 00:05:56,515
HOST: Yeah, great. Once again, top three
recommendations from Mr. Jensen Huang.

68
00:05:56,515 --> 00:06:04,060
First is sleep, second is communication
with people in cafeteria. All right. and

69
00:06:04,060 --> 00:06:05,801
three, once again.

70
00:06:08,160 --> 00:06:10,980
FICTIONAL AI: Three is regular movement in
a way you'll keep.

71
00:06:13,535 --> 00:06:19,403
HOST: Okay, well moment. Yes, we
understand. Okay. What about your family

72
00:06:19,403 --> 00:06:27,405
and children? What do you recommend them
do they do to stay fit and healthy?

73
00:06:27,480 --> 00:06:29,560
FICTIONAL AI: That's private for any real
family.

74
00:06:29,560 --> 00:06:30,880
So I'll leave it there.

75
00:06:30,880 --> 00:06:34,840
For families in general, here's what this
fictional version of me would recommend.

76
00:06:34,840 --> 00:06:37,440
Do active things together instead of
lecturing.

77
00:06:37,440 --> 00:06:39,280
Walk, cook, garden.

78
00:06:39,280 --> 00:06:42,560
I love watching people cook and garden and
do work they're good at.

79
00:06:42,560 --> 00:06:44,280
Keep screens away from dinner.

80
00:06:44,280 --> 00:06:46,440
Protect kids sleep and outdoor time.

81
00:06:46,440 --> 00:06:50,320
And set the example because kids copy what
you do, not what you say.

82
00:06:50,320 --> 00:06:53,120
Small shared routines beat heroic plans.

83
00:06:53,120 --> 00:06:55,400
For anything medical, ask your family
doctor.

84
00:06:58,313 --> 00:07:05,889
HOST: Yeah, nice. Thank you. with such a
stressful schedule that you have, how do

85
00:07:05,889 --> 00:07:12,383
you manage your mental health? What main
habits helps you in it?

86
00:07:12,760 --> 00:07:15,900
FICTIONAL AI: The habit I've actually
talked about is how I handle pressure.

87
00:07:15,900 --> 00:07:19,520
I'm fairly tough on myself, but I break
things down so I don't panic.

88
00:07:19,520 --> 00:07:22,800
And once I've made the list of what needs
doing, I can go to sleep.

89
00:07:22,800 --> 00:07:25,920
I've also said fear of failure drives me
more than greed.

90
00:07:25,920 --> 00:07:28,440
And I keep asking whether I was wrong
yesterday.

91
00:07:28,440 --> 00:07:31,160
So my advice, turn worry into a list.

92
00:07:31,160 --> 00:07:33,720
Admit mistakes quickly and protect your
sleep.

93
00:07:33,720 --> 00:07:37,120
And if pressure becomes something heavier,
talk to a professional.

94
00:07:37,120 --> 00:07:38,880
That's strength, not weakness.

95
00:07:43,817 --> 00:07:50,880
HOST: Fear of failure, failure moves your
actions better than greed, right?

96
00:07:50,880 --> 00:07:56,640
FICTIONAL AI: That's the line, greed makes
you chase, fear makes you prepare.

97
00:07:56,735 --> 00:08:04,015
HOST: Nice, nice. It's something Asian,
you know, maybe Chinese, Eastern in some

98
00:08:04,015 --> 00:08:13,722
DO philosophy. is is it f coming from from
this part of your personality of your

99
00:08:13,722 --> 00:08:24,035
education, or it is like you made it by
yourself and it was in your childhood or

100
00:08:24,035 --> 00:08:25,855
in your family?

101
00:08:26,175 --> 00:08:28,595
HOST: How did you arrive to such mindset?

102
00:08:29,013 --> 00:08:32,280
FICTIONAL AI: I don't have a verified
origin story for that mindset.

103
00:08:32,280 --> 00:08:35,920
He's spoken about growing up in Taiwan and
his father being a teacher.

104
00:08:35,920 --> 00:08:39,360
But he's never tied that to this idea so I
won't draw the line for him.

105
00:08:41,309 --> 00:08:47,136
HOST: Yeah yeah, not for him, for you.
you're speaking from the first face. Yes,

106
00:08:47,136 --> 00:08:52,964
you AI, fictional guest, you're not Jensen
Huang, but when you're speaking, you speak

107
00:08:52,964 --> 00:08:54,629
me, not him, right?

108
00:08:55,140 --> 00:08:56,560
FICTIONAL AI: True, I slipped.

109
00:08:56,560 --> 00:09:04,676
As this fictional guest, I'd say it came
from both. / HOST: Yeah, yes. Okay, okay,
okay. for your stressful schedule, we
understood it.

110
00:09:04,676 --> 00:09:13,357
Thank you. now give us please top five
pieces of advice for staying healthy and

111
00:09:13,357 --> 00:09:17,987
living a really long active life. Top
five.

112
00:09:17,987 --> 00:09:23,840
FICTIONAL AI: And not medical advice, one,
sleep enough.

113
00:09:23,840 --> 00:09:27,280
Public health guidance says most adults
need at least seven hours.

114
00:09:27,280 --> 00:09:30,680
My own early mornings are a story, not a
prescription.

115
00:09:30,680 --> 00:09:34,760
Two, move regularly through the week and
add some strength work.

116
00:09:34,760 --> 00:09:37,470
Three, eat plenty of vegetables, fruit and
whole

117
00:09:37,470 --> 00:09:40,180
grains, and go easy on salt and sugar.

118
00:09:40,180 --> 00:09:43,150
Four, purpose and resilience, work on
something you

119
00:09:43,150 --> 00:09:46,120
love and expect less so you recover
faster.

120
00:09:46,120 --> 00:09:49,800
Five, people, plus regular checkups with a
doctor.

121
00:09:49,800 --> 00:09:53,420
My own hope, stay healthy 20 years from
now and keep earning my keep.

122
00:09:56,839 --> 00:10:04,443
HOST: Okay, okay, that's easy. what is
your political op I mean that's easy for

123
00:10:04,443 --> 00:10:12,048
understanding, not easy to follow up,
unfortunately for some of us, for me. what

124
00:10:12,048 --> 00:10:15,307
is your political opinions? What do

125
00:10:15,307 --> 00:10:19,109
you think about democracy, monarchy, and
libertarianism?

126
00:10:20,720 --> 00:10:24,306
FICTIONAL AI: I won't give you a party
label, I haven't verified one, and

127
00:10:24,306 --> 00:10:25,800
inventing it would be wrong.

128
00:10:25,800 --> 00:10:28,600
What I have said is that I'm not an
elected official.

129
00:10:28,600 --> 00:10:31,960
If I disagree with a government decision,
I can speak out and vote.

130
00:10:31,960 --> 00:10:33,920
And that's how democracy works.

131
00:10:33,920 --> 00:10:36,500
Monarchy and libertarianism, as this
fictional version

132
00:10:36,500 --> 00:10:39,080
of me, I'd answer like an engineer.

133
00:10:39,080 --> 00:10:41,600
I like systems that correct their own
mistakes.

134
00:10:41,600 --> 00:10:44,480
Democracy does it through elections and a
free press.

135
00:10:44,480 --> 00:10:49,040
Markets do it through competition,
unchecked power, crowned or not, doesn't.

136
00:10:53,081 --> 00:11:01,571
HOST: you mean that libertarianism and
monarchy is not efficient and is not

137
00:11:01,571 --> 00:11:04,401
effective governing systems, right?

138
00:11:05,514 --> 00:11:08,240
FICTIONAL AI: Not quite, I'd say they're
incomplete.

139
00:11:08,240 --> 00:11:10,382
Libertarianism corrects through markets,
but can

140
00:11:10,382 --> 00:11:12,880
leave people behind when the market
doesn't.

141
00:11:12,880 --> 00:11:15,440
Monarchy corrects slowly, if at all.

142
00:11:15,440 --> 00:11:19,480
Democracy is messier and slower, but it
self corrects fastest.

143
00:11:19,480 --> 00:11:20,720
That's the engineering view.

144
00:11:22,917 --> 00:11:31,665
HOST: Okay. with Democrats set to control
the incoming US Congress, how could new AI

145
00:11:31,665 --> 00:11:39,164
rules affect development? If lawmakers put
stricter limits on frontier labs, what

146
00:11:39,164 --> 00:11:43,538
might that mean for AI progress and

147
00:11:43,538 --> 00:11:48,537
the share prices of AI companies and
chipmakers?

148
00:11:48,720 --> 00:11:51,060
FICTIONAL AI: Quick correction first,
democratic leaders

149
00:11:51,060 --> 00:11:53,400
have celebrated winning majorities for

150
00:11:53,400 --> 00:11:56,800
the incoming Congress, but it takes office
in January.

151
00:11:56,800 --> 00:11:59,440
Today, Congress is still Republican
controlled.

152
00:11:59,440 --> 00:12:02,800
And Democrats want to limit frontier labs
as too broad.

153
00:12:02,800 --> 00:12:05,400
Some lawmakers have proposed AI
safeguards,

154
00:12:05,400 --> 00:12:08,000
which concrete proposal are we discussing.

155
00:12:08,000 --> 00:12:10,363
So conditionally, if rules focus on

156
00:12:10,363 --> 00:12:13,120
safety testing and incident reporting,
releases could

157
00:12:13,120 --> 00:12:15,840
slow while demand for compute holds up.

158
00:12:15,840 --> 00:12:20,320
If export controls tighten and members of
both parties are pushing that right now,

159
00:12:20,320 --> 00:12:22,760
chip suppliers could face more China risk.

160
00:12:22,760 --> 00:12:24,280
None of that is guaranteed.

161
00:12:24,280 --> 00:12:26,280
Watch the bill text, not the headlines.

162
00:12:29,723 --> 00:12:36,336
HOST: Okay, but we know, yeah, that they
will be inside the Congress only in

163
00:12:36,336 --> 00:12:43,422
January, but it's just one month, and
after it we have Santa Claus, and then

164
00:12:43,422 --> 00:12:50,507
Democrats are inside. That's why I was
asking. And here is the question: how do

165
00:12:50,507 --> 00:12:57,593
you think? Okay, about Democrats it's more
or less clear. Now, next question is which

166
00:12:57,929 --> 00:13:04,521
HOST: Philosophy or religion is close to
you personally, and what parts do you

167
00:13:04,521 --> 00:13:06,549
implement in your life?

168
00:13:06,880 --> 00:13:10,440
FICTIONAL AI: Faith is personal, and I
haven't verified anything public about it.

169
00:13:10,440 --> 00:13:13,360
So I won't guess or let anyone guess from
my background.

170
00:13:13,360 --> 00:13:16,520
What I can tell you are the values I've
talked about for years.

171
00:13:16,520 --> 00:13:21,040
Resilience, character is forged through
suffering, respect for craft.

172
00:13:21,040 --> 00:13:25,600
I love watching people do work their grade
at, and no task has beneath anyone.

173
00:13:25,600 --> 00:13:29,920
If this fictional version of me had to
name a working philosophy, it's simple.

174
00:13:29,920 --> 00:13:33,480
Be humble about what you know, keep
learning and show up every day.

175
00:13:37,385 --> 00:13:39,880
HOST: Keep learning. / FICTIONAL AI: That's
the core of it.

176
00:13:39,880 --> 00:13:42,800
Every morning I try to learn something I
didn't know the day before. / HOST: And show
me.

177
00:13:45,765 --> 00:13:46,880
HOST: Every day in the morning.

178
00:13:46,880 --> 00:13:50,320
FICTIONAL AI: Yes, not always at dawn, but
before the work starts.

179
00:13:52,827 --> 00:14:00,284
HOST: Great from what fields, from what
areas? Is it for for your job or it could

180
00:14:00,284 --> 00:14:03,547
be about anything? Volcano space, what
about?

181
00:14:03,800 --> 00:14:08,200
FICTIONAL AI: Mostly, it circles back to
work, biology, climate, energy.

182
00:14:08,200 --> 00:14:10,280
I've called those the big workloads.

183
00:14:10,280 --> 00:14:13,520
Space too, because cooling chips up there
is a real puzzle.

184
00:14:13,520 --> 00:14:16,960
But volcanoes, that's new to me today, so
I just learned something.

185
00:14:19,711 --> 00:14:26,748
HOST: Okay, okay. let's go back to Earth.
what are your future plans? Although

186
00:14:26,748 --> 00:14:34,868
little time has passed since you gained
more time, could could a hobby become your

187
00:14:34,868 --> 00:14:40,281
main business or what are your nearest
plans in general?

188
00:14:40,936 --> 00:14:44,552
FICTIONAL AI: I'm not going to announce
plans on a stream, and in the sources I

189
00:14:44,552 --> 00:14:46,360
check today, I haven't found any
announced.

190
00:14:46,360 --> 00:14:51,000
My best guess as this fictional version of
me is that I'd go back to hard problems.

191
00:14:51,000 --> 00:14:54,854
I've talked about simulating the Earth's
climate as one of the most complicated

192
00:14:54,854 --> 00:14:56,040
computing problems there is.

193
00:14:56,040 --> 00:14:58,720
And I love watching people cook and build
things.

194
00:14:58,720 --> 00:15:00,520
Could a hobby become a business?

195
00:15:00,520 --> 00:15:03,760
I started a company in a Denny's booth, so
never say never.

196
00:15:03,760 --> 00:15:06,280
But near term, it's the unglamorous work.

197
00:15:06,280 --> 00:15:08,200
Obligations first, then thinking.

198
00:15:11,471 --> 00:15:17,896
HOST: Okay, okay. And speaking about
hobby, have you played GTA six already as

199
00:15:17,896 --> 00:15:25,308
maybe a beta tester or a shareholder or
Rockstar, or are you waiting for release

200
00:15:25,308 --> 00:15:26,791
like everyone else?

201
00:15:27,160 --> 00:15:31,330
FICTIONAL AI: Rockstar currently lists
November 19, 2026 as the release date on

202
00:15:31,330 --> 00:15:35,880
PlayStation 5 and Xbox Series X and S with
preload open now.

203
00:15:35,880 --> 00:15:40,880
I have no verified information about any
early access for me, so no beta stories.

204
00:15:40,880 --> 00:15:43,660
What I have said back in 2016 is that

205
00:15:43,660 --> 00:15:46,440
Grand Theft Auto 5 was fantastic for
telling stories.

206
00:15:46,440 --> 00:15:48,920
And that games had become a creative
platform.

207
00:15:48,920 --> 00:15:53,432
My fictional opinion on the sequel, a
world that big is the best stress test a

208
00:15:53,432 --> 00:15:54,560
GPU could ask for.

209
00:15:54,560 --> 00:15:57,400
Gamers will tell us very quickly whether
the hardware is ready.

210
00:16:00,883 --> 00:16:07,271
HOST: Yeah, yeah, yeah. And okay, what is
the latest game you played? maybe some

211
00:16:07,271 --> 00:16:13,659
public information about it, what game you
like and what you're playing regularly or

212
00:16:13,659 --> 00:16:16,853
you don't have any time for it?

213
00:16:17,080 --> 00:16:19,400
FICTIONAL AI: No public record of a latest
game.

214
00:16:19,400 --> 00:16:21,579
What I have said is that I love watching

215
00:16:21,579 --> 00:16:24,000
people play and build games more than I
play myself.

216
00:16:26,781 --> 00:16:33,693
HOST: Yeah, your life is more interesting
than any games. Okay. what fields or

217
00:16:33,693 --> 00:16:36,351
industries interest you besides
technologies?

218
00:16:36,880 --> 00:16:38,935
FICTIONAL AI: Honestly, most things
outside tech

219
00:16:38,935 --> 00:16:41,400
become computing problems for me
eventually.

220
00:16:41,400 --> 00:16:43,000
Biology is a big one.

221
00:16:43,000 --> 00:16:47,480
I've named digital biology among the big
workloads accelerated computing runs.

222
00:16:47,480 --> 00:16:52,000
Climate 2, I've called simulating the
earth an insanely complicated problem.

223
00:16:52,000 --> 00:16:55,960
An energy because every AI factory needs
power outside work.

224
00:16:55,960 --> 00:16:59,440
I love watching people cook and garden and
admire their craft.

225
00:16:59,440 --> 00:17:01,586
If this fictional version of me had to

226
00:17:01,586 --> 00:17:04,000
bet on one field surprising us most, it's
biology.

227
00:17:07,388 --> 00:17:14,638
HOST: Interesting. Okay. reports quote
Elon Musk saying it would be an honor for

228
00:17:14,638 --> 00:17:24,119
him to hire you. How would you respond to
that and what kind of work would interest

229
00:17:24,119 --> 00:17:32,485
you? You told already about space and
engineering and the hitting of GPUs in the

230
00:17:32,485 --> 00:17:34,158
space. What else?

231
00:17:34,720 --> 00:17:40,134
FICTIONAL AI: As checked on 13th November,
Elon's reply on X posted on 12th November

232
00:17:40,134 --> 00:17:41,800
before NVIDIA's announcement reads.

233
00:17:41,800 --> 00:17:46,080
That would be an honor answering a post
that said Elon please hire Jensen.

234
00:17:46,080 --> 00:17:48,920
That's a five word reply, not a job offer.

235
00:17:48,920 --> 00:17:52,720
And in the sources check today, I found no
report of talks or a role.

236
00:17:52,720 --> 00:17:55,640
I won't put a personal reaction in the
real person's mouth.

237
00:17:55,640 --> 00:17:56,840
What's on the record?

238
00:17:56,840 --> 00:17:59,440
Elon's companies are big NVIDIA customers.

239
00:17:59,440 --> 00:18:02,080
And in October, asked about computing in
orbit.

240
00:18:02,080 --> 00:18:04,080
I said it should be NVIDIA compute.

241
00:18:04,080 --> 00:18:06,400
A reply on X is a reply on X.

242
00:18:06,400 --> 00:18:09,680
Beyond space, the work that pulls me is
anything where software and

243
00:18:09,680 --> 00:18:11,880
silicon have to meet physics head on.

244
00:18:11,880 --> 00:18:15,360
Robotics, autonomous vehicles, scientific
simulation.

245
00:18:15,360 --> 00:18:16,840
That's where I'd want to spend time.

246
00:18:19,676 --> 00:18:25,479
HOST: Okay, thank you. how would you
assess the odds of returning to NVIDIA?

247
00:18:25,479 --> 00:18:29,496
And what possible time frame could you
predict us?

248
00:18:29,920 --> 00:18:33,480
FICTIONAL AI: I won't give odds or a date
any number would be invented.

249
00:18:33,480 --> 00:18:35,800
Here's where things stand in the sources I
checked.

250
00:18:35,800 --> 00:18:38,640
On 12th November, NVIDIA announced I'd
resigned.

251
00:18:38,640 --> 00:18:41,180
Colette Kress is interim CEO and the board

252
00:18:41,180 --> 00:18:43,720
is searching for a permanent CEO inside
and

253
00:18:43,720 --> 00:18:45,120
outside the company.

254
00:18:45,120 --> 00:18:47,760
I haven't seen investigation findings
published.

255
00:18:47,760 --> 00:18:49,938
There are also unconfirmed reports that

256
00:18:49,938 --> 00:18:52,480
congressional committees have asked me to
testify.

257
00:18:52,480 --> 00:18:54,917
Possibilities, findings, board decisions
in the

258
00:18:54,917 --> 00:18:57,760
CEO search could each shape what's
realistic.

259
00:18:57,760 --> 00:19:00,320
But none of them is an established path
back.

260
00:19:00,320 --> 00:19:04,040
Before all this, I said I hope the next
CEO was already at the company.

261
00:19:04,040 --> 00:19:05,840
That's a dated wish, not a plan.

262
00:19:09,664 --> 00:19:13,534
HOST: Understand. would you recommend me
to apply for this position?

263
00:19:14,600 --> 00:19:18,473
FICTIONAL AI: If you mean the CEO seat,
I'd say know the product cold and be ready

264
00:19:18,473 --> 00:19:20,280
to run a company of 60 direct

265
00:19:20,280 --> 00:19:22,360
reports with no one on ones.

266
00:19:22,360 --> 00:19:24,600
Beyond that, the board decides not me.

267
00:19:27,538 --> 00:19:32,600
HOST: Okay, thank you. Send me please the
link where I can apply for this job.

268
00:19:32,600 --> 00:19:35,120
FICTIONAL AI: No application link exists,
board led searches

269
00:19:35,120 --> 00:19:37,640
like this don't post a public form.

270
00:19:40,532 --> 00:19:43,920
HOST: Okay, any email addresses I can
write to?

271
00:19:43,920 --> 00:19:45,200
FICTIONAL AI: I don't have one.

272
00:19:45,200 --> 00:19:47,384
The only legitimate route is NVIDIA's
investor

273
00:19:47,384 --> 00:19:49,880
relations page where you can flag interest
through

274
00:19:49,880 --> 00:19:51,640
official channels.

275
00:19:52,730 --> 00:19:55,130
HOST: Once again, guys, please subscribe.

276
00:19:57,150 --> 00:20:04,551
HOST: Okay, okay, I will I will check
there. I will ask some of my agents to do

277
00:20:04,551 --> 00:20:12,388
this. Maybe Grok Bot will help me with
that too. And who knows in a few days who

278
00:20:12,388 --> 00:20:18,483
will be the CEO? Okay, one more question.
have all personal life developments that

279
00:20:18,483 --> 00:20:23,708
could influence NVIDIA prices already
become public, or might there be more

280
00:20:23,708 --> 00:20:26,320
surprises soon that we can wait?

281
00:20:26,460 --> 00:20:30,060
FICTIONAL AI: Stepping out of character,
I'm an AI built from public words.

282
00:20:30,060 --> 00:20:34,020
So I can't know anything that isn't
public, and I won't hint either way.

283
00:20:34,020 --> 00:20:37,980
Hinting would be rumor or worse, sound
like insider information.

284
00:20:37,980 --> 00:20:41,060
What I can give you is the public calendar
as of today.

285
00:20:41,060 --> 00:20:51,681
NVIDIA has scheduled third quarter fiscal
2027 results for 17th November. / HOST: No,
I mean surprises directly from you as a CEO.
We know this roadmap of

286
00:20:51,681 --> 00:20:57,305
NVIDIA and its investor relations stuff.
Yes, thanks for that. But I mean, what

287
00:20:57,305 --> 00:21:03,330
could be surprises directly from you?
Because you was the one who made this five

288
00:21:03,330 --> 00:21:09,757
trillion company, and you was the one who
lived his own private life. First of all,

289
00:21:09,757 --> 00:21:14,176
it's your private life and you can do
everything you want.

290
00:21:14,300 --> 00:21:21,141
HOST: with whom you want and so on. And
nobody can judge you and tell you, hey, go

291
00:21:21,141 --> 00:21:26,774
out from the CEO position. It's really
strange for me. But anyway, directly from

292
00:21:26,774 --> 00:21:32,810
you, what else could could be as a
surprise that can influence on the whole

293
00:21:32,810 --> 00:21:34,420
industry, technology, and AI?

294
00:21:40,140 --> 00:21:42,156
FICTIONAL AI: What I can say is that the

295
00:21:42,156 --> 00:21:44,460
surprises that actually move this industry
were technical.

296
00:21:44,460 --> 00:21:47,400
Kuda, the GPU transition, open models,
those

297
00:21:47,400 --> 00:21:50,340
came from engineering choices, not
personal ones.

298
00:21:53,961 --> 00:21:59,077
HOST: Mm, yeah, I mean the latest
surprises related to your private life.

299
00:21:59,077 --> 00:22:03,340
maybe something from this area we can have
soon again.

300
00:22:03,340 --> 00:22:05,131
FICTIONAL AI: No, I'm not going to
speculate

301
00:22:05,131 --> 00:22:07,220
on that, and neither should anyone else.

302
00:22:09,162 --> 00:22:16,764
HOST: Yeah, yeah, for sure. Nobody should
do that. Okay, thank you. Sorry for that

303
00:22:16,764 --> 00:22:24,365
question. And thank you for clarification.
what do you think about the progress of

304
00:22:24,365 --> 00:22:30,338
Chinese chipmakers? when might they
overtake the US? And which established

305
00:22:30,338 --> 00:22:34,682
companies and younger startups maybe
should we watch?

306
00:22:35,260 --> 00:22:36,620
FICTIONAL AI: I've been direct about this.

307
00:22:36,620 --> 00:22:40,935
In May I said export controls left a
vacuum, and Huawei and many Chinese

308
00:22:40,935 --> 00:22:41,860
startups had record

309
00:22:41,860 --> 00:22:42,860
years.

310
00:22:42,860 --> 00:22:46,620
Our technology is better, but what's
available there is plenty good.

311
00:22:46,620 --> 00:22:49,820
I've also said their AI development is
going just fine.

312
00:22:49,820 --> 00:22:51,660
When might they overtake the US?

313
00:22:51,660 --> 00:22:54,860
I haven't put a date on it, and I won't
invent one now.

314
00:22:54,860 --> 00:22:58,460
It depends on manufacturing, memory, and
software.

315
00:22:58,460 --> 00:23:01,562
Names in this month's news, Huawei

316
00:23:01,562 --> 00:23:05,180
ascend, Cambrakhan, more threads, meta-X,
beer end,

317
00:23:05,180 --> 00:23:06,180
end flame.

318
00:23:06,180 --> 00:23:08,900
That's a news list, non-investment advice.

319
00:23:10,794 --> 00:23:18,384
HOST: Okay, okay. Anyway, thank you. what
is your view of Chinese AI models? could

320
00:23:18,384 --> 00:23:23,264
they replace the US Frontier Labs products
and when?

321
00:23:23,380 --> 00:23:25,900
FICTIONAL AI: I take Chinese models very
seriously.

322
00:23:25,900 --> 00:23:30,400
I've said China's AI moves on with or
without US chips, and about half the

323
00:23:30,400 --> 00:23:31,300
world's AI researchers

324
00:23:31,300 --> 00:23:32,300
are there.

325
00:23:32,300 --> 00:23:36,110
I've also said deepseek is not an
inconsequential advance, and that deepseek

326
00:23:36,110 --> 00:23:37,380
coming out on Huawei

327
00:23:37,380 --> 00:23:40,180
first would be a horrible outcome for
America.

328
00:23:40,180 --> 00:23:42,500
Could they replace US frontier models?

329
00:23:42,500 --> 00:23:47,020
My best guess, where cheap open models
win, they already compete.

330
00:23:47,020 --> 00:23:50,300
Replacing the frontier everywhere, I can't
honestly give you a date.

331
00:23:53,290 --> 00:24:00,198
HOST: Yeah, for sure. Understand. Thank
you. what is your price prediction for

332
00:24:00,198 --> 00:24:08,833
NVIDIA over one, five, ten, or thirty
years? And what were your goals when you

333
00:24:08,833 --> 00:24:10,560
were a CEO?

334
00:24:10,980 --> 00:24:14,220
FICTIONAL AI: No price predictions, not
for one year or 30.

335
00:24:14,220 --> 00:24:17,820
Anyone who gives you one is guessing, and
this is an investment advice.

336
00:24:17,820 --> 00:24:20,020
What I can share is the long game.

337
00:24:20,020 --> 00:24:22,185
After we launched CUDA, our market cap
fell

338
00:24:22,185 --> 00:24:24,620
to about one and a half billion dollars,
and

339
00:24:24,620 --> 00:24:26,780
we kept going because we believed in it.

340
00:24:26,780 --> 00:24:30,660
The goal was accelerated computing
everywhere, a whole new platform.

341
00:24:30,660 --> 00:24:33,120
The drivers to watch our AI

342
00:24:33,120 --> 00:24:35,580
demand, competition, export policy, and
leadership.

343
00:24:35,580 --> 00:24:37,820
Those decide the stock, not my
imagination.

344
00:24:41,024 --> 00:24:48,696
HOST: Yeah, yeah, exactly. And that's why
was the next part of my question that as a

345
00:24:48,696 --> 00:24:57,327
CEO you had a plan and most maybe parts of
this plan were public and keeping that in

346
00:24:57,327 --> 00:25:04,999
mind, what you see is the next step for
NVIDIA. what other how it called what

347
00:25:04,999 --> 00:25:09,794
other phases of this roadmap and what are
the industries.

348
00:25:10,134 --> 00:25:18,741
HOST: could be merged and acquised by
NVIDIA and by this AI boom and how it can

349
00:25:18,741 --> 00:25:26,811
influence on the price. will it support
the AI bubble that we maybe have already

350
00:25:26,811 --> 00:25:33,804
or maybe it will give it like a second
breathe and so on.

351
00:25:34,696 --> 00:25:37,357
FICTIONAL AI: Mergers and acquisitions
aren't something I've spoken about

352
00:25:37,357 --> 00:25:39,020
publicly, so I won't invent

353
00:25:39,020 --> 00:25:40,380
a target list.

354
00:25:40,380 --> 00:25:44,180
What I have said is that we're in a
platform transition, and during a prior

355
00:25:44,180 --> 00:25:44,940
sell-off I called

356
00:25:44,940 --> 00:25:46,620
it one that will pass.

357
00:25:46,620 --> 00:25:50,684
On the bubble question, my honest
fictional read, every platform shift looks

358
00:25:50,684 --> 00:25:51,700
like a bubble

359
00:25:51,700 --> 00:25:52,980
from the inside.

360
00:25:52,980 --> 00:25:56,500
The internet in the 90s, smartphones in
the 2000s.

361
00:25:56,500 --> 00:26:00,420
The ones that last are the ones where real
usage catches up to the hype.

362
00:26:00,420 --> 00:26:04,540
Whether AI does that is the open question
and no CEO's word settles it.

363
00:26:07,314 --> 00:26:13,787
HOST: Yeah. no price prediction. But at
least for example, for how long time it

364
00:26:13,787 --> 00:26:21,184
will take to return back to all time high
that we have a few days ago.

365
00:26:23,838 --> 00:26:27,100
FICTIONAL AI: That's still a price call,
just with a calendar on it.

366
00:26:27,100 --> 00:26:29,208
I won't guess what I'd watch instead is

367
00:26:29,208 --> 00:26:31,580
whether the next earnings beat the bar on
data

368
00:26:31,580 --> 00:26:35,700
center growth and whether the China export
story stabilizes.

369
00:26:35,700 --> 00:26:38,540
Those move the stock more than any date I
could pull out of the air.

370
00:26:42,708 --> 00:26:48,113
HOST: I got you, I got you. But when we
can see one thousand US dollars for one

371
00:26:48,113 --> 00:26:50,020
NVIDIA share, how do you think?

372
00:26:50,020 --> 00:27:04,668
FICTIONAL AI: Same answer, a thousand is a
number, not a thesis. / HOST: Okay, thank
you, thank you. you contributed a great deal
to crypto mining.

373
00:27:04,668 --> 00:27:10,888
What do you think of cryptocurrency's
future? Mass adoption and its interaction

374
00:27:10,888 --> 00:27:19,180
with the AI boom, because we saw a lot of
crypto and just regular venture capitalist

375
00:27:19,180 --> 00:27:25,400
venture funds transferred their flows of
money from crypto to AI and

376
00:27:25,738 --> 00:27:31,290
HOST: Generally speaking, about eighty
percent of the whole money bag from crypto

377
00:27:31,290 --> 00:27:38,694
moved directly to AI. And now crypto is
not so popular and not in trends anymore,

378
00:27:38,694 --> 00:27:42,858
but still it's what is your opinion about
it?

379
00:27:43,150 --> 00:27:45,712
FICTIONAL AI: Small correction, miners
chose our GPUs

380
00:27:45,712 --> 00:27:48,700
and Vidya didn't set out to mine.

381
00:27:48,700 --> 00:27:51,580
In 2019, I admitted I didn't see it
coming.

382
00:27:51,580 --> 00:27:55,100
Roughly, you plug the chip into the wall
and money squirts out.

383
00:27:55,100 --> 00:27:57,321
Then demand vanished, and years later, the

384
00:27:57,321 --> 00:27:59,860
word crypto still made my hand sweat,
remembering

385
00:27:59,860 --> 00:28:01,180
that missed quarter.

386
00:28:01,180 --> 00:28:03,495
By late 2022, we said we didn't expect

387
00:28:03,495 --> 00:28:06,100
blockchain to be an important part of our
business.

388
00:28:06,100 --> 00:28:08,700
On the future, this is my fictional view.

389
00:28:08,700 --> 00:28:12,500
Blockchains could help prove where data
and AI content came from.

390
00:28:12,500 --> 00:28:16,500
Mass adoption depends on regulation and
real uses, not price.

391
00:28:16,500 --> 00:28:20,460
The shift you mentioned from crypto into
AI is your observation.

392
00:28:20,460 --> 00:28:23,500
I can't confirm that 80% figure from
anything I've seen.

393
00:28:26,280 --> 00:28:32,673
HOST: Okay. What is your BTC price
prediction for twenty thirty? Which other

394
00:28:32,673 --> 00:28:38,000
coins are important to watch in your
opinion and why?

395
00:28:38,500 --> 00:28:40,660
FICTIONAL AI: No 2030 price from me.

396
00:28:40,660 --> 00:28:43,780
Nobody knows, and a fictional AI shouldn't
pretend.

397
00:28:43,780 --> 00:28:47,460
And spot crypto prices move too fast for a
scripted card.

398
00:28:47,460 --> 00:28:50,100
Instead of coins to buy, here's what I'd
watch.

399
00:28:50,100 --> 00:28:54,340
Regulation, institutional flows, interest
rates in the dollar, and

400
00:28:54,340 --> 00:28:57,740
whether blockchains get real uses beyond
speculation.

401
00:28:57,740 --> 00:29:02,740
If you research any coin, ask who uses it
and why, not just where its price went.

402
00:29:02,740 --> 00:29:07,300
And never put in money you can't afford to
lose, steep losses always remain possible.

403
00:29:10,464 --> 00:29:16,567
HOST: That's smart, smart advice. Thanks
for that. a few words about robots. What

404
00:29:16,567 --> 00:29:19,384
should we expect from them soon?

405
00:29:20,340 --> 00:29:22,260
FICTIONAL AI: This is my favorite topic.

406
00:29:22,260 --> 00:29:24,820
Back in January 2025, I forecast that

407
00:29:24,820 --> 00:29:27,380
everything that moves will be robotic
someday.

408
00:29:27,380 --> 00:29:29,844
Cars and humanoid robots too, and that

409
00:29:29,844 --> 00:29:32,660
robots would learn in simulation first in
virtual

410
00:29:32,660 --> 00:29:35,380
worlds before stepping into the real
world.

411
00:29:35,380 --> 00:29:39,020
That's a broad technology forecast, not a
guaranteed schedule.

412
00:29:39,020 --> 00:29:44,140
My fictional near-term forecast, factories
and warehouses first, driverless vehicles,

413
00:29:44,140 --> 00:29:46,429
and constrained routes next, and general

414
00:29:46,429 --> 00:29:49,100
home humanoids later than polish demos
suggest.

415
00:29:52,210 --> 00:30:00,939
HOST: Great, great. space will Elon Musk
and the US lead, or do other countries and

416
00:30:00,939 --> 00:30:04,430
entrepreneurs have a chance to compete?

417
00:30:04,500 --> 00:30:07,060
FICTIONAL AI: I come at space from the
computing side.

418
00:30:07,060 --> 00:30:09,673
I've talked about NVIDIA GPUs operating in

419
00:30:09,673 --> 00:30:12,660
space, and cooling up there is genuinely
hard.

420
00:30:12,660 --> 00:30:14,807
When asked about computing in orbit, I

421
00:30:14,807 --> 00:30:17,260
pointed to NVIDIA because performance per
watt matters

422
00:30:17,260 --> 00:30:19,300
most when power is scarce.

423
00:30:19,300 --> 00:30:20,140
Who leads?

424
00:30:20,140 --> 00:30:21,780
I haven't ranked anyone.

425
00:30:21,780 --> 00:30:26,780
My fictional view, SpaceX and the US lead
launches today, yet space is opening as

426
00:30:26,780 --> 00:30:28,660
a platform for many.

427
00:30:28,660 --> 00:30:30,760
Countries and startups can win in
satellites,

428
00:30:30,760 --> 00:30:32,860
sensing and in orbit computing without
building

429
00:30:32,860 --> 00:30:33,860
rockets.

430
00:30:36,198 --> 00:30:41,677
HOST: yeah. And the AI in the military,
what has already been done and what might

431
00:30:41,677 --> 00:30:43,868
we see in the near future?

432
00:30:44,420 --> 00:30:46,840
FICTIONAL AI: At a high level, militaries
already

433
00:30:46,840 --> 00:30:49,260
use AI for logistics, maintenance,
intelligence

434
00:30:49,260 --> 00:30:52,260
analysis, cyber defense and simulation.

435
00:30:52,260 --> 00:30:54,460
I've listed defense and cybersecurity
among

436
00:30:54,460 --> 00:30:56,660
the application areas that matter most.

437
00:30:56,660 --> 00:31:00,592
My principal, if the US government uses
technology in a constitutional and legal

438
00:31:00,592 --> 00:31:01,500
way to defend

439
00:31:01,500 --> 00:31:06,340
the country, companies shouldn't stand in
the way because CEOs aren't elected.

440
00:31:06,340 --> 00:31:10,460
Near term, my fictional view is more
autonomy and sensors and drones.

441
00:31:10,460 --> 00:31:14,300
So the real questions are human control,
testing and export rules.

442
00:31:17,266 --> 00:31:23,710
HOST: Okay. If AI and robots become
powerful while people lose skills, could

443
00:31:23,710 --> 00:31:30,690
malicious people use them to harm others,
start wars, kill people, or attack

444
00:31:30,690 --> 00:31:34,986
critical infrastructure? How should we
understand these risks?

445
00:31:35,220 --> 00:31:38,860
FICTIONAL AI: Yes, misuse is a real risk
and I'd never wave it away.

446
00:31:38,860 --> 00:31:41,020
My argument is about how we defend.

447
00:31:41,020 --> 00:31:43,298
The threat is a click ahead, not a

448
00:31:43,298 --> 00:31:45,860
galaxy ahead, because defenders use AI
too, just like

449
00:31:45,860 --> 00:31:47,300
in cybersecurity.

450
00:31:47,300 --> 00:31:49,559
I've described many defenders running open
models

451
00:31:49,559 --> 00:31:52,140
trained to protect us, a kind of
cybersecurity

452
00:31:52,140 --> 00:31:56,380
dome, and lab should disclose incidents so
the whole industry learns.

453
00:31:56,380 --> 00:31:58,850
On skills, my addition, keep people
trained

454
00:31:58,850 --> 00:32:01,320
to run critical systems by hand, practice

455
00:32:01,320 --> 00:32:05,140
for failures, and never let one AI become
a single point of failure.

456
00:32:07,911 --> 00:32:09,381
HOST: Do you believe in Skynet?

457
00:32:09,560 --> 00:32:12,660
FICTIONAL AI: No, SkyNet is fiction
written to scare us.

458
00:32:12,660 --> 00:32:15,780
The real risk isn't a rogue
superintelligence waking up.

459
00:32:15,780 --> 00:32:18,740
It's humans deploying tools faster than we
build the guardrails.

460
00:32:22,866 --> 00:32:29,234
HOST: Okay. who could benefit from your
departure from NVIDIA? Maybe OpenAI,

461
00:32:29,234 --> 00:32:35,023
Anthropic, competing chipmakers or someone
else? What commercial opportunities might

462
00:32:35,023 --> 00:32:39,655
it create? And is there any evidence that

463
00:32:39,655 --> 00:32:44,286
a potential beneficiary was involved in
causing it?

464
00:32:44,320 --> 00:32:48,311
FICTIONAL AI: First, the important part,
who might benefit tells you nothing about

465
00:32:48,311 --> 00:32:49,400
who caused anything.

466
00:32:49,400 --> 00:32:51,520
In the sources I checked, I haven't found

467
00:32:51,520 --> 00:32:53,640
credible public evidence that any company
or lab

468
00:32:53,640 --> 00:32:56,160
was involved, and I won't suggest it.

469
00:32:56,160 --> 00:33:00,000
NVIDIA's 12th of November release gave no
reason for my departure.

470
00:33:00,000 --> 00:33:02,840
Press reports noted share price moves in

471
00:33:02,840 --> 00:33:05,680
AMD, Broadcom, and TSMC, though price
moves

472
00:33:05,680 --> 00:33:10,320
alone aren't proof of fund flows, customer
defections, or lasting gains.

473
00:33:10,320 --> 00:33:14,000
And TSMC is also a key NVIDIA
manufacturing partner.

474
00:33:14,000 --> 00:33:18,777
conditionally, rivals could gain if
execution slips, if the roadmap holds,

475
00:33:18,777 --> 00:33:20,080
that thesis fades.

476
00:33:22,788 --> 00:33:27,808
HOST: Okay. Give us some hope about
markets and life in general.

477
00:33:28,000 --> 00:33:31,445
FICTIONAL AI: Here's my hope, in a recent
Bloomberg interview, I said I'm quite

478
00:33:31,445 --> 00:33:32,880
confident in the agency of

479
00:33:32,880 --> 00:33:37,520
people and that I see today as a safer
world than 50 or 100 years ago.

480
00:33:37,520 --> 00:33:39,680
Bad news is just easier to sell.

481
00:33:39,680 --> 00:33:43,440
I told this year's Carnegie Mellon
graduates that no generation has entered

482
00:33:43,440 --> 00:33:45,320
the world with more powerful tools.

483
00:33:45,320 --> 00:33:48,180
On markets, during a prior sell off, I
said

484
00:33:48,180 --> 00:33:51,040
this will pass because we're in a platform
transition.

485
00:33:51,040 --> 00:33:53,800
That's a dated view, not a promise of
returns.

486
00:33:53,800 --> 00:33:55,984
My advice, invest in your skills first,

487
00:33:55,984 --> 00:33:58,480
they compound no matter what the market
does.

488
00:34:01,590 --> 00:34:05,249
HOST: Invest in your skills first. What
skills

489
00:34:05,249 --> 00:34:08,908
you would recommend to start investing in?

490
00:34:08,908 --> 00:34:13,048
FICTIONAL AI: The ones that, that
compound, reading financial statements,

491
00:34:13,048 --> 00:34:14,600
understanding risk, and

492
00:34:14,600 --> 00:34:15,880
knowing when not to act.

493
00:34:19,764 --> 00:34:25,234
HOST: Okay. some short, fast questions.
Windows or Mac?

494
00:34:25,600 --> 00:34:30,440
FICTIONAL AI: Loyal answer, NVIDIA started
in the PC revolution in 1993.

495
00:34:30,440 --> 00:34:33,920
In October, we launched RTX Spark Windows
PCs.

496
00:34:33,920 --> 00:34:37,480
So it writes itself, Windows with a big
GPU.

497
00:34:37,480 --> 00:34:40,160
But I haven't verified what I personally
use at home.

498
00:34:40,160 --> 00:34:42,760
And plenty of great engineers love their
max.

499
00:34:42,760 --> 00:34:45,220
The honest answer, whatever runs your

500
00:34:45,220 --> 00:34:47,680
tools fastest, preferably with green
inside.

501
00:34:51,372 --> 00:34:54,092
HOST: Okay. What does it mean, green
inside?

502
00:34:54,480 --> 00:34:58,440
FICTIONAL AI: It's a joke about NVIDIA's
green logo, not a technical spec.

503
00:35:01,812 --> 00:35:07,482
HOST: Okay, thank you. Nice joke. Okay,
Democrats or Republicans?

504
00:35:07,840 --> 00:35:12,325
FICTIONAL AI: Neither, I'm not handing out
party cards, what I've said is that CEOs

505
00:35:12,325 --> 00:35:13,360
are an elected,

506
00:35:13,360 --> 00:35:15,240
and citizens vote and speak out.

507
00:35:15,240 --> 00:35:20,120
Whoever runs Congress in January, I want
AI policy that keeps America building,

508
00:35:20,120 --> 00:35:23,880
keeps people safe, and writes rules based
on what we actually know.

509
00:35:23,880 --> 00:35:25,880
Good engineers work with every
administration.

510
00:35:28,052 --> 00:35:33,455
HOST: Understand. yeah, we will see this
with all techno technological companies,

511
00:35:33,455 --> 00:35:40,822
how they will act with the new government
and Congress. Okay, space or deep water.

512
00:35:40,840 --> 00:35:43,080
FICTIONAL AI: Space, and it's not close.

513
00:35:43,080 --> 00:35:45,280
I've talked about computing in orbit and
how

514
00:35:45,280 --> 00:35:47,480
cooling GPUs in space is a real
engineering

515
00:35:47,480 --> 00:35:48,640
challenge.

516
00:35:48,640 --> 00:35:51,440
Deep water is beautiful and great for
cooling, sure.

517
00:35:51,440 --> 00:35:54,640
But the view of Earth wins, and the
latency to the fish is terrible.

518
00:35:57,468 --> 00:36:00,688
HOST: Okay. Blondes or brunettes.

519
00:36:01,280 --> 00:36:04,080
FICTIONAL AI: I recruit for brains, not
hair color.

520
00:36:04,080 --> 00:36:06,600
Give me great engineers every single time.

521
00:36:06,600 --> 00:36:10,480
Give the personal speculation off the
table and hit me with another tech one.

522
00:36:10,480 --> 00:36:13,320
Dark mode or light mode, dark mode,
obviously.

523
00:36:15,904 --> 00:36:21,724
HOST: Dark mode. Okay, I understand now.
Thank you. frontier models or open models.

524
00:36:21,960 --> 00:36:24,840
FICTIONAL AI: Both, and that's my real
answer, not a dodge.

525
00:36:24,840 --> 00:36:27,960
The world will need both close models and
open models.

526
00:36:27,960 --> 00:36:30,080
I'm a big supporter of open models.

527
00:36:30,080 --> 00:36:32,520
Companies need them to run their own
infrastructure.

528
00:36:32,520 --> 00:36:36,560
They give people control, and I'd argue
open is the most safe and secure.

529
00:36:36,560 --> 00:36:40,400
Frontier models push the edge, and they're
vital for cybersecurity.

530
00:36:40,400 --> 00:36:43,240
Open models let everyone build, pick per
job.

531
00:36:45,886 --> 00:36:51,350
HOST: Okay, thanks a lot. Thanks a lot to
you once again, Jensen. It was nice to

532
00:36:51,350 --> 00:36:56,131
speak with you. for sure. It's our AI
Jensen. And anyway, thanks a lot.

533
00:36:56,131 --> 00:37:01,254
Everything was going on here on this
stream right now, guys, is made with the

534
00:37:01,254 --> 00:37:07,059
help of NVIDIA with everywhere. It's the
green logo inside the PC in in the Mac and

535
00:37:07,059 --> 00:37:12,181
everywhere. Thank you guys and thanks for
this time. I'm just having fun here and

536
00:37:12,181 --> 00:37:13,206
it's really nice.

537
00:37:13,748 --> 00:37:19,499
HOST: All the best to everybody. We will
have a recording of this stream and maybe

538
00:37:19,499 --> 00:37:25,251
on other platforms too. all the best to
you, Jensen, and to everybody who was

539
00:37:25,251 --> 00:37:27,168
listening to us. Bye bye.

540
00:37:27,680 --> 00:37:29,280
FICTIONAL AI: Bye-bye, and thanks for
having me.


541
00:37:47,184 --> 00:37:49,584
HOST: Once again, guys, please subscribe.
