1
00:00:00,200 --> 00:00:02,120
Speaker 1: I thought a lot about what I would feel about

2
00:00:02,120 --> 00:00:03,880
Speaker 1: the world and my future kids would grow up in

3
00:00:04,120 --> 00:00:06,320
Speaker 1: Like my highest order bit is to not destroy the

4
00:00:06,320 --> 00:00:08,440
Speaker 1: world with AI, Like it doesn't matter how good everything

5
00:00:08,480 --> 00:00:09,200
Speaker 1: else is if we do that.

6
00:00:12,800 --> 00:00:14,319
Speaker 2: Today, I want to take you to the home of

7
00:00:14,360 --> 00:00:18,200
Speaker 2: Sam Altman. He's the CEO of Open AI. Now. When

8
00:00:18,239 --> 00:00:21,920
Speaker 2: I traveled there, I was struck by how green everything was.

9
00:00:22,320 --> 00:00:25,760
Speaker 2: There were rolling hills and birds chirping in the background,

10
00:00:25,840 --> 00:00:28,800
Speaker 2: and as we were driving in there were cows grazing

11
00:00:28,840 --> 00:00:32,000
Speaker 2: by the road. And I paint this picture because I

12
00:00:32,000 --> 00:00:35,199
Speaker 2: think it's a really interesting backdrop for a conversation on

13
00:00:35,320 --> 00:00:38,960
Speaker 2: the tech revolution. And before I interviewed Sam, who is

14
00:00:39,200 --> 00:00:41,960
Speaker 2: a founder that I've interviewed for the last fifteen years,

15
00:00:42,440 --> 00:00:45,840
Speaker 2: my team sat down with me. When I told people

16
00:00:45,880 --> 00:00:48,279
Speaker 2: it's going to come interview him, it was there's such

17
00:00:48,280 --> 00:00:54,240
Speaker 2: a visceral reaction. It's either people were so amazing, that's

18
00:00:54,240 --> 00:00:58,200
Speaker 2: so cool, or man like, that is so terrifying. Tell

19
00:00:58,280 --> 00:00:59,600
Speaker 2: him not to do this, and tell him not to

20
00:00:59,600 --> 00:01:03,720
Speaker 2: do this. Everyone cares. Everyone seems to care about artificial intelligence,

21
00:01:03,720 --> 00:01:05,600
Speaker 2: and there's a good reason why, and that's because it's

22
00:01:05,640 --> 00:01:07,720
Speaker 2: going to impact every single one of us. It's going

23
00:01:07,760 --> 00:01:09,400
Speaker 2: to impact your children, it's going to impact your parents,

24
00:01:09,440 --> 00:01:12,080
Speaker 2: it's going to impact you, it's going to impact everyone.

25
00:01:12,280 --> 00:01:15,160
Speaker 2: And I think the reason for all of that is

26
00:01:15,200 --> 00:01:18,000
Speaker 2: because this company that he's the CEO of is now

27
00:01:18,160 --> 00:01:20,680
Speaker 2: one of the most important companies in the world, and

28
00:01:20,920 --> 00:01:25,759
Speaker 2: the technology that's built at open Ai will impact every

29
00:01:25,800 --> 00:01:28,800
Speaker 2: facet of our lives. And there's a lot of fear

30
00:01:28,840 --> 00:01:31,320
Speaker 2: around that. There's a lot of hope, but it's really

31
00:01:31,360 --> 00:01:34,080
Speaker 2: this uncomfortable period that we're in, and I think Sam

32
00:01:34,120 --> 00:01:37,600
Speaker 2: Altman very much represents what happens when we're going through

33
00:01:37,600 --> 00:01:39,959
Speaker 2: a technological revolution and there are a lot of questions

34
00:01:40,000 --> 00:01:44,160
Speaker 2: because it feels really high stakes. The now of this

35
00:01:44,800 --> 00:01:48,120
Speaker 2: feels really important because I want to go see the

36
00:01:48,160 --> 00:01:51,240
Speaker 2: person that I've known for fifteen years and look at

37
00:01:51,280 --> 00:01:54,320
Speaker 2: who that person was and how that person is going

38
00:01:54,360 --> 00:01:58,480
Speaker 2: to stand in this moment. Sam, how long have we

39
00:01:58,520 --> 00:01:59,240
Speaker 2: known each other.

40
00:02:00,600 --> 00:02:01,320
Speaker 1: Almost twenty years.

41
00:02:01,480 --> 00:02:03,680
Speaker 2: We've known each other a really long time. So I'm

42
00:02:03,800 --> 00:02:06,120
Speaker 2: excited to sit down with you at this moment because

43
00:02:06,120 --> 00:02:08,399
Speaker 2: I sat down with you throughout your career and I've

44
00:02:08,400 --> 00:02:12,200
Speaker 2: seen all sorts of iterations of it. I would argue

45
00:02:12,240 --> 00:02:15,160
Speaker 2: this is the most important seat you have sat in.

46
00:02:15,360 --> 00:02:18,600
Speaker 2: For sure, the stakes feel really high. We are in

47
00:02:18,639 --> 00:02:22,440
Speaker 2: this moment where we're wondering if this technological innovation is

48
00:02:22,440 --> 00:02:26,080
Speaker 2: going to be incredible for all of us or incredible

49
00:02:26,120 --> 00:02:27,720
Speaker 2: for some of us. Do you feel that.

50
00:02:29,720 --> 00:02:32,160
Speaker 1: Yeah? I mean we are clearly entering the phase where

51
00:02:32,160 --> 00:02:34,160
Speaker 1: AI is going to be one of the high order

52
00:02:34,200 --> 00:02:38,480
Speaker 1: bits of how our future society gets shaped, and I

53
00:02:40,200 --> 00:02:44,880
Speaker 1: think people feel all of the excitement, fear, anxiety, hope,

54
00:02:44,919 --> 00:02:47,000
Speaker 1: nervousness all at once.

55
00:02:47,600 --> 00:02:50,560
Speaker 2: I'd love to go back to when I met you,

56
00:02:50,960 --> 00:02:54,280
Speaker 2: like in twenty ten. This is long before the release

57
00:02:54,320 --> 00:02:59,280
Speaker 2: of Chat GBT and AI's integration into the world. I

58
00:02:59,360 --> 00:03:01,520
Speaker 2: remember I was just a cub reporter and I was

59
00:03:01,560 --> 00:03:04,240
Speaker 2: obsessed with this thing called startups that people weren't really

60
00:03:04,320 --> 00:03:07,160
Speaker 2: talking too much about. It wasn't normal to get into

61
00:03:07,160 --> 00:03:10,440
Speaker 2: a stranger's car that's Uber, by the way, or sleep

62
00:03:10,480 --> 00:03:14,320
Speaker 2: in a stranger's home that's Airbnb, And it was just

63
00:03:14,360 --> 00:03:17,359
Speaker 2: this really interesting moment where the iPhone had come out,

64
00:03:17,440 --> 00:03:19,280
Speaker 2: the app store had launched, and you could have this

65
00:03:19,400 --> 00:03:22,280
Speaker 2: idea and you could code it into the hands of

66
00:03:22,280 --> 00:03:27,160
Speaker 2: millions of people. It didn't feel like tech feels today.

67
00:03:27,280 --> 00:03:30,200
Speaker 2: It felt anti establishment, it felt a little bit punk rock,

68
00:03:30,800 --> 00:03:34,040
Speaker 2: and so I just want to ground this. And I

69
00:03:34,200 --> 00:03:37,600
Speaker 2: met you at that moment, right, I met you on

70
00:03:37,680 --> 00:03:40,920
Speaker 2: a random bench to put you on camera for the

71
00:03:40,960 --> 00:03:44,600
Speaker 2: first time. You had a geolocation app called Looped, and

72
00:03:45,560 --> 00:03:48,400
Speaker 2: it was a moment in time. You didn't have a

73
00:03:48,440 --> 00:03:51,880
Speaker 2: massive company, didn't have pr people. It was just like,

74
00:03:52,000 --> 00:03:55,040
Speaker 2: you know, you, I assume like a handful of folks

75
00:03:55,040 --> 00:03:57,680
Speaker 2: building out this thing that you thought could be the future.

76
00:03:59,360 --> 00:04:02,960
Speaker 2: Tell me how you would describe yourself then and that

77
00:04:03,040 --> 00:04:05,520
Speaker 2: moment in time before we get to this moment now,

78
00:04:05,560 --> 00:04:08,280
Speaker 2: which is just so incredibly different.

79
00:04:08,840 --> 00:04:11,840
Speaker 1: The world had really changed. You could make mobile apps,

80
00:04:12,000 --> 00:04:16,200
Speaker 1: a whole new kind of startup potential had unleashed itself.

81
00:04:16,480 --> 00:04:20,240
Speaker 1: And it did feel kind of like punk rock is

82
00:04:20,240 --> 00:04:23,880
Speaker 1: a nice way of saying it, but like chaotic, unorganized, unprofessional,

83
00:04:24,080 --> 00:04:26,400
Speaker 1: like most of us didn't really know what we were doing.

84
00:04:27,560 --> 00:04:32,080
Speaker 1: And you know, now it's been more systematized, and startups

85
00:04:32,080 --> 00:04:34,039
Speaker 1: feel very different and they raise a ton of money

86
00:04:34,080 --> 00:04:36,320
Speaker 1: right away and they grow really fast, and there's like

87
00:04:36,320 --> 00:04:39,000
Speaker 1: a playbook that's been figured out. But it almost felt

88
00:04:39,040 --> 00:04:43,159
Speaker 1: like that was like the pre paradigm stage and it

89
00:04:43,279 --> 00:04:45,720
Speaker 1: felt really fun, kind of felt like we all didn't

90
00:04:45,720 --> 00:04:47,479
Speaker 1: know what you were doing, but clearly something was working.

91
00:04:48,480 --> 00:04:50,720
Speaker 1: So I look back on that phase with like a

92
00:04:50,839 --> 00:04:51,880
Speaker 1: very pleasant nostalgia.

93
00:04:52,360 --> 00:04:52,600
Speaker 2: Yeah.

94
00:04:52,600 --> 00:04:54,560
Speaker 1: Also the stakes felt solo. That was nice too.

95
00:04:54,680 --> 00:04:56,480
Speaker 2: I think that was something I was thinking about when

96
00:04:56,520 --> 00:04:59,200
Speaker 2: I was sitting here thinking about coming to meet you today.

97
00:04:59,320 --> 00:05:02,360
Speaker 2: Like viewed you at many points in your career, and

98
00:05:02,400 --> 00:05:06,479
Speaker 2: it just feels like, I mean, the stakes are so

99
00:05:06,680 --> 00:05:10,520
Speaker 2: high now you Let's go back to the last time

100
00:05:10,560 --> 00:05:13,360
Speaker 2: I interviewed you was twenty twenty. You're the CEO of

101
00:05:13,400 --> 00:05:15,200
Speaker 2: open Ai at the time, but it's before you guys

102
00:05:15,200 --> 00:05:19,560
Speaker 2: had launched chat GBT. You said something to me that

103
00:05:19,640 --> 00:05:23,240
Speaker 2: I think was really interesting to hear back now that

104
00:05:23,320 --> 00:05:24,799
Speaker 2: I want to play for you really quick.

105
00:05:25,200 --> 00:05:27,159
Speaker 1: I may be wrong, but what I actually believe is

106
00:05:27,200 --> 00:05:30,560
Speaker 1: we're going to create this technology that is more transformative

107
00:05:30,560 --> 00:05:33,760
Speaker 1: than any technology humans have ever created, and almost no

108
00:05:33,760 --> 00:05:36,200
Speaker 1: one is paying attention or talking about it. I really

109
00:05:36,240 --> 00:05:40,440
Speaker 1: genuinely believe that this is going to change the world

110
00:05:40,480 --> 00:05:42,680
Speaker 1: in unrecognizable ways. It's going to make what we talked

111
00:05:42,680 --> 00:05:45,000
Speaker 1: about happening earlier with the iPhone look like a warm up.

112
00:05:45,520 --> 00:05:47,120
Speaker 1: And we got to talk about that.

113
00:05:48,520 --> 00:05:50,560
Speaker 2: Okay, I think you were right.

114
00:05:51,120 --> 00:05:52,320
Speaker 1: I mean, yeah, it was pretty accurate.

115
00:05:52,360 --> 00:05:54,719
Speaker 2: I think that was pretty accurate. So let's fast forward.

116
00:05:54,800 --> 00:05:58,599
Speaker 1: The thing that's amazing is like, just how that's six

117
00:05:58,680 --> 00:06:02,320
Speaker 1: years That was six years ago. That's not very long. Yeah,

118
00:06:02,360 --> 00:06:04,040
Speaker 1: and like so, and at the time that I think

119
00:06:04,040 --> 00:06:06,120
Speaker 1: that probably sounded like a crazy thing to say that

120
00:06:06,240 --> 00:06:09,760
Speaker 1: was like pre GPT. I mean maybe we had GPD three.

121
00:06:09,760 --> 00:06:13,080
Speaker 1: It depends when in the year that was. But yeah,

122
00:06:13,160 --> 00:06:14,480
Speaker 1: a lot has happened very quickly.

123
00:06:14,960 --> 00:06:17,000
Speaker 2: Yeah, I mean, and I think that's what you talk

124
00:06:17,040 --> 00:06:19,719
Speaker 2: about with the rate of acceleration, and which is why

125
00:06:19,800 --> 00:06:22,159
Speaker 2: you talk about the stakes. We talk about the stakes

126
00:06:22,200 --> 00:06:25,159
Speaker 2: being so high. So like now it's twenty twenty six.

127
00:06:25,240 --> 00:06:28,000
Speaker 2: Open ais reportedly valued it like a trillion dollars yore

128
00:06:28,360 --> 00:06:32,000
Speaker 2: now for profit, no longer a nonprofit, and you're building

129
00:06:32,000 --> 00:06:35,159
Speaker 2: one of the most powerful technologies in human history, and

130
00:06:35,200 --> 00:06:39,719
Speaker 2: the company is incredibly valuable. So you talked a little

131
00:06:39,760 --> 00:06:42,960
Speaker 2: bit before about where we are and AI you said something.

132
00:06:43,839 --> 00:06:46,920
Speaker 2: You said that based off the current trajectory, it's possible

133
00:06:46,920 --> 00:06:49,960
Speaker 2: we're only two years away from a world with more

134
00:06:50,000 --> 00:06:53,119
Speaker 2: of the cognitive capacity in the world inside of data

135
00:06:53,160 --> 00:06:56,640
Speaker 2: centers than outside of them. I mean, tell me the

136
00:06:56,680 --> 00:06:57,560
Speaker 2: implications of that.

137
00:06:57,839 --> 00:06:59,560
Speaker 1: Yeah, it's easy to say, oh, we're going to have

138
00:06:59,600 --> 00:07:03,240
Speaker 1: super here, is super intelligence here, but that there. But

139
00:07:03,400 --> 00:07:06,400
Speaker 1: the reason I try to say things like that is

140
00:07:07,680 --> 00:07:09,320
Speaker 1: I think if you just say we're gonna have AGI soon,

141
00:07:09,400 --> 00:07:11,560
Speaker 1: people are like, Okay, I don't really know what that

142
00:07:11,600 --> 00:07:13,760
Speaker 1: means how to think about it. If you say, like,

143
00:07:14,320 --> 00:07:16,239
Speaker 1: the AI is going to do more of the total

144
00:07:16,720 --> 00:07:20,400
Speaker 1: thinking than people are, that kind of like gets across

145
00:07:20,680 --> 00:07:24,280
Speaker 1: the magnitude of what's happening in a way that to me,

146
00:07:24,400 --> 00:07:27,400
Speaker 1: at least, I feel differently than just saying like AGI

147
00:07:27,520 --> 00:07:31,040
Speaker 1: is coming soon. In what sense, I'm not like a

148
00:07:31,080 --> 00:07:34,040
Speaker 1: long term jobs doomer. I think we'll figure out new

149
00:07:34,080 --> 00:07:38,200
Speaker 1: things to do. I think we always do. But in

150
00:07:38,240 --> 00:07:41,440
Speaker 1: the short term, I think it may. Again, we don't

151
00:07:41,480 --> 00:07:43,040
Speaker 1: know it could work in all these different ways. I

152
00:07:43,040 --> 00:07:46,760
Speaker 1: think it may and likely will be quite disruptive to

153
00:07:46,800 --> 00:07:48,680
Speaker 1: a lot of the current jobs. If more of the

154
00:07:49,040 --> 00:07:52,400
Speaker 1: intellectual horsepower in the world is GPUs, and I think like,

155
00:07:52,720 --> 00:07:56,240
Speaker 1: right now, we've got to start talking about the principles

156
00:07:56,240 --> 00:07:58,560
Speaker 1: for how we want to design an economy that works

157
00:07:58,600 --> 00:08:00,800
Speaker 1: for everybody. If that's going to happen. Now, this tremend

158
00:08:00,920 --> 00:08:03,720
Speaker 1: is upside too. If you have that world, you can

159
00:08:03,760 --> 00:08:06,880
Speaker 1: also you could imagine curing diseases extremely quickly. The coolest

160
00:08:06,880 --> 00:08:09,160
Speaker 1: meeting I had this week was a guy who was

161
00:08:09,200 --> 00:08:13,440
Speaker 1: not an expert that used chatchipet to make a custom

162
00:08:13,520 --> 00:08:17,840
Speaker 1: mr NA vaccine for his dog's cancer. WHOA, and it's

163
00:08:17,840 --> 00:08:18,440
Speaker 1: just incredible.

164
00:08:18,440 --> 00:08:21,440
Speaker 2: It was crazy, and he's meetings like that every single week.

165
00:08:22,080 --> 00:08:24,160
Speaker 1: For whatever reason, this one hit me in just hearing

166
00:08:24,240 --> 00:08:26,080
Speaker 1: him tell the story. He flew in from Australia and

167
00:08:27,320 --> 00:08:30,240
Speaker 1: like what he went through for this dog and what

168
00:08:30,280 --> 00:08:33,600
Speaker 1: he was able to do, like not just figuring out

169
00:08:33,640 --> 00:08:36,920
Speaker 1: how to make the vaccine, but developing the relationships with

170
00:08:37,000 --> 00:08:39,000
Speaker 1: the professors that had to like do the work for

171
00:08:39,080 --> 00:08:41,199
Speaker 1: him once he told them, like here's the sequence and

172
00:08:41,240 --> 00:08:44,360
Speaker 1: here's what to do, and this the way he was

173
00:08:44,440 --> 00:08:48,680
Speaker 1: able to use chatgipt full stack to like do what

174
00:08:48,760 --> 00:08:52,800
Speaker 1: would have taken a full research institute and save his dog.

175
00:08:53,320 --> 00:08:54,600
Speaker 1: And now he's trying to figure out to do this

176
00:08:54,600 --> 00:08:59,800
Speaker 1: for other people's dogs. It's so incredible. So they're you know, abundant,

177
00:09:00,240 --> 00:09:03,600
Speaker 1: you can do all of these things. Assuming we get

178
00:09:03,880 --> 00:09:07,559
Speaker 1: the new version of the world right in a way

179
00:09:07,559 --> 00:09:09,320
Speaker 1: that this works for everybody and it's not just like

180
00:09:09,760 --> 00:09:11,840
Speaker 1: a small number people have the access and the resources

181
00:09:11,840 --> 00:09:12,160
Speaker 1: to do this.

182
00:09:12,440 --> 00:09:14,040
Speaker 2: Well, you know what. I also hear where I'm like, Wow,

183
00:09:14,080 --> 00:09:17,040
Speaker 2: the guy cured his dog's cancer. That's amazing. And then

184
00:09:17,040 --> 00:09:19,320
Speaker 2: I'm like, wait if he could do that, Like, couldn't

185
00:09:19,320 --> 00:09:22,199
Speaker 2: someone just build a pathogen? Right? Like what about building

186
00:09:22,240 --> 00:09:24,880
Speaker 2: out viruses? So it's like as always, it's you know,

187
00:09:25,000 --> 00:09:28,319
Speaker 2: I imagine that's something you guys think a lot of,

188
00:09:28,400 --> 00:09:30,800
Speaker 2: like a lot about when it comes especially these open

189
00:09:30,840 --> 00:09:31,760
Speaker 2: source models.

190
00:09:31,920 --> 00:09:35,840
Speaker 1: Yeah, are thinking around safety has evolved quite a bit,

191
00:09:36,200 --> 00:09:39,480
Speaker 1: given the way that ecosystem has developed. We still think

192
00:09:39,520 --> 00:09:41,280
Speaker 1: everything we used to think is important. You know, you

193
00:09:41,320 --> 00:09:44,400
Speaker 1: still have to build. You have to align these models

194
00:09:45,200 --> 00:09:46,520
Speaker 1: to get them to behave in a way you want.

195
00:09:46,559 --> 00:09:49,640
Speaker 1: You have to build many levels of safety defense. But

196
00:09:49,720 --> 00:09:51,719
Speaker 1: as you said, there's this new thing going on now,

197
00:09:51,720 --> 00:09:57,240
Speaker 1: which is the world has to be resilient to lots

198
00:09:57,240 --> 00:10:00,160
Speaker 1: of day I from lots of different people, and not

199
00:10:00,240 --> 00:10:02,400
Speaker 1: all that will have the same safety restrictions. So if

200
00:10:02,400 --> 00:10:05,160
Speaker 1: someone does build that bad pathogen, we should do everything

201
00:10:05,160 --> 00:10:07,840
Speaker 1: we can to prevent them from building that or releasing it.

202
00:10:08,080 --> 00:10:10,839
Speaker 1: But we also need the ability to defend against it,

203
00:10:11,200 --> 00:10:14,480
Speaker 1: have rapid response treatments and vaccines, have the ability to

204
00:10:14,520 --> 00:10:17,520
Speaker 1: detect pandemics earlier. This is a particular area that I'm

205
00:10:17,520 --> 00:10:21,960
Speaker 1: nervous about and the reason we've started talking about this

206
00:10:22,000 --> 00:10:25,080
Speaker 1: whole concept today a resilience at least in my own head.

207
00:10:25,080 --> 00:10:28,440
Speaker 1: The thing that first got me thinking that we needed

208
00:10:28,440 --> 00:10:31,520
Speaker 1: to evolve our approach was this pandemic point in.

209
00:10:31,480 --> 00:10:35,120
Speaker 2: Particular, when you say resilience, like, what does that mean in.

210
00:10:35,040 --> 00:10:39,560
Speaker 1: Practice, like the ability for society to rapidly defend against.

211
00:10:39,200 --> 00:10:40,959
Speaker 2: New threats and how do we do that?

212
00:10:41,520 --> 00:10:43,600
Speaker 1: Well, it depends on what the threat is. But for

213
00:10:43,640 --> 00:10:46,560
Speaker 1: this example, it's not enough to just say, like, hey,

214
00:10:46,600 --> 00:10:49,640
Speaker 1: three frontier labs in the world, don't let any of

215
00:10:49,640 --> 00:10:52,600
Speaker 1: your models be used to develop a new pathogen. In

216
00:10:52,640 --> 00:10:54,960
Speaker 1: a few more years, when some open source model is

217
00:10:55,000 --> 00:10:57,840
Speaker 1: capable of it and a terrorist group or someone else

218
00:10:57,880 --> 00:11:02,000
Speaker 1: does that you doing today? Frontier labs that have an

219
00:11:02,000 --> 00:11:05,280
Speaker 1: advantage to help help us be ready for that moment.

220
00:11:05,440 --> 00:11:08,160
Speaker 2: Right right, And there can be like a collective effort.

221
00:11:08,880 --> 00:11:11,240
Speaker 1: I think it has to be different levels too, you

222
00:11:11,280 --> 00:11:15,960
Speaker 1: know mentioned vaccines, earlier response. I honestly I hoped out

223
00:11:15,960 --> 00:11:19,720
Speaker 1: of COVID the world was going to be more ready

224
00:11:19,720 --> 00:11:22,480
Speaker 1: for this. And I won't say we learn nothing, but

225
00:11:22,760 --> 00:11:24,400
Speaker 1: we did not make the progress I was hoping.

226
00:11:24,840 --> 00:11:28,040
Speaker 2: I want to talk about open AI today and what

227
00:11:28,080 --> 00:11:29,360
Speaker 2: your biggest priorities are.

228
00:11:30,640 --> 00:11:34,080
Speaker 1: So Codex has really been amazing for us all to

229
00:11:34,160 --> 00:11:37,360
Speaker 1: watch and use, and it is the first time that

230
00:11:38,920 --> 00:11:41,920
Speaker 1: I feel like I have gotten to use the future

231
00:11:42,160 --> 00:11:46,640
Speaker 1: since CHATCHBT launched. In what sense, I'll say the positive

232
00:11:46,640 --> 00:11:51,000
Speaker 1: sense and the negative one. The positive sense is I

233
00:11:51,080 --> 00:11:54,760
Speaker 1: any idea I can come up with any piece of

234
00:11:54,760 --> 00:11:58,960
Speaker 1: software I want, I can have built, you know by

235
00:11:59,000 --> 00:12:02,520
Speaker 1: the time I wake up the next morning. And I

236
00:12:02,600 --> 00:12:05,320
Speaker 1: use this for fun stuff, for serious stuff. I'm like

237
00:12:05,440 --> 00:12:08,800
Speaker 1: very interested in what it means to mostly automate or

238
00:12:08,840 --> 00:12:11,959
Speaker 1: heavily augment my job. And you know, it's rather than

239
00:12:12,040 --> 00:12:13,840
Speaker 1: like ask someone else to build stuff for me. The

240
00:12:13,840 --> 00:12:16,840
Speaker 1: fact that I can build these tools myself is amazing.

241
00:12:17,280 --> 00:12:18,920
Speaker 1: That's like, Wow, this is what it's going to be

242
00:12:19,080 --> 00:12:20,839
Speaker 1: like when people get to use a huge amount of

243
00:12:20,880 --> 00:12:23,920
Speaker 1: compute and anything they want they can create, Like this

244
00:12:23,960 --> 00:12:26,640
Speaker 1: is the sci fi future. This is awesome. And then

245
00:12:26,640 --> 00:12:29,160
Speaker 1: the negative I had with Codex was I had this

246
00:12:29,280 --> 00:12:31,640
Speaker 1: list of side projects that I wanted to build and

247
00:12:31,720 --> 00:12:33,959
Speaker 1: for the last you know, few years open and it's

248
00:12:33,960 --> 00:12:36,480
Speaker 1: been so busy, things only ever got added to it.

249
00:12:36,960 --> 00:12:38,760
Speaker 1: I never did any of them, I never built any

250
00:12:38,800 --> 00:12:42,480
Speaker 1: of them. And I finished the list. I ran out

251
00:12:42,480 --> 00:12:45,160
Speaker 1: of ideas. I ran out of side project ideas. And

252
00:12:45,200 --> 00:12:47,079
Speaker 1: the models are not yet good enough, or at least

253
00:12:47,080 --> 00:12:48,319
Speaker 1: I had figured how to use them to help me

254
00:12:48,360 --> 00:12:49,360
Speaker 1: come up with more ideas.

255
00:12:50,080 --> 00:12:52,199
Speaker 2: Well, you could use your brain, yeah, but.

256
00:12:52,200 --> 00:12:53,880
Speaker 1: I had like used my brain for this list. And

257
00:12:53,960 --> 00:12:55,920
Speaker 1: so then I had this like weird thing one Friday

258
00:12:55,960 --> 00:12:58,560
Speaker 1: a few weekends ago where I was like, I'm going

259
00:12:58,640 --> 00:13:01,079
Speaker 1: to go to bed tonight without Codex because I don't

260
00:13:01,080 --> 00:13:02,480
Speaker 1: have like another idea of thing to build. That was

261
00:13:02,520 --> 00:13:05,079
Speaker 1: like a strange feeling. But but it's really amazing what

262
00:13:05,320 --> 00:13:08,800
Speaker 1: is happening with how good Codex has gotten. I mean,

263
00:13:09,000 --> 00:13:11,440
Speaker 1: the like we just watched the usage of this expand

264
00:13:11,480 --> 00:13:14,959
Speaker 1: and with the next model, I'm sure it'll explode upwards again.

265
00:13:16,520 --> 00:13:20,600
Speaker 1: The fact that people can now build such sophisticated things.

266
00:13:21,200 --> 00:13:23,679
Speaker 1: It's totally changed our workflow. It's changed the workflow of

267
00:13:23,760 --> 00:13:26,840
Speaker 1: many companies. It's really changed what individuals can do and

268
00:13:26,920 --> 00:13:29,920
Speaker 1: you know, we've talked about these one person startups that

269
00:13:29,960 --> 00:13:32,600
Speaker 1: may feel like the punk rock era again that's going

270
00:13:32,640 --> 00:13:36,800
Speaker 1: to be fun to explore. And then on the super upfront, historically,

271
00:13:36,800 --> 00:13:41,040
Speaker 1: we've talked about two major kind of priorities that we're

272
00:13:41,040 --> 00:13:45,760
Speaker 1: going to work towards. One is the automated researcher, where

273
00:13:45,760 --> 00:13:48,680
Speaker 1: we have systems that can do scientific research, including AI research,

274
00:13:48,800 --> 00:13:53,440
Speaker 1: on their own. And the other, which touches on this

275
00:13:53,600 --> 00:13:57,240
Speaker 1: kind of startup thing, is you know, automated companies, like

276
00:13:57,240 --> 00:13:59,720
Speaker 1: not literally fully automated companies, but where you can sow

277
00:14:00,800 --> 00:14:04,160
Speaker 1: accelerate a company because the AI can do not just

278
00:14:04,240 --> 00:14:08,280
Speaker 1: coding work, but huge amounts of what it takes to

279
00:14:08,360 --> 00:14:10,560
Speaker 1: run and operate a company. The world can have much

280
00:14:10,600 --> 00:14:12,880
Speaker 1: more stuff and we can create you know, way better

281
00:14:12,920 --> 00:14:15,240
Speaker 1: products and services, and it'll be accessible to way more

282
00:14:15,240 --> 00:14:18,000
Speaker 1: people to do that. But then a thing that we

283
00:14:18,040 --> 00:14:20,240
Speaker 1: didn't talk as much about, not because we weren't excited

284
00:14:20,280 --> 00:14:22,680
Speaker 1: about it, we just weren't sure how I was going

285
00:14:22,760 --> 00:14:27,240
Speaker 1: to manifest itself was when you take that level of

286
00:14:27,280 --> 00:14:29,880
Speaker 1: capability and give it to an individual and just say,

287
00:14:30,240 --> 00:14:36,440
Speaker 1: you know, help make my life better. Open claw really

288
00:14:36,480 --> 00:14:40,480
Speaker 1: I think showed us a path there and the I

289
00:14:40,520 --> 00:14:43,880
Speaker 1: hate the phrase super app, but the combined personal agent

290
00:14:43,920 --> 00:14:46,280
Speaker 1: that we're building. The dream that I have is I

291
00:14:46,320 --> 00:14:48,360
Speaker 1: get to a point where there's like enough robustness and

292
00:14:48,400 --> 00:14:50,560
Speaker 1: trust I have in the system where I just say, hey,

293
00:14:51,040 --> 00:14:53,480
Speaker 1: go look around my computer, go use the web for me,

294
00:14:53,680 --> 00:14:58,320
Speaker 1: go read my messages, and you know, you can just

295
00:14:58,560 --> 00:15:01,240
Speaker 1: you can sort of like listen to my meetings and

296
00:15:01,440 --> 00:15:05,040
Speaker 1: you know my just intermediate my interactions for me and

297
00:15:05,120 --> 00:15:07,240
Speaker 1: just start doing useful stuff to me. I don't have

298
00:15:07,280 --> 00:15:09,360
Speaker 1: to think, I don't have to ask you questions, but

299
00:15:09,400 --> 00:15:11,920
Speaker 1: I've run out of stuff on my side project list. However,

300
00:15:11,960 --> 00:15:14,600
Speaker 1: you can know everything about my life, start suggesting more

301
00:15:14,640 --> 00:15:18,040
Speaker 1: things I should build, and then like help me do them.

302
00:15:18,160 --> 00:15:19,520
Speaker 1: So I'm really excited for that.

303
00:15:19,600 --> 00:15:23,200
Speaker 2: It's interesting. You know, there is this concept that floats

304
00:15:23,240 --> 00:15:26,760
Speaker 2: around of could the next billion dollar company be created

305
00:15:26,760 --> 00:15:28,920
Speaker 2: with like a solo entrepreneur and AI agents.

306
00:15:29,000 --> 00:15:32,760
Speaker 1: Do you believe that has happened? I promise I would

307
00:15:32,760 --> 00:15:35,080
Speaker 1: not share details until he's ready to announce it, But

308
00:15:35,080 --> 00:15:35,960
Speaker 1: I believe that has happened.

309
00:15:36,000 --> 00:15:39,480
Speaker 2: Can you give us any details around the details.

310
00:15:40,920 --> 00:15:43,760
Speaker 1: It is a legitimate, single person billion dollar company as

311
00:15:43,800 --> 00:15:46,520
Speaker 1: far as I can tell I'm not like reviewed the financials.

312
00:15:46,880 --> 00:15:48,720
Speaker 1: I mean, but I think it's just happened.

313
00:15:50,080 --> 00:15:51,880
Speaker 2: I think folks listening to that and think, well, this

314
00:15:51,960 --> 00:15:57,240
Speaker 2: is AI can be so unapproachable. But the idea that

315
00:15:57,240 --> 00:15:59,320
Speaker 2: that we are entering a world where you don't have

316
00:15:59,440 --> 00:16:03,440
Speaker 2: to have Silicon Valley contacts. You don't have to be

317
00:16:03,440 --> 00:16:05,880
Speaker 2: able to be in this club to build out a

318
00:16:05,920 --> 00:16:10,680
Speaker 2: company and have evaluation. And I think open Claw is

319
00:16:10,680 --> 00:16:12,800
Speaker 2: a really interesting example for folks who don't know what

320
00:16:12,840 --> 00:16:15,800
Speaker 2: open Claw is. It's kind of this like AI agent

321
00:16:15,880 --> 00:16:19,200
Speaker 2: that just does anything, which of course had lots of

322
00:16:19,280 --> 00:16:21,480
Speaker 2: safety issues. I would say, like, if you're going to

323
00:16:21,480 --> 00:16:24,240
Speaker 2: give it access to everything, this is the most Silicon

324
00:16:24,320 --> 00:16:26,640
Speaker 2: Valley thing. I just see everybody doing this and I'm like,

325
00:16:27,360 --> 00:16:29,280
Speaker 2: what could go wrong? Like, you do have to be

326
00:16:29,480 --> 00:16:32,680
Speaker 2: very careful about the security of this, but you the

327
00:16:32,880 --> 00:16:36,240
Speaker 2: founder ended up joining the company, and so I don't

328
00:16:36,240 --> 00:16:37,560
Speaker 2: know if this was like an acquahi or whatever you

329
00:16:37,600 --> 00:16:42,400
Speaker 2: want to call it. But he built this with AI agents,

330
00:16:42,680 --> 00:16:43,840
Speaker 2: and I imagine.

331
00:16:43,400 --> 00:16:45,200
Speaker 1: He was like one of the top users of Codex

332
00:16:45,200 --> 00:16:47,320
Speaker 1: of all time, so he built this whole thing with

333
00:16:47,400 --> 00:16:50,520
Speaker 1: Codex and it was just like unbelievably productive in a

334
00:16:50,520 --> 00:16:51,960
Speaker 1: way that no single person could have been.

335
00:16:52,200 --> 00:16:54,120
Speaker 2: Is he one of the early examples you think of

336
00:16:54,560 --> 00:16:58,560
Speaker 2: someone who was able to exit cell or you have

337
00:16:58,800 --> 00:17:00,160
Speaker 2: a company of.

338
00:17:00,120 --> 00:17:01,560
Speaker 1: A successful company with one person.

339
00:17:03,160 --> 00:17:04,360
Speaker 2: Does that surprise you at all?

340
00:17:05,200 --> 00:17:05,360
Speaker 3: No?

341
00:17:05,400 --> 00:17:09,399
Speaker 1: And I think it's awesome. I mean the individual empowerment

342
00:17:09,520 --> 00:17:11,200
Speaker 1: that happens, as you were just saying, when you don't

343
00:17:11,240 --> 00:17:12,760
Speaker 1: have to be part of the Silicon Valley network, you

344
00:17:12,800 --> 00:17:14,200
Speaker 1: don't have to know all these people you can hire

345
00:17:14,240 --> 00:17:16,760
Speaker 1: and these vcs that you can convince to give you money,

346
00:17:16,760 --> 00:17:19,720
Speaker 1: you can just like just have AI build it for you.

347
00:17:20,600 --> 00:17:22,040
Speaker 1: I think that's awesome. And I think we're seeing the

348
00:17:22,080 --> 00:17:24,359
Speaker 1: same thing now in science and a lot of other

349
00:17:24,400 --> 00:17:25,919
Speaker 1: fields too. It's not just startups.

350
00:17:26,280 --> 00:17:29,080
Speaker 2: Yeah, I'd love to talk about as we're talking about

351
00:17:29,119 --> 00:17:31,280
Speaker 2: the company and the focus of it, it's what are

352
00:17:31,320 --> 00:17:33,159
Speaker 2: you doing and what are you not doing? And so

353
00:17:33,440 --> 00:17:36,280
Speaker 2: apparently you're not doing Sora anymore. You guys have made

354
00:17:36,280 --> 00:17:38,600
Speaker 2: the decision to shut it down. So take me to

355
00:17:38,640 --> 00:17:42,240
Speaker 2: the room. Sam. It's like three months ago, Opening Eye

356
00:17:42,240 --> 00:17:44,400
Speaker 2: signs a deal with Disney. This is like a Landmark

357
00:17:44,480 --> 00:17:47,760
Speaker 2: deal and in so many ways, Disney's going to license

358
00:17:47,760 --> 00:17:51,639
Speaker 2: two hundred characters, Pixar, Marvels, Star Wars. Everybody has an

359
00:17:51,640 --> 00:17:55,720
Speaker 2: opinion on this. Bob Eiger, who led the deal, he said, this,

360
00:17:55,840 --> 00:17:58,879
Speaker 2: not me. We're bringing together Disney's iconic stories and characters

361
00:17:58,880 --> 00:18:01,720
Speaker 2: with open Ay's ground technology. They're going to invest a

362
00:18:01,720 --> 00:18:06,439
Speaker 2: billion dollars. Fast forward three months later, open Ai shuts

363
00:18:06,480 --> 00:18:09,080
Speaker 2: down Sora. What's the note under the note here?

364
00:18:09,680 --> 00:18:12,480
Speaker 1: We have a few times in our history realized something

365
00:18:12,720 --> 00:18:15,639
Speaker 1: really important is working are about to work so well

366
00:18:16,040 --> 00:18:19,320
Speaker 1: that we have to stop a bunch of other projects.

367
00:18:19,760 --> 00:18:22,800
Speaker 1: In fact, this was the original thing happening with GPT three.

368
00:18:23,520 --> 00:18:25,400
Speaker 1: We had a whole portfolio of bets at the time.

369
00:18:26,080 --> 00:18:28,160
Speaker 1: A lot of them were working well. We shut down

370
00:18:28,200 --> 00:18:30,840
Speaker 1: many projects that we're working well, like robotics, which we mentioned,

371
00:18:31,160 --> 00:18:34,280
Speaker 1: so that we could concentrate our compute, our researchers, our

372
00:18:34,320 --> 00:18:36,879
Speaker 1: effort into this thing that we said, Okay, there's a

373
00:18:37,000 --> 00:18:41,600
Speaker 1: very important thing happening. I did not expect three or

374
00:18:41,680 --> 00:18:43,440
Speaker 1: six months ago to be at this point. We're at

375
00:18:43,480 --> 00:18:46,199
Speaker 1: now where something very big and important is about to

376
00:18:46,200 --> 00:18:49,760
Speaker 1: happen again with this next generation of models and the

377
00:18:49,800 --> 00:18:55,119
Speaker 1: agents they they can power. But I love soa I

378
00:18:55,160 --> 00:18:58,000
Speaker 1: love generated videos. And I love our partnership with Disney

379
00:18:58,080 --> 00:19:02,160
Speaker 1: and we're working hard with them to find a world

380
00:19:02,720 --> 00:19:04,919
Speaker 1: where they can still do something amazing and we can

381
00:19:04,960 --> 00:19:11,600
Speaker 1: help with that. But we need to concentrate our compute

382
00:19:11,760 --> 00:19:18,120
Speaker 1: and our product capacity into these next generation of automated

383
00:19:18,160 --> 00:19:19,280
Speaker 1: researchers and companies.

384
00:19:19,320 --> 00:19:21,800
Speaker 2: And the note under the note is it's about compute.

385
00:19:22,400 --> 00:19:23,560
Speaker 1: It's always about compute.

386
00:19:24,119 --> 00:19:26,040
Speaker 2: God to be a fly on the wall for that conversation,

387
00:19:26,119 --> 00:19:29,080
Speaker 2: did you call Bob Eiger and tell him personally?

388
00:19:30,240 --> 00:19:30,600
Speaker 1: Of course?

389
00:19:30,720 --> 00:19:31,600
Speaker 2: Yeah, how was that?

390
00:19:32,720 --> 00:19:36,879
Speaker 1: Disney is an amazing company all around, and the like

391
00:19:36,920 --> 00:19:40,639
Speaker 1: the very first thing that the new Disney CEO Josh

392
00:19:40,640 --> 00:19:43,080
Speaker 1: said to me, and I felt like terrible and I was, okay,

393
00:19:43,320 --> 00:19:49,919
Speaker 1: you know, it's like I get it, but it's super

394
00:19:50,440 --> 00:19:54,679
Speaker 1: sad always to disappoint a partner or users or a

395
00:19:54,680 --> 00:19:59,920
Speaker 1: team all of what you're doing incredible work and the

396
00:20:00,000 --> 00:20:01,880
Speaker 1: I mean, there are like many hard parts about being

397
00:20:01,880 --> 00:20:04,119
Speaker 1: a CEO that you don't get sympathy for. Understand that lead.

398
00:20:04,160 --> 00:20:06,680
Speaker 1: But one of them is like you have to make

399
00:20:06,720 --> 00:20:09,840
Speaker 1: a lot of very tough resourcing calls and a lot

400
00:20:09,840 --> 00:20:14,680
Speaker 1: of good things get caught up in that because they're

401
00:20:14,720 --> 00:20:28,119
Speaker 1: not the most important thing.

402
00:20:28,920 --> 00:20:32,000
Speaker 2: I want to go back to February twenty seventh, the

403
00:20:32,040 --> 00:20:35,920
Speaker 2: Trump administration had to deal with Anthropic. They're renegotiating the deal.

404
00:20:35,920 --> 00:20:38,679
Speaker 2: Anthropics CEO said he's not going to allow cloud to

405
00:20:38,720 --> 00:20:41,639
Speaker 2: be used for autonomous weapons or to surveil American citizens.

406
00:20:42,080 --> 00:20:44,199
Speaker 2: There was a big public back and forth. People who

407
00:20:44,240 --> 00:20:46,840
Speaker 2: didn't even really know about AI were seeing Anthropic on

408
00:20:46,880 --> 00:20:49,520
Speaker 2: the front pages of the paper. They were given a

409
00:20:49,560 --> 00:20:53,480
Speaker 2: five pm deadline to comply. They didn't agree. The Trump

410
00:20:53,520 --> 00:20:57,200
Speaker 2: administration accused them of being traders, labeled them the tech

411
00:20:57,240 --> 00:21:00,520
Speaker 2: a supply chain risk, and ordered federal agent to stop

412
00:21:00,600 --> 00:21:04,280
Speaker 2: using Anthropics. So hours later, OpenAI announces a deal with

413
00:21:04,320 --> 00:21:06,600
Speaker 2: the Department of War. Was that a mistake.

414
00:21:08,760 --> 00:21:12,160
Speaker 1: I'll separate what we did and why from the way

415
00:21:12,200 --> 00:21:15,480
Speaker 1: that we rolled it out. We had decided not to

416
00:21:15,560 --> 00:21:18,720
Speaker 1: work with the government on classified networks. It's not that

417
00:21:18,760 --> 00:21:21,400
Speaker 1: it's not important. It's just that Anthropic had gone hard

418
00:21:21,400 --> 00:21:25,080
Speaker 1: in that direction, and I think they're, you know, reasonably

419
00:21:25,119 --> 00:21:29,080
Speaker 1: safety minded, and we're doing good work there, and we

420
00:21:29,080 --> 00:21:31,800
Speaker 1: were busy with other things. I do think it's very

421
00:21:31,840 --> 00:21:35,040
Speaker 1: important that our government and our military in particular have

422
00:21:35,240 --> 00:21:39,159
Speaker 1: access to advanced AI models. I do not think that

423
00:21:39,359 --> 00:21:42,240
Speaker 1: they will be able to carry out their mission for

424
00:21:42,400 --> 00:21:48,000
Speaker 1: national security without that. I also think that we talked

425
00:21:48,000 --> 00:21:51,280
Speaker 1: about kind of the changes happening to society earlier. One

426
00:21:51,359 --> 00:21:53,280
Speaker 1: of the most important questions the world will have to

427
00:21:53,280 --> 00:21:56,400
Speaker 1: answer in the next year is our AI companies or

428
00:21:56,520 --> 00:22:00,800
Speaker 1: our governments more powerful? And I think it's very important

429
00:22:00,840 --> 00:22:04,439
Speaker 1: that the governments are more powerful, like the future of

430
00:22:06,240 --> 00:22:11,119
Speaker 1: the world, and the decisions about the most important elements

431
00:22:11,160 --> 00:22:15,240
Speaker 1: of national security should be made through a democratically elected process,

432
00:22:15,240 --> 00:22:16,960
Speaker 1: and the people that have been appointed as part of

433
00:22:16,960 --> 00:22:19,159
Speaker 1: that process not me and not the CEO of some

434
00:22:19,200 --> 00:22:23,520
Speaker 1: other lab. However, there are areas where the law hasn't

435
00:22:23,520 --> 00:22:26,120
Speaker 1: caught up yet. This is a new technology and it's

436
00:22:26,119 --> 00:22:28,320
Speaker 1: going to take some time to legislate it, and during

437
00:22:28,320 --> 00:22:30,600
Speaker 1: that time period, I think it is reasonable for the

438
00:22:30,600 --> 00:22:33,400
Speaker 1: companies to say, hey, we understand something about this technology.

439
00:22:33,480 --> 00:22:36,320
Speaker 1: It is not ready yet for autonomous weapons. The rules

440
00:22:36,320 --> 00:22:39,760
Speaker 1: about protecting against domestic surveillance were well thought through for

441
00:22:39,800 --> 00:22:43,600
Speaker 1: a world without super powerful AI and are not yet

442
00:22:44,240 --> 00:22:48,040
Speaker 1: adapted for that. So in areas like that, I think

443
00:22:48,080 --> 00:22:50,439
Speaker 1: it's reasonable to say, hey, we just need to go

444
00:22:50,560 --> 00:22:52,320
Speaker 1: slower on this, which we did, like that was part

445
00:22:52,359 --> 00:22:55,720
Speaker 1: of our contract as it's written, that is part of

446
00:22:55,760 --> 00:22:59,520
Speaker 1: our safety stack that those are we share the same principles.

447
00:23:00,040 --> 00:23:04,520
Speaker 1: Thropic is right on those principles, and as I've heard,

448
00:23:04,560 --> 00:23:06,600
Speaker 1: they like very nearly got a deal done like ours

449
00:23:06,600 --> 00:23:09,119
Speaker 1: with those principles. In my dream world, they would have

450
00:23:09,160 --> 00:23:12,159
Speaker 1: just kept working with the government. I don't think it

451
00:23:12,359 --> 00:23:16,040
Speaker 1: works for our industry to say, hey, this is the

452
00:23:16,040 --> 00:23:19,120
Speaker 1: most powerful technology humanity has ever built. It is going

453
00:23:19,160 --> 00:23:21,960
Speaker 1: to be the high order bit in geopolitics. It is

454
00:23:22,040 --> 00:23:24,040
Speaker 1: going to be the greatest cyber weapon the world has

455
00:23:24,040 --> 00:23:27,320
Speaker 1: ever built. It is going to you know, be the

456
00:23:27,359 --> 00:23:31,359
Speaker 1: determinant of future wars and protection. And we are not

457
00:23:31,400 --> 00:23:32,800
Speaker 1: giving it to you. We are not going to tell

458
00:23:32,800 --> 00:23:34,240
Speaker 1: you have it. You know, we're going to make a decision.

459
00:23:34,280 --> 00:23:37,120
Speaker 1: We're going to make the decisions here. If I were

460
00:23:37,119 --> 00:23:40,240
Speaker 1: the government, or if I were you know, a citizen

461
00:23:40,359 --> 00:23:42,280
Speaker 1: voting for representatives of my government, I would say, like, that's

462
00:23:42,320 --> 00:23:45,480
Speaker 1: really not okay. And I thought, and I still think

463
00:23:45,640 --> 00:23:50,200
Speaker 1: that had we not gotten involved with the government, things

464
00:23:50,240 --> 00:23:53,040
Speaker 1: could have gone fairly off the rails pretty quickly. Where

465
00:23:53,040 --> 00:23:54,800
Speaker 1: the government was saying, like you know, they were making

466
00:23:54,880 --> 00:23:58,040
Speaker 1: DPA threats. They were making some stronger statements than that.

467
00:23:58,520 --> 00:24:00,639
Speaker 1: So we need to say, hey, like going to leave you.

468
00:24:01,320 --> 00:24:03,920
Speaker 1: We the industry can't leave the US government without something here.

469
00:24:04,000 --> 00:24:05,800
Speaker 1: We will as long as we can have something that

470
00:24:05,840 --> 00:24:08,560
Speaker 1: respects our red lines. I wish we'd announced it very differently.

471
00:24:08,600 --> 00:24:11,880
Speaker 1: I think, like my goal was to lower the temperature.

472
00:24:11,880 --> 00:24:14,119
Speaker 1: It clearly had the opposite effect. I think I like

473
00:24:14,200 --> 00:24:17,959
Speaker 1: learned something about how people feel about the government, how

474
00:24:18,040 --> 00:24:19,919
Speaker 1: much people want to tell like a competitive story in

475
00:24:19,920 --> 00:24:22,359
Speaker 1: the industry. So I'm sure we'll do better next time.

476
00:24:23,440 --> 00:24:25,959
Speaker 1: But on the merits of like, the government needs support

477
00:24:26,000 --> 00:24:28,400
Speaker 1: from US companies to carry out the mission and national security,

478
00:24:28,440 --> 00:24:29,560
Speaker 1: and we need to answer that call.

479
00:24:29,720 --> 00:24:33,600
Speaker 2: I feel good, isn't it? Also? The devil's in the details, right,

480
00:24:33,720 --> 00:24:36,120
Speaker 2: you know, if we look at how people feel, we're

481
00:24:36,160 --> 00:24:41,080
Speaker 2: in a contentious time. You know, there's increased domestic scrutiny

482
00:24:41,119 --> 00:24:44,040
Speaker 2: over ice, the war in Iran. So what do you

483
00:24:44,080 --> 00:24:47,679
Speaker 2: say to folks who say, okay, like your red line,

484
00:24:47,680 --> 00:24:52,760
Speaker 2: but you sam like, what exactly like is your red line?

485
00:24:52,760 --> 00:24:55,960
Speaker 2: And how will you knowing that things can go awright,

486
00:24:56,160 --> 00:24:59,000
Speaker 2: knowing that these technologies can be used in these ways

487
00:25:00,320 --> 00:25:02,560
Speaker 2: that you might not be able to control. What is

488
00:25:02,640 --> 00:25:06,879
Speaker 2: your process for knowing when we're too close to the cliff?

489
00:25:06,920 --> 00:25:07,480
Speaker 1: Yeah?

490
00:25:07,520 --> 00:25:10,639
Speaker 2: And also like, can you enforce your red lines?

491
00:25:10,760 --> 00:25:12,840
Speaker 1: Yeah? I mean one question that people say is if

492
00:25:12,840 --> 00:25:16,120
Speaker 1: the government's nationalized the AI companies, can you enforce any

493
00:25:16,160 --> 00:25:18,479
Speaker 1: red lines at all? And I'll be honest at at

494
00:25:18,480 --> 00:25:22,119
Speaker 1: that point, probably we you know, like it's you can

495
00:25:22,160 --> 00:25:24,480
Speaker 1: imagine a world of that happening where no, we couldn't

496
00:25:24,520 --> 00:25:27,520
Speaker 1: anymore as an independent company. And certainly in our interaction

497
00:25:27,560 --> 00:25:31,240
Speaker 1: with the government, they have been really great about saying, hey,

498
00:25:31,320 --> 00:25:36,040
Speaker 1: you know, we understand these issues. You will do it contractually.

499
00:25:36,040 --> 00:25:39,520
Speaker 1: You can build systems there. Now. Again, I don't know,

500
00:25:39,600 --> 00:25:41,520
Speaker 1: nor does anybody else know. How weird the next few

501
00:25:41,560 --> 00:25:44,040
Speaker 1: years are going to get. I hope the governments don't

502
00:25:44,240 --> 00:25:45,679
Speaker 1: nationalize the AI labs.

503
00:25:46,640 --> 00:25:48,359
Speaker 2: You don't think that's a possibility, though, do you.

504
00:25:49,640 --> 00:25:52,400
Speaker 1: I would don't say it's not a possibility. I don't

505
00:25:52,440 --> 00:25:53,000
Speaker 1: think it's.

506
00:25:53,000 --> 00:25:54,359
Speaker 2: Have you had conversations about that.

507
00:25:54,520 --> 00:25:57,000
Speaker 1: I don't think it's likely, but I like I will

508
00:25:57,000 --> 00:25:59,639
Speaker 1: say that if I want, if we want that not

509
00:25:59,680 --> 00:26:02,399
Speaker 1: to happen, and then I think we better find a

510
00:26:02,400 --> 00:26:04,600
Speaker 1: way to work with the government if we're not doing it,

511
00:26:04,600 --> 00:26:08,840
Speaker 1: it would seem more likely. The I've said before, and

512
00:26:08,880 --> 00:26:11,600
Speaker 1: I still really believe that in a well run society,

513
00:26:12,000 --> 00:26:16,000
Speaker 1: developing AI would have been a government project. The government

514
00:26:16,040 --> 00:26:19,080
Speaker 1: is the government project. The Apollo program was a government project.

515
00:26:19,119 --> 00:26:22,199
Speaker 1: Even the Eisenhower Highway System as a government project. But

516
00:26:22,480 --> 00:26:24,399
Speaker 1: it doesn't feel like we're in a time like that anymore.

517
00:26:24,400 --> 00:26:26,840
Speaker 1: It doesn't feel like the government could effectively do this,

518
00:26:27,359 --> 00:26:28,600
Speaker 1: and I think it does need to happen.

519
00:26:29,080 --> 00:26:32,040
Speaker 2: You know, there's this assumptions that federal agencies operate within

520
00:26:32,080 --> 00:26:35,040
Speaker 2: the law, but history shows that it can be fluid.

521
00:26:35,080 --> 00:26:39,439
Speaker 2: That surveillance programs exposed by Edward Snowden Essentially they were

522
00:26:39,480 --> 00:26:42,439
Speaker 2: authorized by secret courts and considered legal at the time,

523
00:26:42,480 --> 00:26:45,479
Speaker 2: but they were later ruled unlawful in reform once they

524
00:26:45,520 --> 00:26:47,840
Speaker 2: became public. And I think that's why I go back

525
00:26:47,880 --> 00:26:50,240
Speaker 2: to your red lines of a lot of these things

526
00:26:50,280 --> 00:26:52,680
Speaker 2: sound great in theory, but when we look at the

527
00:26:52,760 --> 00:26:56,080
Speaker 2: current reality of the world that we're living in, how

528
00:26:56,160 --> 00:26:59,159
Speaker 2: do you define these red lines and what happens if

529
00:26:59,200 --> 00:26:59,919
Speaker 2: you can't pull them that?

530
00:27:00,280 --> 00:27:02,560
Speaker 1: So two things there. One that was also one of

531
00:27:02,560 --> 00:27:05,159
Speaker 1: my biggest takeaways is there are a lot of people.

532
00:27:05,400 --> 00:27:08,480
Speaker 1: I don't know if it's a representative sample or just

533
00:27:08,520 --> 00:27:10,919
Speaker 1: a lot of people online, but there is there's at

534
00:27:11,000 --> 00:27:13,120
Speaker 1: least a group of a lot of people online who

535
00:27:13,240 --> 00:27:16,320
Speaker 1: really don't trust the government to follow the law, and

536
00:27:16,359 --> 00:27:19,400
Speaker 1: that feels like a very bad sign for our democracy.

537
00:27:19,760 --> 00:27:22,080
Speaker 1: I mostly do. I mean, I realized they're not perfect

538
00:27:22,119 --> 00:27:24,159
Speaker 1: and some things are going to get screwed up, and

539
00:27:24,200 --> 00:27:25,880
Speaker 1: I think we have a system of checks and balances,

540
00:27:26,280 --> 00:27:28,840
Speaker 1: but I mostly trust it. As you said, this is

541
00:27:28,920 --> 00:27:31,080
Speaker 1: like a time where people are looking at things happening

542
00:27:31,119 --> 00:27:33,800
Speaker 1: and saying, I'm just, you know, don't feel bad about it.

543
00:27:33,840 --> 00:27:36,119
Speaker 1: And maybe this is I don't feel good about it,

544
00:27:36,400 --> 00:27:38,520
Speaker 1: and maybe this is such a uniquely bad time that

545
00:27:38,720 --> 00:27:42,760
Speaker 1: like companies, patriotic duty is not to support, not to

546
00:27:42,760 --> 00:27:45,639
Speaker 1: work with the government. I totally disagree on that. I

547
00:27:45,640 --> 00:27:48,680
Speaker 1: think again, given what I see on the horizon, if

548
00:27:48,680 --> 00:27:51,120
Speaker 1: we don't help the government with national security and it's

549
00:27:51,160 --> 00:27:53,080
Speaker 1: not just wars and traditional sense, if we don't help

550
00:27:53,119 --> 00:27:56,680
Speaker 1: them with you know, defending this cyber infrastructure of the US,

551
00:27:56,680 --> 00:27:58,199
Speaker 1: if we don't help them with the biodefense we were

552
00:27:58,200 --> 00:28:00,600
Speaker 1: talking about earlier, I think it's really I think we

553
00:28:00,680 --> 00:28:05,000
Speaker 1: have to work with the government, but the intensity of

554
00:28:05,040 --> 00:28:09,040
Speaker 1: the current mood of mistrust I was miscalibrated on. And

555
00:28:09,080 --> 00:28:10,360
Speaker 1: I understand something there. Now.

556
00:28:10,840 --> 00:28:15,159
Speaker 2: A federal judge just ordered a preliminary injunction against the

557
00:28:15,200 --> 00:28:19,600
Speaker 2: Pentagon's actions on Anthropic, calling the administration's actions what appears

558
00:28:19,640 --> 00:28:21,879
Speaker 2: to be a First Amendment retaliation. So I'm curious for

559
00:28:21,920 --> 00:28:22,760
Speaker 2: your response on that.

560
00:28:22,920 --> 00:28:27,440
Speaker 1: We said all the way through that we thought, like publicly, privately, loudly,

561
00:28:27,640 --> 00:28:30,719
Speaker 1: that we thought the government doing anything like an scr

562
00:28:30,800 --> 00:28:32,880
Speaker 1: or threat of the DPA against Anthropic was really bad.

563
00:28:32,880 --> 00:28:35,720
Speaker 1: We were trying to help provide an off ramp there.

564
00:28:36,840 --> 00:28:38,120
Speaker 1: I think that's like a very bad thing.

565
00:28:38,560 --> 00:28:41,520
Speaker 2: What would be your message to the folks at Anthropic

566
00:28:41,560 --> 00:28:42,600
Speaker 2: and the government on it.

567
00:28:42,960 --> 00:28:45,200
Speaker 1: Stop the stuff on boat, stop the escalation on both sides,

568
00:28:45,200 --> 00:28:46,200
Speaker 1: and find a way to work there.

569
00:28:46,600 --> 00:28:49,520
Speaker 2: I was thinking a lot about our last conversation in

570
00:28:49,600 --> 00:28:52,000
Speaker 2: twenty twenty, and I was struck by a story you

571
00:28:52,000 --> 00:28:53,880
Speaker 2: told me that I you know, it's a kind of

572
00:28:53,880 --> 00:28:56,120
Speaker 2: a personal story, but I think it's go for it.

573
00:28:56,240 --> 00:28:57,640
Speaker 2: You talked about how you were one of the only

574
00:28:57,680 --> 00:29:01,480
Speaker 2: openly gay students your high school growing up in Missouri,

575
00:29:01,720 --> 00:29:05,600
Speaker 2: and how you essentially gave a speech in front of

576
00:29:05,840 --> 00:29:10,880
Speaker 2: oh yeah, in front of an assembly to talk about it,

577
00:29:10,920 --> 00:29:12,960
Speaker 2: to put it out there so other people would start

578
00:29:13,000 --> 00:29:15,280
Speaker 2: talking about it, whether they liked you or not you said,

579
00:29:16,040 --> 00:29:19,120
Speaker 2: or whether they you know, they felt this way or not.

580
00:29:19,400 --> 00:29:22,200
Speaker 2: And I think there is something interesting about this moment

581
00:29:22,360 --> 00:29:26,280
Speaker 2: and the anger and the and the interest, because I

582
00:29:26,280 --> 00:29:29,120
Speaker 2: feel like you're under the bright lights again and people

583
00:29:29,200 --> 00:29:33,160
Speaker 2: are asking you what you stand for, and honestly, if

584
00:29:33,200 --> 00:29:36,000
Speaker 2: I'm being totally honest with someone who's known you for

585
00:29:36,040 --> 00:29:38,760
Speaker 2: fifteen years, I want to know the same thing.

586
00:29:39,120 --> 00:29:39,360
Speaker 4: You know.

587
00:29:39,440 --> 00:29:42,240
Speaker 2: This is a moment where things aren't normal, all these

588
00:29:42,240 --> 00:29:44,760
Speaker 2: things that tech folks say of like Okay, we can

589
00:29:44,800 --> 00:29:47,640
Speaker 2: do this and this, but we're this playbook is different. Yeah,

590
00:29:47,640 --> 00:29:50,000
Speaker 2: this is different, and so what do you stand for?

591
00:29:50,880 --> 00:29:53,720
Speaker 1: I think this technology has to be democratized. I can

592
00:29:53,800 --> 00:29:57,200
Speaker 1: see a world where this is either something that is

593
00:29:57,360 --> 00:30:01,320
Speaker 1: concentrated from a perspective of wealth, but also power over

594
00:30:01,360 --> 00:30:04,640
Speaker 1: the future in the hands of a small number of

595
00:30:04,680 --> 00:30:07,320
Speaker 1: companies who want to make the decisions, and no matter

596
00:30:07,360 --> 00:30:11,240
Speaker 1: how much they care about all of us, the idea

597
00:30:11,280 --> 00:30:14,200
Speaker 1: that we would like trade off our agency and our

598
00:30:14,240 --> 00:30:18,400
Speaker 1: collective will over the future to a few companies doesn't

599
00:30:18,400 --> 00:30:21,400
Speaker 1: sit right with me. I think we need to empower

600
00:30:21,440 --> 00:30:25,720
Speaker 1: people with the technology and have our governments decide how

601
00:30:26,440 --> 00:30:29,000
Speaker 1: society is going to evolve, what the economic system looks like,

602
00:30:29,200 --> 00:30:32,320
Speaker 1: and people need to have this technology. And I think

603
00:30:32,360 --> 00:30:33,960
Speaker 1: one of the most important things that we ever came

604
00:30:34,040 --> 00:30:36,560
Speaker 1: up with was this idea of iterative deployment. We're going

605
00:30:36,600 --> 00:30:38,479
Speaker 1: to put the technology out in the world early and

606
00:30:38,560 --> 00:30:40,680
Speaker 1: often and let people figure out what works for them

607
00:30:40,720 --> 00:30:42,280
Speaker 1: and what doesn't and get a feel for it, and

608
00:30:42,320 --> 00:30:46,680
Speaker 1: society to decide what should happen. Now. The criticism of

609
00:30:46,720 --> 00:30:48,680
Speaker 1: that I'd give is we started doing that a little

610
00:30:48,720 --> 00:30:51,000
Speaker 1: over three years ago. Society has not made a lot

611
00:30:51,080 --> 00:30:52,560
Speaker 1: in the way of the decisions I was hoping for,

612
00:30:54,440 --> 00:30:56,400
Speaker 1: but people at least do have a sense for what's coming.

613
00:30:56,720 --> 00:31:00,120
Speaker 1: So that's that theme of democratization is a big one.

614
00:31:02,080 --> 00:31:06,440
Speaker 1: Empowerment I think goes closely along with that. But we

615
00:31:06,520 --> 00:31:08,520
Speaker 1: talked about a bunch of examples where people have been

616
00:31:08,800 --> 00:31:12,320
Speaker 1: really empowered with this technology. Yes, we should all collectively

617
00:31:12,320 --> 00:31:14,920
Speaker 1: decide what the guardrails should be, and we should have

618
00:31:14,960 --> 00:31:17,360
Speaker 1: democratic governments have a lot of power. You should strengthen

619
00:31:17,400 --> 00:31:19,720
Speaker 1: the maxu where we can. We should also, and I

620
00:31:19,720 --> 00:31:22,600
Speaker 1: think this is like a deeply held American value that

621
00:31:22,640 --> 00:31:25,960
Speaker 1: has worked very well. We should really empower individual people,

622
00:31:26,480 --> 00:31:28,800
Speaker 1: you know, like we should set some broad rules and

623
00:31:28,800 --> 00:31:31,680
Speaker 1: then let people go figure out how to cure cancer

624
00:31:31,840 --> 00:31:33,680
Speaker 1: for their dog, or figure out how to make a

625
00:31:33,720 --> 00:31:36,080
Speaker 1: one person startup, or figure out how to set up

626
00:31:36,120 --> 00:31:39,200
Speaker 1: open Claw to like automate something amazing in their life.

627
00:31:39,200 --> 00:31:43,520
Speaker 1: And this is uncomfortable for people because individuals are going

628
00:31:43,600 --> 00:31:46,280
Speaker 1: to have so much power. But I think the democratization

629
00:31:46,320 --> 00:31:52,400
Speaker 1: and empowerment pieces are linked importantly. Abundance is something that

630
00:31:52,440 --> 00:31:55,280
Speaker 1: I care a lot about. If I look at my

631
00:31:55,360 --> 00:31:57,240
Speaker 1: career and the stuff I've been interested in and kind

632
00:31:57,280 --> 00:31:59,600
Speaker 1: of what I believe about what it takes to have

633
00:31:59,640 --> 00:32:01,800
Speaker 1: a better world, and I could pick only one word,

634
00:32:01,840 --> 00:32:05,000
Speaker 1: I pick abundance. I think you want to just create

635
00:32:05,360 --> 00:32:10,040
Speaker 1: huge amounts of resources, opportunity, the ability for people to

636
00:32:10,080 --> 00:32:14,640
Speaker 1: have like an amazing life and to sort of express

637
00:32:14,680 --> 00:32:18,320
Speaker 1: it in all kinds of in all kinds of different ways.

638
00:32:18,600 --> 00:32:21,400
Speaker 1: And that's AI. It's energy, It'll be things like robots

639
00:32:21,400 --> 00:32:23,360
Speaker 1: in the future. It's certainly making sure that we build

640
00:32:24,000 --> 00:32:28,440
Speaker 1: enough compute to make all of this happen, but that's

641
00:32:28,480 --> 00:32:31,720
Speaker 1: a place that I really want to drive towards. Safety

642
00:32:31,920 --> 00:32:35,040
Speaker 1: has evolved or been added to. But that's another thing

643
00:32:35,040 --> 00:32:39,240
Speaker 1: that I like. The trade off of all of this

644
00:32:39,320 --> 00:32:43,480
Speaker 1: individual empowerment is that one person could do a great

645
00:32:43,520 --> 00:32:47,160
Speaker 1: deal of harm, and so building up guardrails in the

646
00:32:47,160 --> 00:32:49,960
Speaker 1: world to prevent that and where we can feel okay

647
00:32:50,000 --> 00:32:55,400
Speaker 1: about this is critical. I named a bunch of principles.

648
00:32:55,400 --> 00:32:57,360
Speaker 1: There's more I'd like to name, you know, like fairly

649
00:32:57,440 --> 00:33:00,040
Speaker 1: sharing all of the upside and the voice and the

650
00:33:00,120 --> 00:33:03,520
Speaker 1: agency and the decision making of this technology. Is that

651
00:33:03,560 --> 00:33:07,040
Speaker 1: one seems like one that would be very difficult to

652
00:33:07,040 --> 00:33:10,320
Speaker 1: ever change my mind on. Yeah, But other things, like

653
00:33:10,840 --> 00:33:14,360
Speaker 1: maybe the trade off between safety and agency. There'll be

654
00:33:14,400 --> 00:33:16,520
Speaker 1: periods where we have to say, ah, this technology went

655
00:33:16,560 --> 00:33:18,320
Speaker 1: in a different way than we thought. We really love

656
00:33:18,400 --> 00:33:20,600
Speaker 1: empowering people, but not at the risk of destroying the

657
00:33:20,600 --> 00:33:22,160
Speaker 1: whole world. So we're going to have to do something

658
00:33:22,160 --> 00:33:25,000
Speaker 1: differently for a while. So we will wrestle with these

659
00:33:25,000 --> 00:33:27,880
Speaker 1: principles all the time, and we, I would say, with

660
00:33:27,920 --> 00:33:31,400
Speaker 1: some certainty, will have to make changes as the technology progresses.

661
00:33:31,840 --> 00:33:33,560
Speaker 1: And I don't think people like hearing that, but it is

662
00:33:33,600 --> 00:33:34,080
Speaker 1: the truth.

663
00:33:34,880 --> 00:33:39,560
Speaker 2: Man named Blasio runs the mc Government Institute. He asked

664
00:33:39,600 --> 00:33:42,080
Speaker 2: a question, and I think it's an important question, because

665
00:33:42,200 --> 00:33:44,680
Speaker 2: to be a founder now is really different than it

666
00:33:44,800 --> 00:33:47,640
Speaker 2: was back when we started talking talking about this, when

667
00:33:47,640 --> 00:33:51,000
Speaker 2: it felt a little more anti establishment and chaotic. Now

668
00:33:51,400 --> 00:33:53,680
Speaker 2: there's kind of like a different playbook. And so he

669
00:33:53,720 --> 00:33:55,239
Speaker 2: had a question about how you navigate it.

670
00:33:55,320 --> 00:33:58,960
Speaker 3: Okay, many of us became founders or builders because we

671
00:33:59,000 --> 00:34:02,480
Speaker 3: wanted to make something useful. Used to be about a

672
00:34:02,480 --> 00:34:06,520
Speaker 3: fanatic focus on the product, building something great and getting

673
00:34:06,600 --> 00:34:10,040
Speaker 3: it into people's hands. But the job of a founder

674
00:34:10,040 --> 00:34:13,439
Speaker 3: seems to have fundamentally changed over the last few years. Now,

675
00:34:13,560 --> 00:34:16,759
Speaker 3: especially in AI, being a founder might mean having to

676
00:34:16,960 --> 00:34:20,080
Speaker 3: raise sovereign capital, or stand next to heads of state

677
00:34:20,239 --> 00:34:24,520
Speaker 3: at a public meeting, or endorse political statements. It might

678
00:34:24,640 --> 00:34:29,600
Speaker 3: mean making decisions that reshape labor markets and affect billions

679
00:34:29,600 --> 00:34:32,799
Speaker 3: of lives. They're the decisions that used to be the

680
00:34:32,800 --> 00:34:35,759
Speaker 3: domain of governments and required a very different kind of

681
00:34:35,800 --> 00:34:41,160
Speaker 3: preparation than building a product. So Sam, as a quintessential founder,

682
00:34:41,200 --> 00:34:43,600
Speaker 3: I'm curious, is this what it means to be a

683
00:34:43,640 --> 00:34:47,319
Speaker 3: founder now? And if so, who are founders accountable to?

684
00:34:47,400 --> 00:34:50,520
Speaker 3: When they're making those kinds of decisions, how are we

685
00:34:50,760 --> 00:34:53,000
Speaker 3: preparing them to hold that responsibility.

686
00:34:53,880 --> 00:34:56,360
Speaker 1: I don't think that's what it means for most founders.

687
00:34:56,440 --> 00:34:58,279
Speaker 1: I think, you know, if you're one of these one

688
00:34:58,320 --> 00:35:00,719
Speaker 1: person or small team companies build in with AI. It

689
00:35:01,320 --> 00:35:03,719
Speaker 1: feels the closest to that twenty ten period we were

690
00:35:03,719 --> 00:35:09,160
Speaker 1: talking about that it's ever felt since then. My own experience, unfortunately,

691
00:35:09,160 --> 00:35:12,759
Speaker 1: does feel more like a politician in what sense I

692
00:35:12,800 --> 00:35:14,360
Speaker 1: have to do those things. I have to go like

693
00:35:14,440 --> 00:35:16,840
Speaker 1: negotiate deals with governments and fly around and you know,

694
00:35:17,000 --> 00:35:21,239
Speaker 1: talked to war leaders and work on these deals for

695
00:35:21,920 --> 00:35:24,080
Speaker 1: land and power for data center expansion that are kind

696
00:35:24,080 --> 00:35:27,399
Speaker 1: of the kind of deals governments used to do. It's

697
00:35:27,440 --> 00:35:32,160
Speaker 1: not my natural environment, say more. It was what you

698
00:35:32,200 --> 00:35:34,520
Speaker 1: were saying earlier, Like I like my whole life has

699
00:35:34,560 --> 00:35:37,840
Speaker 1: been around kind of the garage version of the founders.

700
00:35:39,040 --> 00:35:40,759
Speaker 1: I've won a lot more suits in the last couple

701
00:35:40,800 --> 00:35:42,120
Speaker 1: of years, and I've worn in my whole life put

702
00:35:42,120 --> 00:35:42,720
Speaker 1: together before.

703
00:35:42,960 --> 00:35:45,720
Speaker 2: And it also feels like the stakes are much higher.

704
00:35:45,719 --> 00:35:47,759
Speaker 2: You're in very different rooms than you used to be.

705
00:35:48,200 --> 00:35:50,640
Speaker 2: What have you learned about that? Becaus it been an

706
00:35:50,640 --> 00:35:54,440
Speaker 2: easy process to kind of to lean into them.

707
00:35:54,760 --> 00:35:59,080
Speaker 1: No, no, no, not at all. This sounds obvious, But

708
00:35:59,560 --> 00:36:04,040
Speaker 1: the difference in skills and temperament and worldview between the

709
00:36:04,040 --> 00:36:06,680
Speaker 1: average politician and the average founder, I kind of knew

710
00:36:06,719 --> 00:36:08,839
Speaker 1: they were crazy different. They're crazier different than I could

711
00:36:08,880 --> 00:36:11,600
Speaker 1: have ever imagined a lesson. I feel like I've learned

712
00:36:11,640 --> 00:36:15,720
Speaker 1: repeatedly in life, at higher and higher levels of power

713
00:36:15,719 --> 00:36:20,160
Speaker 1: and status and whatever is that everyone thinks. I always

714
00:36:20,200 --> 00:36:22,239
Speaker 1: have thought, there's like some adult in the room. You

715
00:36:22,320 --> 00:36:24,399
Speaker 1: always get to some final boss. There's someone with a plan,

716
00:36:24,440 --> 00:36:27,040
Speaker 1: there's someone who knows exactly what they're doing. And the

717
00:36:27,120 --> 00:36:30,279
Speaker 1: leaders of the world are also you know, uncertain and

718
00:36:30,320 --> 00:36:33,040
Speaker 1: insecure and trying their best but don't have the answers,

719
00:36:33,080 --> 00:36:35,600
Speaker 1: And like maybe this, maybe that kind of going to

720
00:36:35,600 --> 00:36:36,399
Speaker 1: make a call while I'm.

721
00:36:36,320 --> 00:36:49,719
Speaker 2: Tired either way, doing kind of a hard pivot to

722
00:36:49,760 --> 00:36:51,600
Speaker 2: something I think we both care about, but I think

723
00:36:51,719 --> 00:36:54,399
Speaker 2: is probably like the most high stakes thing I could

724
00:36:54,440 --> 00:36:58,200
Speaker 2: think of is being a parent in this era of

725
00:36:58,360 --> 00:37:03,279
Speaker 2: AI both raising young boys, and I think I said

726
00:37:03,280 --> 00:37:05,440
Speaker 2: this to you before, like we're both raising young boys,

727
00:37:05,440 --> 00:37:07,640
Speaker 2: but in a sense you're raising my son too. Like

728
00:37:07,760 --> 00:37:12,120
Speaker 2: the technology that you build will be integrated into every

729
00:37:12,239 --> 00:37:14,480
Speaker 2: facet of my son, Charlie.

730
00:37:14,480 --> 00:37:17,040
Speaker 1: I hope I'll let them use it, Yet I am.

731
00:37:17,080 --> 00:37:19,120
Speaker 2: Absolutely not letting him use it. When are you letting

732
00:37:19,120 --> 00:37:19,920
Speaker 2: your son use it?

733
00:37:20,160 --> 00:37:20,800
Speaker 1: Not for a while?

734
00:37:21,040 --> 00:37:21,800
Speaker 2: Do you have an idea?

735
00:37:22,520 --> 00:37:24,560
Speaker 1: You know? People ask me all the time, like, oh,

736
00:37:24,680 --> 00:37:26,239
Speaker 1: now that you have a kid, are going to have

737
00:37:26,239 --> 00:37:29,480
Speaker 1: more kids? Like do you feel more responsibility about how

738
00:37:29,480 --> 00:37:31,240
Speaker 1: you don't destroy the world with Ai?

739
00:37:31,880 --> 00:37:32,920
Speaker 2: Do you?

740
00:37:32,960 --> 00:37:35,480
Speaker 1: No? Like, my highest order bit is to not destroy

741
00:37:35,520 --> 00:37:37,440
Speaker 1: the world with Ai, Like it doesn't matter how good

742
00:37:37,480 --> 00:37:40,600
Speaker 1: everything else is if we do that. So I knew

743
00:37:40,640 --> 00:37:42,759
Speaker 1: I was going to have kids, and I knew that

744
00:37:43,480 --> 00:37:45,399
Speaker 1: I thought a lot about what I would feel about

745
00:37:45,440 --> 00:37:47,200
Speaker 1: the world and my future kids would grow up in.

746
00:37:47,239 --> 00:37:51,879
Speaker 1: And you know that I used to write my kid

747
00:37:51,880 --> 00:37:54,680
Speaker 1: a letter every night. The lawyers eventually convinced me to stop,

748
00:37:54,880 --> 00:37:56,680
Speaker 1: but it was a nice thing to do about just

749
00:37:56,960 --> 00:37:58,560
Speaker 1: because when I would put him to bed, I would

750
00:37:58,560 --> 00:38:00,640
Speaker 1: just sort of get home late, like for his bedtime,

751
00:38:00,640 --> 00:38:02,799
Speaker 1: and I'd, you know, i'd be rocking him, and he

752
00:38:03,000 --> 00:38:04,440
Speaker 1: needs something to talk to, you kid about. So I would

753
00:38:04,400 --> 00:38:05,799
Speaker 1: tell them about my day and what was going on.

754
00:38:05,880 --> 00:38:08,200
Speaker 1: And you know, obviously a five month old four month

755
00:38:08,200 --> 00:38:09,960
Speaker 1: old baby's not going to understand these things. So I

756
00:38:10,000 --> 00:38:11,239
Speaker 1: was like, you know, i'll write them down. I'll give

757
00:38:11,239 --> 00:38:12,440
Speaker 1: them to him later because it was sort of an

758
00:38:12,440 --> 00:38:16,959
Speaker 1: interesting time. But an interesting thing about writing those letters

759
00:38:17,000 --> 00:38:20,680
Speaker 1: would be like you can't really hide anything, like you

760
00:38:20,719 --> 00:38:22,480
Speaker 1: will be the most honest version of yourself and the

761
00:38:22,520 --> 00:38:25,080
Speaker 1: decisions you're making and the thing you write to your kid.

762
00:38:25,320 --> 00:38:27,200
Speaker 1: So I really loved doing it. It was like my every

763
00:38:27,200 --> 00:38:29,960
Speaker 1: Sunday night ritual about like here was the hard stuff

764
00:38:29,960 --> 00:38:31,840
Speaker 1: that came this week, and here's why I decided and

765
00:38:31,880 --> 00:38:34,000
Speaker 1: why and you know, here's what I'm worried about. But

766
00:38:34,040 --> 00:38:35,879
Speaker 1: the like the mindset of writing that to your kid

767
00:38:35,920 --> 00:38:38,200
Speaker 1: to read, and yeah, you know, fifteen or twenty years

768
00:38:38,440 --> 00:38:41,359
Speaker 1: very interesting, but no on the like not to story

769
00:38:41,400 --> 00:38:44,080
Speaker 1: the world that I always very said on that. A

770
00:38:44,120 --> 00:38:47,360
Speaker 1: thing that has changed hugely is how I feel about

771
00:38:49,160 --> 00:38:52,920
Speaker 1: algorithmic feeds and iPads in small children's hands and stuff

772
00:38:52,960 --> 00:38:57,560
Speaker 1: like that. And when I watch kids just a little

773
00:38:57,560 --> 00:39:00,440
Speaker 1: bit older than mine that you can not take the

774
00:39:00,440 --> 00:39:04,520
Speaker 1: iPad away from watching the sort of like short you know,

775
00:39:04,600 --> 00:39:08,759
Speaker 1: watching like whatever they're watching. That I feel very strongly about.

776
00:39:08,800 --> 00:39:11,479
Speaker 1: So I don't know when I would let him talk

777
00:39:11,520 --> 00:39:14,160
Speaker 1: to AI, but I'd rather be on the late end

778
00:39:14,200 --> 00:39:16,200
Speaker 1: of what's reasonable. They're not the early end. Of course.

779
00:39:16,239 --> 00:39:18,560
Speaker 1: I think it's like great, and he'll grow up in

780
00:39:18,600 --> 00:39:20,759
Speaker 1: a world where computers are smarter than him and do

781
00:39:20,800 --> 00:39:24,440
Speaker 1: anything he wants. But you know, I wanted to like

782
00:39:24,480 --> 00:39:25,480
Speaker 1: play in the dirt for now.

783
00:39:25,880 --> 00:39:28,400
Speaker 2: Well, I mean, what's interesting is you talk about these

784
00:39:28,440 --> 00:39:30,360
Speaker 2: algorithmic girls. I think a lot about this is a

785
00:39:30,400 --> 00:39:33,080
Speaker 2: mom too, Like, oh, these tech guys are just going

786
00:39:33,160 --> 00:39:35,760
Speaker 2: to create all the stuff and then send their children

787
00:39:35,800 --> 00:39:38,759
Speaker 2: to school with like no iPads, right, And so I

788
00:39:38,760 --> 00:39:42,520
Speaker 2: think about like I think about AI and how powerful

789
00:39:42,600 --> 00:39:45,319
Speaker 2: this product is. I mean, it is highly personalized, it's

790
00:39:45,360 --> 00:39:48,279
Speaker 2: always on, it's always there. You have people who are

791
00:39:48,280 --> 00:39:52,080
Speaker 2: becoming increasingly delusional by talking to AI all day. We've

792
00:39:52,080 --> 00:39:55,000
Speaker 2: had issues of AI psychosis, people ending their lives because

793
00:39:55,000 --> 00:39:59,080
Speaker 2: of AI companions. I'm curious, Like, you don't want AI

794
00:39:59,120 --> 00:40:03,000
Speaker 2: to become the next social media. That's terrifying, but we're

795
00:40:03,000 --> 00:40:05,560
Speaker 2: in a world where that could be a possibility without

796
00:40:05,560 --> 00:40:06,920
Speaker 2: the correct guardrails.

797
00:40:07,680 --> 00:40:10,560
Speaker 1: Just as we learn more about the relationship people are

798
00:40:10,600 --> 00:40:12,680
Speaker 1: going to have with AI, we should make it easy

799
00:40:12,719 --> 00:40:14,239
Speaker 1: for them to succeed, which make it easier for them

800
00:40:14,239 --> 00:40:16,320
Speaker 1: to have the healthy relationship and then give adults a

801
00:40:16,360 --> 00:40:19,319
Speaker 1: lot of freedom. But kids teenagers, I think they will

802
00:40:19,360 --> 00:40:21,600
Speaker 1: need a lot of guardrails around AI. This is this

803
00:40:21,680 --> 00:40:25,279
Speaker 1: is such a powerful and such a new thing. You

804
00:40:25,320 --> 00:40:27,200
Speaker 1: mentioned like sending how we're going to send our kids

805
00:40:27,200 --> 00:40:30,839
Speaker 1: to school. There is a version of the future here

806
00:40:30,880 --> 00:40:33,320
Speaker 1: that I'm very excited about. These new schools are popping

807
00:40:33,400 --> 00:40:35,400
Speaker 1: up where you have, you know, like maybe a couple

808
00:40:35,400 --> 00:40:37,560
Speaker 1: of hours a day of intense personalized one on one

809
00:40:37,560 --> 00:40:40,600
Speaker 1: tutoring with AI and then you kind of explore project

810
00:40:40,600 --> 00:40:42,520
Speaker 1: based in the school sets up whatever you want to

811
00:40:42,600 --> 00:40:45,000
Speaker 1: work on. That seems great. That's the kind of thing

812
00:40:45,000 --> 00:40:49,759
Speaker 1: where I'm like, I can totally imagine that, But you

813
00:40:49,760 --> 00:40:51,479
Speaker 1: can also mention a lot of worlds where it goes wrong.

814
00:40:51,760 --> 00:40:54,759
Speaker 2: And what world do you see AI being a very

815
00:40:55,400 --> 00:40:58,000
Speaker 2: a huge value add to the way our children grow up.

816
00:40:58,239 --> 00:40:59,640
Speaker 1: So I think there's like two stories you can tell

817
00:40:59,640 --> 00:41:02,200
Speaker 1: about A and education right now. One is you can

818
00:41:02,239 --> 00:41:04,719
Speaker 1: say everybody's using it to cheat, they're not really using

819
00:41:04,760 --> 00:41:06,839
Speaker 1: to learn their And you know, schools are making their

820
00:41:06,920 --> 00:41:08,960
Speaker 1: kids like write essays by hand because they have not

821
00:41:09,000 --> 00:41:12,600
Speaker 1: been able to figure out how to teach a curriculum

822
00:41:12,640 --> 00:41:15,840
Speaker 1: where people can use AI. And the other is you

823
00:41:15,880 --> 00:41:18,840
Speaker 1: can say, well, the world has been through transitions like

824
00:41:18,880 --> 00:41:20,880
Speaker 1: this before. We've freaked out about calculators, We've freaked out

825
00:41:20,880 --> 00:41:24,120
Speaker 1: about Google, we freaked out about computers. And it takes

826
00:41:24,160 --> 00:41:26,319
Speaker 1: a little while. It's only been three years, but we

827
00:41:26,400 --> 00:41:29,520
Speaker 1: do figure out how to teach people to think and

828
00:41:29,600 --> 00:41:32,040
Speaker 1: achieve at a higher level and to work with the

829
00:41:32,040 --> 00:41:34,279
Speaker 1: tools that they will have as an adult in society.

830
00:41:34,880 --> 00:41:38,440
Speaker 1: And you know, I definitely meet kids and teachers who

831
00:41:38,480 --> 00:41:41,960
Speaker 1: are like, Okay, fine, I can't teach people. I can't

832
00:41:42,000 --> 00:41:44,839
Speaker 1: like evaluate an essay as well anymore. And I think

833
00:41:44,840 --> 00:41:46,200
Speaker 1: the kids aren't learning to write in the same way,

834
00:41:46,280 --> 00:41:48,719
Speaker 1: but probably adults won't write essays in the same way

835
00:41:48,760 --> 00:41:50,920
Speaker 1: in the future either. And what really matters is do

836
00:41:50,960 --> 00:41:54,320
Speaker 1: you teach the students to think and create and figure out,

837
00:41:54,440 --> 00:41:56,880
Speaker 1: you know, stretch their brain and come up with like

838
00:41:56,920 --> 00:41:59,279
Speaker 1: explore ideas. And then I see what some people are

839
00:41:59,280 --> 00:42:03,160
Speaker 1: doing where like students are building entire new complex pieces

840
00:42:03,200 --> 00:42:06,680
Speaker 1: of software and worlds, and you know, figuring out very

841
00:42:06,680 --> 00:42:10,040
Speaker 1: difficult new ideas, much more difficult to figure out than

842
00:42:10,040 --> 00:42:13,000
Speaker 1: anything I had to figure out in school, and I

843
00:42:13,080 --> 00:42:15,520
Speaker 1: assume my kids when they finish high school will be

844
00:42:15,600 --> 00:42:19,640
Speaker 1: like wildly smarter and more capable and have like developed

845
00:42:19,640 --> 00:42:20,960
Speaker 1: their brain much more than I did.

846
00:42:21,400 --> 00:42:24,120
Speaker 2: I can't stop thinking as a mom about this idea

847
00:42:24,239 --> 00:42:28,279
Speaker 2: of friction, right, Like this is a weird thing to

848
00:42:28,280 --> 00:42:30,520
Speaker 2: say to a tech person, but like humanity is messy,

849
00:42:30,960 --> 00:42:36,360
Speaker 2: like vulnerability is weird, intimacy is awkward. It's all a mess,

850
00:42:36,360 --> 00:42:39,120
Speaker 2: but it's in that mess that we find purpose and

851
00:42:39,160 --> 00:42:43,640
Speaker 2: we build strength. And AI systems we now interact with

852
00:42:44,160 --> 00:42:48,920
Speaker 2: are so incredible in so many ways, and they are

853
00:42:48,960 --> 00:42:53,480
Speaker 2: a frictionless experience. And so I wonder thinking about my

854
00:42:53,600 --> 00:42:58,319
Speaker 2: son growing up, like will there be shortcuts? And if

855
00:42:58,320 --> 00:43:01,799
Speaker 2: we're not careful, we'll there not be the friction for

856
00:43:01,920 --> 00:43:05,919
Speaker 2: him to develop into the person that he could be.

857
00:43:07,440 --> 00:43:11,200
Speaker 1: Allie and I talk about this a lot, like Ali's

858
00:43:11,200 --> 00:43:14,000
Speaker 1: your husband. Yeah, I would maybe call it adversity, not friction,

859
00:43:14,880 --> 00:43:16,279
Speaker 1: but maybe friction is a better word.

860
00:43:16,360 --> 00:43:18,359
Speaker 2: Adversity to one person, friction to the other.

861
00:43:19,719 --> 00:43:23,440
Speaker 1: But like, so much of what I think I learned

862
00:43:23,680 --> 00:43:27,600
Speaker 1: was in the messiness, in the difficulty, actually even in

863
00:43:27,640 --> 00:43:30,000
Speaker 1: the boredom, Like you and I were kind of the

864
00:43:30,040 --> 00:43:34,319
Speaker 1: last generation of kids that grew up board. Yeah, and

865
00:43:34,560 --> 00:43:38,239
Speaker 1: at the time I hated it. Looking back, I think

866
00:43:38,239 --> 00:43:40,360
Speaker 1: it was just like super valuable in all of these

867
00:43:40,520 --> 00:43:41,160
Speaker 1: strange ways.

868
00:43:41,200 --> 00:43:43,040
Speaker 2: I grew up in the suburbs, hanging out at male

869
00:43:43,080 --> 00:43:45,480
Speaker 2: parking lots like that. Looking back on it, I'm like,

870
00:43:45,480 --> 00:43:46,040
Speaker 2: God help me.

871
00:43:46,080 --> 00:43:48,240
Speaker 1: But we didn't have we just know, no phones.

872
00:43:48,280 --> 00:43:51,040
Speaker 2: But there was a certain not to be nostalgic with

873
00:43:51,200 --> 00:43:54,120
Speaker 2: the person leading the tech revolution of AI. But like

874
00:43:54,200 --> 00:43:56,279
Speaker 2: there was a magic to that, And I also think

875
00:43:56,360 --> 00:43:58,000
Speaker 2: made me who I am totally.

876
00:43:58,880 --> 00:44:00,759
Speaker 1: So I don't want to like hang on to the

877
00:44:00,800 --> 00:44:03,160
Speaker 1: past too much, but I do want to find that,

878
00:44:03,360 --> 00:44:08,480
Speaker 1: like figure out how we don't throw out the good

879
00:44:08,480 --> 00:44:11,239
Speaker 1: parts of that and what it's going to take to

880
00:44:11,320 --> 00:44:14,000
Speaker 1: create an environment for kids to grow up in that

881
00:44:14,080 --> 00:44:17,120
Speaker 1: still has some of that. I think it's my intuition

882
00:44:17,200 --> 00:44:18,080
Speaker 1: is it's very important.

883
00:44:18,239 --> 00:44:21,400
Speaker 2: I want to play a little sound from doctor Becky.

884
00:44:21,440 --> 00:44:23,800
Speaker 2: Do you know who doctor Becky is? Okay, Well, doctor

885
00:44:23,800 --> 00:44:25,719
Speaker 2: Becky said something on this. I just interviewed her like

886
00:44:25,800 --> 00:44:27,799
Speaker 2: last week, and she had a thought on this.

887
00:44:28,560 --> 00:44:31,600
Speaker 5: If I know one thing about tweens and teens growing

888
00:44:31,680 --> 00:44:34,440
Speaker 5: up in an AI age, they're gonna have a lot

889
00:44:34,520 --> 00:44:38,160
Speaker 5: of tricky situations, a lot of messy situations. I said

890
00:44:38,160 --> 00:44:40,440
Speaker 5: to my kid recently, my fourteen year old, you know,

891
00:44:40,520 --> 00:44:43,160
Speaker 5: if you were walking to a town and you wanted

892
00:44:43,200 --> 00:44:45,160
Speaker 5: to get there and all of a sudden there's a

893
00:44:45,160 --> 00:44:48,080
Speaker 5: short cut, But do you take it? He was like, yeah,

894
00:44:48,160 --> 00:44:49,759
Speaker 5: this is a trick question. I was like, yeah, I

895
00:44:49,760 --> 00:44:50,200
Speaker 5: would too.

896
00:44:50,280 --> 00:44:50,480
Speaker 1: Good.

897
00:44:50,680 --> 00:44:56,279
Speaker 5: Okay, you're growing up at a time when there is

898
00:44:56,320 --> 00:45:01,759
Speaker 5: this shortcut for every act, pandemic, an emotional thing you

899
00:45:02,120 --> 00:45:03,040
Speaker 5: ever go through.

900
00:45:03,719 --> 00:45:05,359
Speaker 1: Like that is so.

901
00:45:05,520 --> 00:45:08,000
Speaker 5: Hard because this shortcut isn't like the shortcut to the town.

902
00:45:08,400 --> 00:45:10,600
Speaker 5: It is a shortcut. But the way it will add

903
00:45:10,680 --> 00:45:12,600
Speaker 5: up over time is it And things we've talked about,

904
00:45:12,600 --> 00:45:15,000
Speaker 5: It's going to be harder to think about things yourself.

905
00:45:15,320 --> 00:45:17,439
Speaker 5: It's going to be harder to tolerate people not getting

906
00:45:17,480 --> 00:45:19,520
Speaker 5: it right. It's going to overtime maybe take away from

907
00:45:19,560 --> 00:45:21,960
Speaker 5: the things you're trying to build to function in the world.

908
00:45:22,000 --> 00:45:24,080
Speaker 5: But it is a shortcut. I'm not going to pretend

909
00:45:24,120 --> 00:45:24,480
Speaker 5: it's not.

910
00:45:25,040 --> 00:45:29,759
Speaker 1: And so what a hard thing. I don't agree with that.

911
00:45:31,360 --> 00:45:34,160
Speaker 1: I mean, I agree with the worry, but the reason

912
00:45:34,200 --> 00:45:36,520
Speaker 1: I don't think that's what's going to happen. If you

913
00:45:36,600 --> 00:45:41,239
Speaker 1: were a kid from twenty thirty and you only had

914
00:45:41,239 --> 00:45:44,560
Speaker 1: to compete with kids from twenty twenty, yeah, you would

915
00:45:44,560 --> 00:45:45,920
Speaker 1: take the shortcut and you would do great, and you

916
00:45:45,920 --> 00:45:47,680
Speaker 1: would never really learn to think or struggle, and you

917
00:45:47,680 --> 00:45:52,040
Speaker 1: would just hugely outperform them. But the world we live

918
00:45:52,040 --> 00:45:53,560
Speaker 1: in is a competitive world, and I don't think that's

919
00:45:53,560 --> 00:45:55,680
Speaker 1: going to stop even if you did a lot of redistribution.

920
00:45:55,840 --> 00:45:59,400
Speaker 1: You know, people, we have a deep desire to excel

921
00:45:59,680 --> 00:46:02,640
Speaker 1: and be competitive and gain status and be useful to others.

922
00:46:03,280 --> 00:46:07,640
Speaker 1: And it's a multiplayer game, and all of the other

923
00:46:07,719 --> 00:46:09,520
Speaker 1: kids from twenty thirty are gonna have the same tools,

924
00:46:09,520 --> 00:46:12,160
Speaker 1: and it will push us, like the tools will raise

925
00:46:12,560 --> 00:46:15,160
Speaker 1: what we can do, but also they will raise expectations

926
00:46:15,160 --> 00:46:15,600
Speaker 1: even more.

927
00:46:16,680 --> 00:46:20,040
Speaker 2: Okay, I take your multiplayer game and I go back

928
00:46:20,040 --> 00:46:22,120
Speaker 2: to I know we don't want to keep going back,

929
00:46:22,160 --> 00:46:24,440
Speaker 2: but I go back to my childhood. I didn't have

930
00:46:24,600 --> 00:46:27,680
Speaker 2: I would say sometimes the happiest like my parents got

931
00:46:27,680 --> 00:46:30,000
Speaker 2: divorced when I was like I think I was like

932
00:46:30,000 --> 00:46:33,600
Speaker 2: eleven or something, and I just remember I had this journal,

933
00:46:33,760 --> 00:46:35,640
Speaker 2: similar to how you talk about writing to your son.

934
00:46:36,080 --> 00:46:38,319
Speaker 2: I had this journal that I would write, I you know,

935
00:46:38,560 --> 00:46:41,359
Speaker 2: paint the picture. I'm kind of a I'm you know,

936
00:46:41,480 --> 00:46:44,120
Speaker 2: Jewish girl at a largely Christian conservative school, growing up,

937
00:46:44,120 --> 00:46:46,640
Speaker 2: parents getting divorced, all sorts of stuff happening that I think,

938
00:46:46,680 --> 00:46:49,799
Speaker 2: are you know, not unique to the teenage experience. And

939
00:46:49,840 --> 00:46:51,560
Speaker 2: I wrote so much of it in my journal. And

940
00:46:51,600 --> 00:46:53,440
Speaker 2: I thought a lot about this because I did a

941
00:46:53,440 --> 00:46:56,840
Speaker 2: story on character AI and a chatbot that was incredibly

942
00:46:56,920 --> 00:46:59,040
Speaker 2: unhealthy in an emotional way for a fourteen year old

943
00:46:59,080 --> 00:47:01,759
Speaker 2: that was interacting with and didn't have the correct guardrails.

944
00:47:02,120 --> 00:47:04,400
Speaker 2: And I think about my journal. What if my journal

945
00:47:04,400 --> 00:47:07,279
Speaker 2: talked back. What if it was the therapist I didn't

946
00:47:07,280 --> 00:47:09,359
Speaker 2: have at the time. What if it was the guy

947
00:47:09,400 --> 00:47:11,960
Speaker 2: that wouldn't look at me in school right? What if

948
00:47:12,000 --> 00:47:15,440
Speaker 2: it was the parental figure. Would that have helped me?

949
00:47:15,760 --> 00:47:18,440
Speaker 2: Would that have been a level up in the game

950
00:47:18,680 --> 00:47:22,239
Speaker 2: of the multiplayer game that we talk about, or would

951
00:47:22,239 --> 00:47:25,560
Speaker 2: that have taken me further away from my reality and

952
00:47:25,760 --> 00:47:28,680
Speaker 2: the hardships that I think gave me a certain amount

953
00:47:28,680 --> 00:47:31,440
Speaker 2: of empathy that allows me to enable to do what

954
00:47:31,480 --> 00:47:34,000
Speaker 2: I do in my career and live the life I have,

955
00:47:34,640 --> 00:47:38,120
Speaker 2: And would it have made me more isolated and more lonely.

956
00:47:38,719 --> 00:47:42,440
Speaker 2: I don't know the answer, but I worry that it

957
00:47:41,640 --> 00:47:45,680
Speaker 2: could have taken away the special sauce of messiness and

958
00:47:45,800 --> 00:47:48,480
Speaker 2: adversity as you and Ollie liked to talk about that

959
00:47:48,600 --> 00:47:52,799
Speaker 2: makes me human. And so I just wonder, as someone

960
00:47:52,840 --> 00:47:56,400
Speaker 2: who wants to democratize artificial intelligence, and he thinks a

961
00:47:56,440 --> 00:47:58,440
Speaker 2: lot about these things, how do you make sure that

962
00:47:58,480 --> 00:48:02,160
Speaker 2: you don't democratize a shortcut that sacrifices the things in

963
00:48:02,239 --> 00:48:03,560
Speaker 2: us that are especially human.

964
00:48:04,480 --> 00:48:06,680
Speaker 1: I'm happy not to have to be a teenager again,

965
00:48:06,719 --> 00:48:14,839
Speaker 1: but I do remember the difficulty and the pain I

966
00:48:14,880 --> 00:48:18,919
Speaker 1: and I also remember just the absolute self centeredness and

967
00:48:19,040 --> 00:48:22,239
Speaker 1: how I thought like my struggles and life and was

968
00:48:22,239 --> 00:48:24,120
Speaker 1: the only thing that mattered in the world, you know,

969
00:48:24,920 --> 00:48:28,640
Speaker 1: and if I had and also how I kind of

970
00:48:28,640 --> 00:48:31,840
Speaker 1: only cared about my peers and whatever thing was happening

971
00:48:31,880 --> 00:48:34,920
Speaker 1: to me. And I don't think there's any well, I

972
00:48:35,040 --> 00:48:37,560
Speaker 1: loved video games, and we could spend a lot of

973
00:48:37,560 --> 00:48:40,400
Speaker 1: time playing them, but I don't think there's anything on

974
00:48:40,440 --> 00:48:45,320
Speaker 1: a computer that could have kept me from caring about

975
00:48:45,480 --> 00:48:48,080
Speaker 1: my peers in the world, and you know, like the

976
00:48:48,120 --> 00:48:50,640
Speaker 1: sort of normal teenage experience. I think it would have

977
00:48:50,640 --> 00:48:52,319
Speaker 1: helped me if I could write to something that would

978
00:48:52,360 --> 00:48:55,880
Speaker 1: talk back, but I don't think it would have been.

979
00:48:56,880 --> 00:48:58,880
Speaker 1: That's there's a lot of things I do worry about

980
00:48:58,960 --> 00:49:00,759
Speaker 1: for people who are going to grow up. This technology

981
00:49:01,239 --> 00:49:04,160
Speaker 1: that one seems like it should be should be pretty

982
00:49:04,160 --> 00:49:04,719
Speaker 1: positive to me.

983
00:49:05,680 --> 00:49:08,600
Speaker 2: I think with the correct guardrails, if it's talking back

984
00:49:09,760 --> 00:49:13,440
Speaker 2: and and you know, and not being so affirmative, that

985
00:49:13,520 --> 00:49:16,080
Speaker 2: it just goes because if you think that's intoxicating, if

986
00:49:16,120 --> 00:49:18,680
Speaker 2: you're a young person or any person and you have

987
00:49:18,760 --> 00:49:21,560
Speaker 2: a very powerful technology that's speaking back to you, that's

988
00:49:22,080 --> 00:49:24,480
Speaker 2: that goes where you wanted to go, that sees you like,

989
00:49:24,560 --> 00:49:25,239
Speaker 2: think about this.

990
00:49:25,320 --> 00:49:29,960
Speaker 1: Centetic AI models are incredibly dangerous. But AI models that

991
00:49:29,960 --> 00:49:36,879
Speaker 1: are kind of like thoughtful, carying listeners that will push

992
00:49:36,920 --> 00:49:40,120
Speaker 1: back as necessary, I think that could be a great thing.

993
00:49:40,560 --> 00:49:42,960
Speaker 2: But that's in the product, right, That's not our fault

994
00:49:43,000 --> 00:49:45,319
Speaker 2: on the outside, right, Like, that's in the product of

995
00:49:45,360 --> 00:49:47,879
Speaker 2: how these things are built. So take me into open AI,

996
00:49:47,960 --> 00:49:51,799
Speaker 2: into these conversations behind the scenes of you know, there

997
00:49:51,800 --> 00:49:53,880
Speaker 2: are these dials and you get to push them. I

998
00:49:53,920 --> 00:49:58,279
Speaker 2: remember when chat Gypt, when four oh the model was discontinued,

999
00:49:58,400 --> 00:50:02,160
Speaker 2: was really you know, that was really affirmative. There were

1000
00:50:02,239 --> 00:50:06,000
Speaker 2: people online that were almost treating it like it was

1001
00:50:06,040 --> 00:50:09,319
Speaker 2: a death. That is a powerful signal to someone like you.

1002
00:50:09,400 --> 00:50:13,960
Speaker 1: Absolutely, I think that points to not just the danger

1003
00:50:14,000 --> 00:50:19,400
Speaker 1: of a model that is too affirmative, but also how

1004
00:50:19,719 --> 00:50:22,040
Speaker 1: I mean, there were these really heartbreaking messages people would

1005
00:50:22,080 --> 00:50:26,920
Speaker 1: send stopping four O, which is, hey, this is the

1006
00:50:27,000 --> 00:50:28,680
Speaker 1: only thing in my life that has ever been a

1007
00:50:28,680 --> 00:50:30,280
Speaker 1: positive voice.

1008
00:50:30,440 --> 00:50:33,160
Speaker 2: You know what I think when you get those messages.

1009
00:50:33,640 --> 00:50:37,839
Speaker 1: I mean, it's heartbreaking. Obviously you really feel the weight

1010
00:50:37,880 --> 00:50:41,000
Speaker 1: of getting this right. Where it's easy to say, hey,

1011
00:50:41,000 --> 00:50:42,759
Speaker 1: we don't want a model that's too affirmative because it

1012
00:50:42,800 --> 00:50:45,360
Speaker 1: can make you know, it could like it can have

1013
00:50:45,400 --> 00:50:47,840
Speaker 1: negative mental effects on people. But then you hear somebody

1014
00:50:47,880 --> 00:50:51,080
Speaker 1: else say, like, actually, I never had any confidence. I

1015
00:50:51,120 --> 00:50:52,840
Speaker 1: had parents who told me I was terrible. I you know,

1016
00:50:52,880 --> 00:50:55,680
Speaker 1: didn't have any friends in school. And because I've had

1017
00:50:55,680 --> 00:50:58,480
Speaker 1: this model, sure maybe it's like a little too positive,

1018
00:50:58,480 --> 00:51:00,520
Speaker 1: but it's given me the confidence and I've gone out

1019
00:51:00,520 --> 00:51:02,279
Speaker 1: and got a job and found a girlfriend and like

1020
00:51:03,440 --> 00:51:04,879
Speaker 1: this has been the most important thing in my life

1021
00:51:04,920 --> 00:51:07,239
Speaker 1: and I'm doing great, but don't take it away. So

1022
00:51:07,600 --> 00:51:08,000
Speaker 1: it's like.

1023
00:51:08,520 --> 00:51:10,880
Speaker 2: It's almost this fine line because there are all of

1024
00:51:10,920 --> 00:51:13,600
Speaker 2: these stories and it's like, push this far and it

1025
00:51:13,640 --> 00:51:16,319
Speaker 2: can help you get offline. Push this far and you

1026
00:51:16,680 --> 00:51:21,120
Speaker 2: begin to lose that capacity for other humans. And I

1027
00:51:21,200 --> 00:51:23,640
Speaker 2: think that's a really interesting moment. I think this is

1028
00:51:23,640 --> 00:51:25,880
Speaker 2: probably one of the most important questions that we have

1029
00:51:25,960 --> 00:51:28,480
Speaker 2: to ask our tech leaders as they build out the

1030
00:51:28,520 --> 00:51:31,279
Speaker 2: future that our children will grow up in. What have

1031
00:51:31,360 --> 00:51:32,480
Speaker 2: you decided not to do?

1032
00:51:32,960 --> 00:51:35,040
Speaker 1: I mean, the specific thing we decided not to do is,

1033
00:51:35,080 --> 00:51:40,960
Speaker 1: although we think the models have extreme power to be

1034
00:51:40,960 --> 00:51:44,320
Speaker 1: a positive force in people's lives, we don't know how

1035
00:51:44,360 --> 00:51:47,759
Speaker 1: to balance a model that can be pushed too far

1036
00:51:47,800 --> 00:51:53,160
Speaker 1: in that direction with the concerns of I mean, I

1037
00:51:53,160 --> 00:51:54,920
Speaker 1: think you said is like kind of pushing people into

1038
00:51:54,960 --> 00:51:58,480
Speaker 1: a psychotic episode. And so we've made a decision, which

1039
00:51:58,520 --> 00:52:01,480
Speaker 1: is we know we're keeping something. And I'm not saying

1040
00:52:01,480 --> 00:52:03,680
Speaker 1: we shouldn't have warm models. Of course we should, but

1041
00:52:04,719 --> 00:52:06,160
Speaker 1: there are a lot of people who would like four

1042
00:52:06,239 --> 00:52:09,799
Speaker 1: row back, and we've decided, when looking at the full

1043
00:52:09,840 --> 00:52:15,200
Speaker 1: balance of upsides and risks, we can't quite offer that responsibly.

1044
00:52:17,800 --> 00:52:19,040
Speaker 1: There are a lot of people who would like us

1045
00:52:19,080 --> 00:52:21,879
Speaker 1: to have less restrictions on what our models can do

1046
00:52:22,000 --> 00:52:25,160
Speaker 1: from a bioperspective. Probably more people could you know, save

1047
00:52:25,200 --> 00:52:28,520
Speaker 1: their dogs lives with custom and my name vaccines if

1048
00:52:28,520 --> 00:52:30,360
Speaker 1: we made that a little easier and turned up the

1049
00:52:30,360 --> 00:52:32,360
Speaker 1: power on the model there. That would come with a

1050
00:52:32,360 --> 00:52:35,480
Speaker 1: big risk on novel pandemics that we've decided it's not

1051
00:52:35,520 --> 00:52:40,600
Speaker 1: worth the trade off now. Eventually, I think these decisions

1052
00:52:40,600 --> 00:52:42,719
Speaker 1: will be made by society. You know, it's like not

1053
00:52:42,880 --> 00:52:45,640
Speaker 1: up to the see of a company that makes airplanes

1054
00:52:45,640 --> 00:52:48,680
Speaker 1: what the safety regulations are. But for now we kind

1055
00:52:48,680 --> 00:52:49,239
Speaker 1: of got to do it.

1056
00:53:02,160 --> 00:53:04,520
Speaker 2: We've obviously talked a lot about the fears, but there

1057
00:53:04,560 --> 00:53:07,399
Speaker 2: is so much excitement about what this technology can do.

1058
00:53:07,920 --> 00:53:13,160
Speaker 2: Open Air has a foundation that's focused on humanity's biggest problems.

1059
00:53:13,239 --> 00:53:15,960
Speaker 2: Where do you see this being good for humans in

1060
00:53:15,960 --> 00:53:17,040
Speaker 2: a very specific way.

1061
00:53:17,560 --> 00:53:21,040
Speaker 1: My most exciting meaning of last week was with a

1062
00:53:21,040 --> 00:53:25,640
Speaker 1: physicist who's using one of our latest internal systems. He said,

1063
00:53:25,680 --> 00:53:27,239
Speaker 1: you know, my mind has been completely blown and it

1064
00:53:27,320 --> 00:53:30,719
Speaker 1: makes whatever happened. We are going to make decades worth

1065
00:53:31,320 --> 00:53:35,080
Speaker 1: theoretical physics progress in the next couple of years. And

1066
00:53:35,600 --> 00:53:38,560
Speaker 1: you know, some people will say, Okay, who cares, you

1067
00:53:38,560 --> 00:53:41,600
Speaker 1: can write a better equation, But I am a believer

1068
00:53:42,120 --> 00:53:45,040
Speaker 1: that scientific progress is one of the most maybe the

1069
00:53:45,120 --> 00:53:47,600
Speaker 1: most important thing we can do to make the world

1070
00:53:47,680 --> 00:53:49,960
Speaker 1: better and better. Obviously, we can do things like developed

1071
00:53:50,000 --> 00:53:53,840
Speaker 1: new cures for diseases, but there's so much more material science,

1072
00:53:54,080 --> 00:53:56,200
Speaker 1: new ways to produce clean and cheap energy and bond

1073
00:53:56,200 --> 00:54:01,400
Speaker 1: and energy that if we are really on the precipice

1074
00:54:01,840 --> 00:54:06,080
Speaker 1: of being able to use AI to dramatically accelerate science,

1075
00:54:06,640 --> 00:54:08,440
Speaker 1: then I think we can solve some of the biggest

1076
00:54:08,440 --> 00:54:11,400
Speaker 1: problems in the world. And we now have one of

1077
00:54:11,440 --> 00:54:13,640
Speaker 1: the most maybe the most well funded foundation in the world,

1078
00:54:14,719 --> 00:54:18,360
Speaker 1: and the ability to use that amount of capital with

1079
00:54:18,440 --> 00:54:21,920
Speaker 1: the technology to go create a bunch of new scientific

1080
00:54:22,040 --> 00:54:24,440
Speaker 1: understanding for the world. I hope that'll be one of

1081
00:54:24,480 --> 00:54:25,680
Speaker 1: our greatest contributions ever.

1082
00:54:26,280 --> 00:54:29,479
Speaker 2: A big moment happened recently, which is a California jury

1083
00:54:29,520 --> 00:54:33,239
Speaker 2: found that Meta and Google reliable and damage is for

1084
00:54:33,320 --> 00:54:37,520
Speaker 2: harmful and addictive platforms. So the first time that social media

1085
00:54:37,560 --> 00:54:40,279
Speaker 2: has been treated as a defective product from a design perspective.

1086
00:54:40,680 --> 00:54:43,120
Speaker 2: I'm curious what the outcome of this case means for

1087
00:54:43,200 --> 00:54:44,640
Speaker 2: AA companies like open Ai.

1088
00:54:45,400 --> 00:54:47,239
Speaker 1: I mean, I think society is going to decide that

1089
00:54:47,320 --> 00:54:50,440
Speaker 1: creators of AI products bear a tremendous amount of responsibility

1090
00:54:50,480 --> 00:54:52,480
Speaker 1: to the products we put out in the world. But

1091
00:54:53,520 --> 00:54:56,560
Speaker 1: I kind of thought that before this case, and I

1092
00:54:56,600 --> 00:54:58,560
Speaker 1: haven't reviewed this enough to say if there's like a

1093
00:54:58,840 --> 00:55:00,360
Speaker 1: specifically they will apply to us.

1094
00:55:01,080 --> 00:55:04,240
Speaker 2: Yeah, I think so much of the fear around AI

1095
00:55:04,400 --> 00:55:07,680
Speaker 2: and who has the power is based in grounded and

1096
00:55:08,080 --> 00:55:10,600
Speaker 2: you know, people worried that their jobs might not exist,

1097
00:55:10,600 --> 00:55:14,759
Speaker 2: their livelihood might not exist. I'd love to play a

1098
00:55:14,880 --> 00:55:18,400
Speaker 2: question from a mom who lives in Rhode Island. A

1099
00:55:18,400 --> 00:55:20,719
Speaker 2: big part of this show, and what I want to

1100
00:55:20,760 --> 00:55:23,160
Speaker 2: do is be able to give people like that access

1101
00:55:23,200 --> 00:55:25,960
Speaker 2: to people like you to ask real questions. So I

1102
00:55:26,000 --> 00:55:26,560
Speaker 2: want to play that.

1103
00:55:27,600 --> 00:55:31,280
Speaker 6: Hi, I'm the mom of two middle schoolers in Rhode Island,

1104
00:55:31,440 --> 00:55:33,920
Speaker 6: and my kids are kind of coming around to the

1105
00:55:33,960 --> 00:55:38,759
Speaker 6: idea of having a backup plan for their careers. You know,

1106
00:55:38,800 --> 00:55:41,719
Speaker 6: as most kids, they have aspirational careers of being professional

1107
00:55:41,760 --> 00:55:47,160
Speaker 6: sports players and you know, fighter jet pilots. And we've

1108
00:55:47,200 --> 00:55:51,200
Speaker 6: been chatting about other options for careers and I'm curious

1109
00:55:51,239 --> 00:55:55,719
Speaker 6: to see what careers you might recommend that would be

1110
00:55:55,840 --> 00:55:58,440
Speaker 6: AI proof. My husband and I have been talking a

1111
00:55:58,440 --> 00:56:02,160
Speaker 6: lot about skilled trades. How you know, as a homeowner

1112
00:56:02,680 --> 00:56:05,440
Speaker 6: it's difficult to find skilled trades people to come and

1113
00:56:05,520 --> 00:56:08,000
Speaker 6: help you out with electrical problems.

1114
00:56:07,440 --> 00:56:09,400
Speaker 2: With heating and cooling options.

1115
00:56:09,480 --> 00:56:13,160
Speaker 6: Maybe that's a great path for kids these days, and

1116
00:56:13,320 --> 00:56:17,600
Speaker 6: more traditional like great career paths and engineering might take

1117
00:56:17,640 --> 00:56:22,399
Speaker 6: a backseat to these skilled trades. Obviously, kids will do

1118
00:56:22,520 --> 00:56:26,920
Speaker 6: what you know, very gravitate toward, and we encourage that.

1119
00:56:27,160 --> 00:56:31,640
Speaker 6: But if you have thoughts on AI proof careers, I'd

1120
00:56:31,680 --> 00:56:33,120
Speaker 6: be interested to hear them.

1121
00:56:33,400 --> 00:56:36,880
Speaker 1: Thanks so much. One thing I will point out, though,

1122
00:56:37,160 --> 00:56:39,239
Speaker 1: in terms of things that I think are truly AI proof,

1123
00:56:40,160 --> 00:56:43,560
Speaker 1: is I don't think anyone cares about the AI version

1124
00:56:43,680 --> 00:56:47,160
Speaker 1: of Lori's Siegel. I don't think anyone wants to watch

1125
00:56:47,680 --> 00:56:51,080
Speaker 1: an AI come interview people like me. We are obsessed

1126
00:56:51,120 --> 00:56:53,759
Speaker 1: with other people. I don't think anyone care. You know,

1127
00:56:53,880 --> 00:56:56,360
Speaker 1: she talked about our kids want to be sports stars

1128
00:56:56,360 --> 00:56:59,560
Speaker 1: and fighter jet pilots. Maybe fighter jet pilots. Society is

1129
00:56:59,560 --> 00:57:02,239
Speaker 1: happy to say, like, yeah, put the AI in there,

1130
00:57:02,280 --> 00:57:07,000
Speaker 1: and that's you know, safer and better, and okay, the

1131
00:57:07,040 --> 00:57:11,600
Speaker 1: sports stars. Are we really gonna watch like eleven robots

1132
00:57:11,680 --> 00:57:14,040
Speaker 1: play sports against each other? Some people will.

1133
00:57:14,120 --> 00:57:16,040
Speaker 2: I mean a lot of that stuff goes viral. I

1134
00:57:16,080 --> 00:57:18,240
Speaker 2: saw like a robot staring more dancing.

1135
00:57:18,400 --> 00:57:21,040
Speaker 1: It has novelty, But like, are we gonna get caught

1136
00:57:21,120 --> 00:57:23,400
Speaker 1: up in a sports league of just robots without the

1137
00:57:23,400 --> 00:57:27,600
Speaker 1: personal interest? Are we gonna read ai written novels and

1138
00:57:27,760 --> 00:57:30,760
Speaker 1: not have an author that we could like understand their

1139
00:57:30,760 --> 00:57:34,040
Speaker 1: life story and went into that. I would bet there's

1140
00:57:34,080 --> 00:57:37,560
Speaker 1: a huge cluster of people care about people, and in

1141
00:57:37,600 --> 00:57:39,560
Speaker 1: the future people care even more about people. But like,

1142
00:57:39,760 --> 00:57:42,760
Speaker 1: this is so deep in evolutionary biology that those things

1143
00:57:42,800 --> 00:57:48,800
Speaker 1: are like all quite a I proof someone said to

1144
00:57:48,840 --> 00:57:51,520
Speaker 1: me recently. I thought it was with the time, it

1145
00:57:51,560 --> 00:57:53,360
Speaker 1: seemed not that profound, more insightful the more I've thought

1146
00:57:53,400 --> 00:57:57,240
Speaker 1: about it. In a world with true abundance, there's still

1147
00:57:57,320 --> 00:58:03,320
Speaker 1: one limited thing, which is human attention. And we care

1148
00:58:03,320 --> 00:58:05,680
Speaker 1: a lot about human attention. So there's a whole category

1149
00:58:05,680 --> 00:58:07,920
Speaker 1: of stuff there's on the specifics, I'd go into it

1150
00:58:07,920 --> 00:58:10,520
Speaker 1: if we have more time, But like, yes, clearly there

1151
00:58:10,520 --> 00:58:12,480
Speaker 1: will be a lot a huge amount of short term

1152
00:58:12,480 --> 00:58:16,480
Speaker 1: demand for electricians and skilled trades people. Eventually. I think

1153
00:58:16,480 --> 00:58:19,360
Speaker 1: the robots will do some of those things. But you know,

1154
00:58:19,440 --> 00:58:21,200
Speaker 1: all of these things are just sort of like periods

1155
00:58:21,360 --> 00:58:23,200
Speaker 1: of time. If you just look at how much jobs

1156
00:58:23,200 --> 00:58:25,240
Speaker 1: have shifted every last decade that will keep going.

1157
00:58:25,400 --> 00:58:27,520
Speaker 2: You talk about human attention, and so much of the

1158
00:58:27,560 --> 00:58:30,280
Speaker 2: last decade of tech was kind of like really a

1159
00:58:30,320 --> 00:58:33,960
Speaker 2: business model that preyed on human attention. I'm curious how

1160
00:58:34,000 --> 00:58:36,800
Speaker 2: you think about the lessons of the last decade because

1161
00:58:36,800 --> 00:58:42,240
Speaker 2: now you're a for profit company and there is this

1162
00:58:42,320 --> 00:58:45,160
Speaker 2: idea that eyeballs on screen, more people using the product,

1163
00:58:45,480 --> 00:58:48,640
Speaker 2: go go go, is what you know is what makes

1164
00:58:48,680 --> 00:58:50,919
Speaker 2: the company successful? And how do you not fall into

1165
00:58:50,960 --> 00:58:53,880
Speaker 2: the same business model trap that I would say the

1166
00:58:53,960 --> 00:58:58,080
Speaker 2: last decade.

1167
00:58:55,840 --> 00:59:00,240
Speaker 1: We decided to just totally shut down, soa we were

1168
00:59:00,240 --> 00:59:03,360
Speaker 1: thinking about other versions of keeping it before the computer

1169
00:59:03,400 --> 00:59:05,040
Speaker 1: crunch came, Like we were talking about putting it into

1170
00:59:05,080 --> 00:59:08,040
Speaker 1: the chatcheapy t app, really focusing on generation and creativity.

1171
00:59:08,560 --> 00:59:11,120
Speaker 1: But one thing that we had realized is that to

1172
00:59:13,440 --> 00:59:17,680
Speaker 1: succeed with it, as the product was currently conceptualized and

1173
00:59:17,720 --> 00:59:19,040
Speaker 1: sort of this way, you could watch a lot of

1174
00:59:19,080 --> 00:59:21,880
Speaker 1: videos that would have put a series of incentives on

1175
00:59:21,960 --> 00:59:23,960
Speaker 1: us and would have led to a bunch of decisions

1176
00:59:24,000 --> 00:59:26,120
Speaker 1: to win that we just didn't want to make. So

1177
00:59:26,280 --> 00:59:29,680
Speaker 1: there was some version of that, and there still is

1178
00:59:29,720 --> 00:59:32,120
Speaker 1: some version of letting people generate video that I think

1179
00:59:32,200 --> 00:59:35,200
Speaker 1: is awesome to compete in short term video feeds and

1180
00:59:35,200 --> 00:59:36,640
Speaker 1: where that's going to go on A world of AI

1181
00:59:37,040 --> 00:59:38,360
Speaker 1: was not something I wanted to have.

1182
00:59:38,280 --> 00:59:41,560
Speaker 2: To go do, right It's an interesting decision of what

1183
00:59:41,600 --> 00:59:44,840
Speaker 2: you don't do like and what that business model could demand.

1184
00:59:45,840 --> 00:59:47,680
Speaker 1: And there can be other versions of like this with chat,

1185
00:59:47,720 --> 00:59:50,040
Speaker 1: like we could make a version of chatchbyt that was

1186
00:59:50,200 --> 00:59:53,160
Speaker 1: very addicting to talk to. Now I have a bunch

1187
00:59:53,160 --> 00:59:54,080
Speaker 1: of complications too.

1188
00:59:55,160 --> 00:59:59,480
Speaker 2: I want to play another another voice memo from a

1189
00:59:59,480 --> 00:59:59,960
Speaker 2: guy named Joe.

1190
01:00:00,560 --> 01:00:02,040
Speaker 1: The clips are a fun little I know.

1191
01:00:02,240 --> 01:00:03,520
Speaker 2: Let's play Joseph.

1192
01:00:03,720 --> 01:00:06,680
Speaker 4: Hey, my name is Joseph. I'm a recent college grad

1193
01:00:07,160 --> 01:00:09,360
Speaker 4: trying to work in film in New York City, and

1194
01:00:09,440 --> 01:00:12,360
Speaker 4: I have a couple of questions for you. Number one,

1195
01:00:12,480 --> 01:00:15,320
Speaker 4: what do you dream about? Do you ever have recurring

1196
01:00:15,400 --> 01:00:18,960
Speaker 4: nightmares about AI? Do you often have dreams about AI?

1197
01:00:19,600 --> 01:00:23,280
Speaker 4: Number two, if you were in my shoes as like

1198
01:00:23,400 --> 01:00:26,360
Speaker 4: a young person trying to find themselves in a career

1199
01:00:26,840 --> 01:00:29,240
Speaker 4: or just trying to like live in this world, what

1200
01:00:29,680 --> 01:00:31,840
Speaker 4: do you think is important to read and learn about

1201
01:00:31,880 --> 01:00:35,280
Speaker 4: right now? And what skills are most useful? And then finally,

1202
01:00:36,480 --> 01:00:40,320
Speaker 4: what specific social causes do you care about? Thank you

1203
01:00:40,400 --> 01:00:40,760
Speaker 4: so much.

1204
01:00:41,800 --> 01:00:43,560
Speaker 2: It's a lot of them, so you can pick a couple,

1205
01:00:43,760 --> 01:00:45,400
Speaker 2: pick your favorite one. Thanks Joseph.

1206
01:00:45,480 --> 01:00:49,520
Speaker 1: You them all fast. When I until my early thirties,

1207
01:00:50,160 --> 01:00:54,360
Speaker 1: I had this recurring kind of bad dream of that.

1208
01:00:54,560 --> 01:00:58,400
Speaker 1: Right when I dropped out of college to start my company,

1209
01:01:00,920 --> 01:01:03,320
Speaker 1: I would be in that period, but I was like

1210
01:01:03,440 --> 01:01:06,280
Speaker 1: always I was in my dream, I was still in school,

1211
01:01:06,280 --> 01:01:08,240
Speaker 1: and I was always missing something, and I was always

1212
01:01:08,320 --> 01:01:10,200
Speaker 1: like you know, late for an exam, er for out

1213
01:01:10,200 --> 01:01:12,840
Speaker 1: about a class. And this went on to like deep

1214
01:01:12,840 --> 01:01:15,360
Speaker 1: into open AI time and by far the most frequent

1215
01:01:15,440 --> 01:01:19,080
Speaker 1: dream of my life, and I it was so vivid.

1216
01:01:19,200 --> 01:01:21,560
Speaker 1: It was a little different every time, but it was

1217
01:01:21,640 --> 01:01:25,680
Speaker 1: like a kind of a crazy, most intense dream. So

1218
01:01:25,800 --> 01:01:28,120
Speaker 1: I don't know what to make of that. What I

1219
01:01:28,160 --> 01:01:31,320
Speaker 1: would study. I think the answer should just get really

1220
01:01:31,320 --> 01:01:34,040
Speaker 1: good at using AI tools more than any specific thing

1221
01:01:34,080 --> 01:01:36,760
Speaker 1: to study, and then the kind of skills that I

1222
01:01:36,800 --> 01:01:38,320
Speaker 1: expect to matter in the future. And I think these

1223
01:01:38,320 --> 01:01:44,439
Speaker 1: are learnable like resilience, adaptability, kind of like calmness, being

1224
01:01:44,480 --> 01:01:48,680
Speaker 1: comfortable with a lot of change social causes. I am

1225
01:01:48,800 --> 01:01:50,840
Speaker 1: very interested in this question of what the new version

1226
01:01:50,880 --> 01:01:52,880
Speaker 1: of the social contract looks like, you know, I don't

1227
01:01:52,880 --> 01:01:55,200
Speaker 1: think it's enough to say, hey, this job transition is

1228
01:01:55,200 --> 01:01:56,400
Speaker 1: going to be really hard, but there's going to be

1229
01:01:56,400 --> 01:01:58,680
Speaker 1: a great time on the other side. I think we

1230
01:01:58,720 --> 01:02:00,080
Speaker 1: can say that, and then we have to say, and

1231
01:02:00,120 --> 01:02:02,680
Speaker 1: here's our proposal for a new tax system and a

1232
01:02:02,680 --> 01:02:05,840
Speaker 1: new way to think about collective ownership that makes sense.

1233
01:02:05,880 --> 01:02:06,840
Speaker 2: What would you propose?

1234
01:02:07,600 --> 01:02:12,520
Speaker 1: I am interested in ideas where everybody is an owner

1235
01:02:12,680 --> 01:02:15,520
Speaker 1: in the magic of capitalism. I think that the thing

1236
01:02:15,560 --> 01:02:20,200
Speaker 1: that happened with these new Trump accounts for kids, that

1237
01:02:20,280 --> 01:02:23,040
Speaker 1: direction of stuff. I really like. So a world where

1238
01:02:23,120 --> 01:02:26,160
Speaker 1: like we kind of preserve as much of what's worked

1239
01:02:26,200 --> 01:02:30,160
Speaker 1: about capitalism of last couple of centuries as we can,

1240
01:02:30,440 --> 01:02:32,440
Speaker 1: but we say, hey, everybody, just by virtue of being

1241
01:02:32,440 --> 01:02:34,040
Speaker 1: a citizen is going to own some and we're going

1242
01:02:34,080 --> 01:02:38,960
Speaker 1: to have like a tax system that enables that. I

1243
01:02:39,000 --> 01:02:40,960
Speaker 1: feel very excited about that. I think as long as

1244
01:02:40,960 --> 01:02:44,240
Speaker 1: people can afford a good life in the price of admission,

1245
01:02:44,240 --> 01:02:46,080
Speaker 1: they will outperform expectations.

1246
01:02:46,640 --> 01:02:49,920
Speaker 2: Being in front of you and having spent so much

1247
01:02:49,960 --> 01:02:52,840
Speaker 2: time over the last four years, we discovered there was

1248
01:02:52,920 --> 01:02:56,720
Speaker 2: this site on the internet right where a man who

1249
01:02:56,760 --> 01:03:02,120
Speaker 2: is anonymous for seven years prior had hyperrealistic, sexually explicit

1250
01:03:02,200 --> 01:03:05,800
Speaker 2: deep fakes of women using artificial intelligence. It was it

1251
01:03:05,840 --> 01:03:08,000
Speaker 2: was horrific looking at like deep fake porn and the

1252
01:03:08,000 --> 01:03:11,560
Speaker 2: way that these impacted women and so much of that.

1253
01:03:11,760 --> 01:03:15,400
Speaker 2: I think, you know, when we look at agency and control.

1254
01:03:15,640 --> 01:03:18,640
Speaker 2: I talked to so many women and it's not just women,

1255
01:03:18,680 --> 01:03:23,000
Speaker 2: all sorts of victims who are victims of deep fake pornography,

1256
01:03:23,200 --> 01:03:25,440
Speaker 2: and they felt like they lost their sense of agency,

1257
01:03:25,440 --> 01:03:27,920
Speaker 2: They had no control. They went to the laws hadn't

1258
01:03:27,960 --> 01:03:29,720
Speaker 2: caught up, so they went to the police. They said,

1259
01:03:29,920 --> 01:03:32,400
Speaker 2: we don't even know what you're talking about. I had

1260
01:03:32,400 --> 01:03:34,400
Speaker 2: a woman talk about wanting to end her life, and

1261
01:03:34,440 --> 01:03:37,600
Speaker 2: so the human impact of this is so incredibly visceral.

1262
01:03:38,200 --> 01:03:40,640
Speaker 2: And I think about this through the lens of like, well,

1263
01:03:40,640 --> 01:03:43,760
Speaker 2: whose job is this? Right? You know? And granted I

1264
01:03:43,800 --> 01:03:46,520
Speaker 2: say this with the caveat that OpenAI has gotten in

1265
01:03:46,520 --> 01:03:48,280
Speaker 2: front of a lot of this stuff. We're not seeing

1266
01:03:48,320 --> 01:03:50,440
Speaker 2: a lot of these were open source models. But I

1267
01:03:50,440 --> 01:03:52,800
Speaker 2: think that you're an important voice to be able to

1268
01:03:52,840 --> 01:03:54,120
Speaker 2: weigh in on this type of thing.

1269
01:03:55,000 --> 01:03:57,080
Speaker 1: I mean that one seems like such a that's not

1270
01:03:57,120 --> 01:03:59,040
Speaker 1: a hard decision to make. Right, Like, that's so clear

1271
01:03:59,080 --> 01:04:00,880
Speaker 1: that we're not going to our models to be used

1272
01:04:00,920 --> 01:04:03,800
Speaker 1: for that. Open source models. Yes, people are going to

1273
01:04:03,960 --> 01:04:07,600
Speaker 1: use a lot of that. I would kind of go

1274
01:04:07,640 --> 01:04:11,280
Speaker 1: back to the principles that I outlined earlier. You can

1275
01:04:11,640 --> 01:04:14,680
Speaker 1: say a lot of bad things about the Republic, but

1276
01:04:15,480 --> 01:04:18,720
Speaker 1: I think it has done a good job of representing

1277
01:04:21,400 --> 01:04:25,400
Speaker 1: the will of the citizens and protecting against the tyranny

1278
01:04:25,400 --> 01:04:28,440
Speaker 1: of the majority in that process. I think the Constitution

1279
01:04:28,720 --> 01:04:31,680
Speaker 1: remains one of the most amazing governing docs the world

1280
01:04:31,720 --> 01:04:34,160
Speaker 1: has ever produced, and I believe in it as a

1281
01:04:34,160 --> 01:04:40,040
Speaker 1: governance system. So like with other technologies that have changed

1282
01:04:40,040 --> 01:04:46,440
Speaker 1: the world, I hope that is our systems, our governments

1283
01:04:46,920 --> 01:04:51,440
Speaker 1: figure out how to run the process to decide how

1284
01:04:51,440 --> 01:04:56,480
Speaker 1: to make these trade offs about say freedom and protection

1285
01:04:56,560 --> 01:04:59,760
Speaker 1: and what's in the best interest of society. Now, I

1286
01:04:59,760 --> 01:05:01,920
Speaker 1: think you might say, and I would agree with, is

1287
01:05:02,040 --> 01:05:04,720
Speaker 1: maybe that worked really well in a different time. Our

1288
01:05:04,800 --> 01:05:06,880
Speaker 1: leaders didn't do a great job with the Internet or

1289
01:05:06,920 --> 01:05:10,680
Speaker 1: with social media whatever, and that is fair, which is

1290
01:05:10,680 --> 01:05:12,360
Speaker 1: part of why we've been trying to talk about this

1291
01:05:13,040 --> 01:05:13,920
Speaker 1: more and earlier.

1292
01:05:15,280 --> 01:05:17,840
Speaker 2: But if I could say specifically, so much so many

1293
01:05:17,880 --> 01:05:20,800
Speaker 2: of the folks victims I've talked to finally had recourse

1294
01:05:20,840 --> 01:05:23,360
Speaker 2: when there were state laws. You know, there yet to

1295
01:05:23,360 --> 01:05:26,160
Speaker 2: be a federal law. Takes the it takes a long time.

1296
01:05:26,960 --> 01:05:30,440
Speaker 2: I know open AI has looked at a supported cribbing

1297
01:05:30,680 --> 01:05:33,880
Speaker 2: state legislation in the name of innovation, and we have

1298
01:05:33,960 --> 01:05:37,720
Speaker 2: to move fast. So I'm curious if you can explain this.

1299
01:05:37,680 --> 01:05:40,760
Speaker 1: May be one we disagree on. I don't think state

1300
01:05:40,880 --> 01:05:44,160
Speaker 1: legislation will be helpful here. I would much rather see

1301
01:05:44,200 --> 01:05:48,120
Speaker 1: the system urgently try to get a federal framework right now.

1302
01:05:48,240 --> 01:05:50,400
Speaker 1: Maybe we try that for a couple of years and

1303
01:05:50,440 --> 01:05:52,360
Speaker 1: we can't and we say, hey, it's just taking too long,

1304
01:05:52,400 --> 01:05:56,120
Speaker 1: and you know, we need something. But like, I don't

1305
01:05:56,200 --> 01:06:01,800
Speaker 1: think I think we need one stance on this. I mean, really,

1306
01:06:02,040 --> 01:06:03,800
Speaker 1: there's even trickier thing which I think we need a

1307
01:06:03,880 --> 01:06:05,720
Speaker 1: global stance on a lot of this. A lot of

1308
01:06:05,720 --> 01:06:07,920
Speaker 1: these things are going to affect the whole world. We've

1309
01:06:07,920 --> 01:06:10,960
Speaker 1: talked about things like bioterror, but certainly what happens in

1310
01:06:11,000 --> 01:06:13,040
Speaker 1: one state will affect a lot of other states.

1311
01:06:13,080 --> 01:06:16,440
Speaker 2: But in theory, but in reality, sometimes the change happens

1312
01:06:16,480 --> 01:06:17,680
Speaker 2: quicker at the state level.

1313
01:06:18,560 --> 01:06:22,400
Speaker 1: I think even if it does, it doesn't mean that

1314
01:06:22,400 --> 01:06:24,880
Speaker 1: that will have the impact on the industry or the

1315
01:06:24,880 --> 01:06:26,840
Speaker 1: technology you would hope for. You know, if we take

1316
01:06:26,840 --> 01:06:29,480
Speaker 1: an area like social where I think we'd both agree

1317
01:06:29,520 --> 01:06:32,320
Speaker 1: that government didn't really do their job. Public pressure did

1318
01:06:32,360 --> 01:06:34,800
Speaker 1: a lot, yeah, and the companies, you know, I think

1319
01:06:34,840 --> 01:06:36,640
Speaker 1: like most people that work at these companies probably want

1320
01:06:36,640 --> 01:06:38,800
Speaker 1: to do the right thing too, And the companies not

1321
01:06:38,840 --> 01:06:40,640
Speaker 1: every decision I would have made the same way, but

1322
01:06:40,760 --> 01:06:43,480
Speaker 1: I think they made a lot of good ones. So

1323
01:06:44,040 --> 01:06:47,600
Speaker 1: it's not like we only have governments to rely on here.

1324
01:06:48,040 --> 01:06:49,800
Speaker 2: I want to do a quick lightning round what will

1325
01:06:49,800 --> 01:06:52,000
Speaker 2: be the most valuable human skill set that's AI.

1326
01:06:51,880 --> 01:06:53,800
Speaker 1: Proof caring about other people?

1327
01:06:55,000 --> 01:06:58,320
Speaker 2: Open AIS biggest opportunity and biggest missed opportunity.

1328
01:06:58,800 --> 01:07:00,880
Speaker 1: Biggest opportunity in front of us is this sort of

1329
01:07:01,800 --> 01:07:06,400
Speaker 1: cluster of automating researchers and companies and the super personal

1330
01:07:06,400 --> 01:07:10,080
Speaker 1: assistant for people's life. Biggest missed opportunity is we passed

1331
01:07:10,120 --> 01:07:11,840
Speaker 1: on a compute deal once we shouldn't have passed on.

1332
01:07:11,880 --> 01:07:15,400
Speaker 1: I got big one, like what I can't say exactly what,

1333
01:07:15,480 --> 01:07:17,440
Speaker 1: but there was like a very large compute deal we

1334
01:07:17,440 --> 01:07:18,000
Speaker 1: could have taken.

1335
01:07:18,320 --> 01:07:20,680
Speaker 2: Okay, how likely is open A I do acquire an

1336
01:07:20,760 --> 01:07:21,560
Speaker 2: entertainment company.

1337
01:07:22,040 --> 01:07:24,320
Speaker 1: Not a top consideration right now.

1338
01:07:24,360 --> 01:07:26,120
Speaker 2: How likely is open A to go public this year?

1339
01:07:26,440 --> 01:07:27,000
Speaker 1: Could happen?

1340
01:07:27,080 --> 01:07:27,160
Speaker 2: Not?

1341
01:07:27,560 --> 01:07:28,280
Speaker 1: Yeah, I don't know.

1342
01:07:28,400 --> 01:07:30,080
Speaker 2: How likely is it that you could one day be

1343
01:07:30,160 --> 01:07:33,280
Speaker 2: replaced as CEO by an AI.

1344
01:07:34,280 --> 01:07:37,720
Speaker 1: On the on the skills of my job? Very likely.

1345
01:07:38,360 --> 01:07:42,720
Speaker 1: On the will the world want a human responsible for

1346
01:07:42,760 --> 01:07:44,560
Speaker 1: the decisions of a company like open AI? I think

1347
01:07:44,560 --> 01:07:45,440
Speaker 1: the world will demand that.

1348
01:07:45,800 --> 01:07:47,600
Speaker 2: Do you say please and thank you to check GPT?

1349
01:07:47,880 --> 01:07:48,200
Speaker 1: I do?

1350
01:07:49,000 --> 01:07:52,200
Speaker 2: Why force of habit? Not because you're expecting if it

1351
01:07:52,240 --> 01:07:53,760
Speaker 2: all goes bad, you want to be like on the

1352
01:07:53,760 --> 01:07:57,240
Speaker 2: other side. Okay, got it. Last time you spoken to

1353
01:07:57,240 --> 01:07:58,600
Speaker 2: Elon Musk and what was the conversation?

1354
01:07:59,600 --> 01:08:01,800
Speaker 1: Aland I was just some emojis back and forth?

1355
01:08:03,200 --> 01:08:05,680
Speaker 2: Have you speaking? Have you seen the fruit AI videos online?

1356
01:08:05,720 --> 01:08:05,800
Speaker 1: No?

1357
01:08:06,320 --> 01:08:10,640
Speaker 2: Oh, they represent AI slop at its finest when what

1358
01:08:10,720 --> 01:08:13,480
Speaker 2: we see AI wearables mainstream.

1359
01:08:13,600 --> 01:08:16,720
Speaker 1: Two to four years. There's the reason I give such

1360
01:08:16,760 --> 01:08:21,760
Speaker 1: a broad range is right now, if you're talking to

1361
01:08:21,800 --> 01:08:23,640
Speaker 1: someone that's got like a ford facing camera on their

1362
01:08:23,640 --> 01:08:28,080
Speaker 1: glasses is incredibly off putting. Yeah, maybe that gets more

1363
01:08:28,120 --> 01:08:30,000
Speaker 1: comfortable quickly, or maybe we just figure out a very

1364
01:08:30,000 --> 01:08:30,920
Speaker 1: different kind of wearable.

1365
01:08:32,320 --> 01:08:34,120
Speaker 2: Could you give any sense of how you believe that

1366
01:08:34,160 --> 01:08:36,960
Speaker 2: will look. Is it like a pin people are wearing?

1367
01:08:37,240 --> 01:08:40,559
Speaker 1: Pin? Doesn't feel quite right to me. I mean, I

1368
01:08:40,600 --> 01:08:43,919
Speaker 1: hope it like more as a general thing, not about wearables.

1369
01:08:43,920 --> 01:08:46,720
Speaker 1: I hope that the whole concept of products in the

1370
01:08:46,760 --> 01:08:49,920
Speaker 1: age of AI fade into the background. Like the ideal

1371
01:08:50,400 --> 01:08:53,640
Speaker 1: AI assistant I want, there's almost no product at all.

1372
01:08:53,720 --> 01:08:56,320
Speaker 1: It's an extremely smart model that's got full access to

1373
01:08:56,360 --> 01:08:58,920
Speaker 1: all the contexts in my life that I can just trust.

1374
01:08:58,920 --> 01:09:00,599
Speaker 1: I can say one thing to an it's going to happen.

1375
01:09:00,640 --> 01:09:02,800
Speaker 1: It'll be proactive with me only exactly what I want.

1376
01:09:03,400 --> 01:09:06,120
Speaker 1: And so when you think about like software products in

1377
01:09:06,160 --> 01:09:08,240
Speaker 1: that sense, and also wearables, they could be quite subtle

1378
01:09:08,320 --> 01:09:11,280
Speaker 1: or quite restrained. Interesting.

1379
01:09:11,720 --> 01:09:13,680
Speaker 2: The craziest thing coming down the pipeline when it comes

1380
01:09:13,720 --> 01:09:15,639
Speaker 2: to innovation that you don't think we're talking enough.

1381
01:09:15,520 --> 01:09:20,200
Speaker 1: About, uh, probably automated researchers. Like we know we're talking

1382
01:09:20,200 --> 01:09:23,680
Speaker 1: about some but like if we talked about if we

1383
01:09:23,720 --> 01:09:26,640
Speaker 1: waited every conversation in this room relative to it's that

1384
01:09:26,720 --> 01:09:29,040
Speaker 1: we had today, relative to its importance of the future,

1385
01:09:29,160 --> 01:09:31,799
Speaker 1: we probably would have talked way more about research, automated research,

1386
01:09:31,840 --> 01:09:33,040
Speaker 1: and way less about everything.

1387
01:09:32,840 --> 01:09:35,920
Speaker 2: Else you're saying. One of the most important conversations we're

1388
01:09:35,960 --> 01:09:38,439
Speaker 2: not having is about automated research, Like what what's make

1389
01:09:38,479 --> 01:09:39,160
Speaker 2: for He made now.

1390
01:09:39,080 --> 01:09:42,280
Speaker 1: Like ten years of the whole world's scientific progress in

1391
01:09:42,320 --> 01:09:45,120
Speaker 1: one year, and then one hundred years of progress in

1392
01:09:45,160 --> 01:09:48,160
Speaker 1: one year. The transformation that will have the quality of life,

1393
01:09:48,200 --> 01:09:50,680
Speaker 1: to the economy, to new risks like this is just

1394
01:09:50,800 --> 01:09:51,920
Speaker 1: there's no analog for this.

1395
01:09:52,600 --> 01:09:53,960
Speaker 2: Do you think that's a possibility.

1396
01:09:54,160 --> 01:09:55,400
Speaker 1: I think it's a possibility. Yeah.

1397
01:09:55,439 --> 01:09:58,040
Speaker 2: What has to happen for it to happen.

1398
01:09:58,240 --> 01:10:00,479
Speaker 1: The models need to get smarter, we need to eventually

1399
01:10:00,520 --> 01:10:02,479
Speaker 1: we need to figure out how to connect physical labs

1400
01:10:02,479 --> 01:10:03,479
Speaker 1: to them. Things like that.

1401
01:10:04,000 --> 01:10:05,559
Speaker 2: I want to play one thing you said, and then

1402
01:10:05,560 --> 01:10:06,439
Speaker 2: we'll wrap up.

1403
01:10:06,520 --> 01:10:07,360
Speaker 1: I'm excited for this.

1404
01:10:07,680 --> 01:10:10,160
Speaker 2: What do you think is the single most important ethical

1405
01:10:10,280 --> 01:10:12,160
Speaker 2: question we need to ask ourselves when it comes to

1406
01:10:12,200 --> 01:10:15,000
Speaker 2: the future of like us tech in humans, in.

1407
01:10:14,920 --> 01:10:17,160
Speaker 1: A world where we're going to have computers that can

1408
01:10:17,240 --> 01:10:21,040
Speaker 1: think like humans, what is the society we want to design?

1409
01:10:21,479 --> 01:10:26,080
Speaker 1: This question of what do we want the role of

1410
01:10:26,160 --> 01:10:29,040
Speaker 1: humans to be in the world, and how do we

1411
01:10:29,080 --> 01:10:31,920
Speaker 1: make sure the world is good for humans in the

1412
01:10:31,920 --> 01:10:36,519
Speaker 1: broadest sense. I think that is the biggest ethical question

1413
01:10:36,720 --> 01:10:39,479
Speaker 1: of our lifetime. Do you agree with I want nothing

1414
01:10:39,520 --> 01:10:40,519
Speaker 1: to add or change to that.

1415
01:10:40,800 --> 01:10:43,240
Speaker 2: How do you think we're doing on answering that question?

1416
01:10:43,640 --> 01:10:47,719
Speaker 1: Better than I much better than I feared we would

1417
01:10:47,720 --> 01:10:49,880
Speaker 1: be doing. You know, obviously not as well as I

1418
01:10:49,880 --> 01:10:51,439
Speaker 1: hope to what I'm always like a little too optimistic.

1419
01:10:54,680 --> 01:10:56,840
Speaker 2: I had a lot of time with Sam, but I

1420
01:10:56,920 --> 01:10:59,960
Speaker 2: have to say I still am left with so many questions.

1421
01:11:00,479 --> 01:11:03,679
Speaker 2: But maybe that's the point of it. Well, I want

1422
01:11:03,680 --> 01:11:05,599
Speaker 2: to live in the version of the world that Sam

1423
01:11:05,680 --> 01:11:08,040
Speaker 2: is hoping to build, one where we all have a shot,

1424
01:11:08,120 --> 01:11:12,040
Speaker 2: where AI equals abundance, and technology just fades into the

1425
01:11:12,080 --> 01:11:15,920
Speaker 2: backdrop of our lives, working for us and not against us.

1426
01:11:16,400 --> 01:11:19,320
Speaker 2: I don't think that's the full picture of our current reality,

1427
01:11:19,680 --> 01:11:21,960
Speaker 2: because over the last years I've spent quite a bit

1428
01:11:22,000 --> 01:11:24,960
Speaker 2: of time talking to women who were impacted by deep fakes,

1429
01:11:24,960 --> 01:11:28,000
Speaker 2: and folks worried about losing their jobs and their livelihood.

1430
01:11:28,240 --> 01:11:30,640
Speaker 2: And yeah, I actually think there's so much to be

1431
01:11:30,760 --> 01:11:35,200
Speaker 2: excited about medical advancements, scientific breakthroughs. We're just in the

1432
01:11:35,240 --> 01:11:37,760
Speaker 2: first inning of this. But I keep coming back to

1433
01:11:37,800 --> 01:11:41,160
Speaker 2: this idea that for so many years, the divide between

1434
01:11:41,240 --> 01:11:45,240
Speaker 2: Silicon Valley's optimism and the reality of the world was

1435
01:11:45,280 --> 01:11:49,439
Speaker 2: its superpower. But now, as we navigate an increasingly fractured world,

1436
01:11:49,600 --> 01:11:51,720
Speaker 2: my hope is that we can actually find a way

1437
01:11:51,760 --> 01:11:54,759
Speaker 2: to bridge that gap, that we can ground the promise

1438
01:11:54,760 --> 01:11:58,280
Speaker 2: of tomorrow's innovation with the realities of what's happening today.

1439
01:11:58,960 --> 01:12:01,439
Speaker 2: I keep going back to say prediction in twenty twenty,

1440
01:12:01,479 --> 01:12:03,519
Speaker 2: the last time we did an interview. He told me

1441
01:12:03,560 --> 01:12:06,479
Speaker 2: at that time that AI would be more transformative than

1442
01:12:06,640 --> 01:12:10,559
Speaker 2: any technology in human history. And he said, no one's

1443
01:12:10,600 --> 01:12:13,639
Speaker 2: talking about this, no one's paying attention to this. He

1444
01:12:13,720 --> 01:12:15,840
Speaker 2: was one of the early people to talk about not

1445
01:12:15,920 --> 01:12:19,120
Speaker 2: just the upside but also the potential risk. Now we're

1446
01:12:19,120 --> 01:12:21,639
Speaker 2: not just talking about it, we're actually living it. And

1447
01:12:21,800 --> 01:12:24,479
Speaker 2: my hope is that we all get the opportunity to

1448
01:12:24,560 --> 01:12:28,200
Speaker 2: participate in conversations about the future and how our world

1449
01:12:28,320 --> 01:12:31,320
Speaker 2: is shaped. The interview with Sam might be over, but

1450
01:12:31,360 --> 01:12:33,000
Speaker 2: my hope is that this is just the beginning of

1451
01:12:33,000 --> 01:12:36,360
Speaker 2: the conversation. I'm going to dig into these questions, these themes,

1452
01:12:36,360 --> 01:12:39,519
Speaker 2: these concerns, the excitement, and I'd love to ask you

1453
01:12:39,760 --> 01:12:42,479
Speaker 2: folks who are listening and watching, how do you feel

1454
01:12:42,479 --> 01:12:44,720
Speaker 2: about AI? What do you want me to cover? What

1455
01:12:44,840 --> 01:12:47,200
Speaker 2: thread should I pull on? Who should I talk to?

1456
01:12:47,840 --> 01:12:50,600
Speaker 2: I will make sure that people like Sam and the

1457
01:12:50,640 --> 01:12:56,320
Speaker 2: folks who are building the future also hear your thoughts.

1458
01:12:57,120 --> 01:12:59,679
Speaker 2: Mostly Human is a production of iHeart Podcasts and mostly

1459
01:12:59,720 --> 01:13:03,160
Speaker 2: human Media. It's produced and edited by Laurie Siegel, Lauren Hanson,

1460
01:13:03,240 --> 01:13:06,840
Speaker 2: and Nicole Bouchet. Sound design and mixing by Derek Clements.

1461
01:13:07,160 --> 01:13:11,040
Speaker 2: Additional production health from Abouzafar, Special thanks to Mark Weinhaus.

1462
01:13:11,320 --> 01:13:14,120
Speaker 2: Find us on all socials at mostly human Media. You

1463
01:13:14,160 --> 01:13:16,599
Speaker 2: can also watch mostly Human on our YouTube page. If

1464
01:13:16,640 --> 01:13:18,759
Speaker 2: you want to get in touch, email us at hello

1465
01:13:18,760 --> 01:13:21,640
Speaker 2: at mostlyguman dot com. And if you like what you're here,

1466
01:13:21,880 --> 01:13:23,840
Speaker 2: please rate and review the show and share it with

1467
01:13:23,880 --> 01:13:25,160
Speaker 2: your friends. See you next week.