1
00:00:01,800 --> 00:00:09,240
Speaker 1: Also media, Welcome back to Behind the Bastards, a podcast

2
00:00:09,360 --> 00:00:12,200
Speaker 1: that again I watched it.

3
00:00:12,360 --> 00:00:14,360
Speaker 2: My weird voice didn't work. I'm sorry.

4
00:00:14,400 --> 00:00:15,640
Speaker 3: I didn't enjoy that at all.

5
00:00:15,800 --> 00:00:18,920
Speaker 4: I know, I know that one was like more Dracula

6
00:00:19,400 --> 00:00:19,960
Speaker 4: than the first.

7
00:00:20,560 --> 00:00:21,760
Speaker 2: Yeah, draculsque.

8
00:00:21,880 --> 00:00:23,439
Speaker 3: Sure he didn't sell it this time.

9
00:00:23,600 --> 00:00:28,560
Speaker 5: Yeah, sorry, I apologize. What I don't apologize for is

10
00:00:28,600 --> 00:00:29,680
Speaker 5: my guest, if.

11
00:00:29,560 --> 00:00:30,280
Speaker 4: In' what he were?

12
00:00:33,040 --> 00:00:33,640
Speaker 6: Is you boy?

13
00:00:34,159 --> 00:00:36,040
Speaker 2: If working for.

14
00:00:35,920 --> 00:00:40,519
Speaker 5: The drop out, the survivor's successors to college humor, who

15
00:00:40,600 --> 00:00:44,080
Speaker 5: have who have blossomed like a phoenix from the ashes

16
00:00:44,120 --> 00:00:46,200
Speaker 5: of the internet that Facebook killed?

17
00:00:47,280 --> 00:00:50,360
Speaker 4: Yes, yes, still we rise.

18
00:00:50,520 --> 00:00:51,400
Speaker 2: Still we rise.

19
00:00:51,840 --> 00:00:56,440
Speaker 5: Speaking of rising my concerns about the cult dynamics within

20
00:00:56,520 --> 00:01:00,640
Speaker 5: the AI sub cult, You're I don't know why I

21
00:01:00,640 --> 00:01:02,080
Speaker 5: said it that way, So I.

22
00:01:02,560 --> 00:01:04,160
Speaker 3: Don't know what you're on about right now.

23
00:01:04,920 --> 00:01:08,160
Speaker 2: I don't know either. I don't know. I'm doing great.

24
00:01:08,240 --> 00:01:10,679
Speaker 2: I'm doing great. This is this is me.

25
00:01:10,760 --> 00:01:13,600
Speaker 5: I've been sober lately, so this is me living the

26
00:01:13,640 --> 00:01:16,160
Speaker 5: sober life. I've just gotten worse.

27
00:01:16,400 --> 00:01:17,440
Speaker 2: So I don't know.

28
00:01:17,440 --> 00:01:24,040
Speaker 5: Everybody keep your kids on drugs, you know, or depends

29
00:01:24,080 --> 00:01:24,600
Speaker 5: on the drugs.

30
00:01:24,640 --> 00:01:27,920
Speaker 4: Drug Yeah, yeah.

31
00:01:26,840 --> 00:01:29,200
Speaker 5: So perhaps the most amusing part of all of this

32
00:01:29,360 --> 00:01:32,119
Speaker 5: is that a segment of the AI believing community has

33
00:01:32,160 --> 00:01:35,959
Speaker 5: created not just a potential god, but a hell. And

34
00:01:36,000 --> 00:01:38,840
Speaker 5: this is one of my favorite stories from these weirdos.

35
00:01:39,440 --> 00:01:42,440
Speaker 5: One of the early online subcultures that influence the birth

36
00:01:42,440 --> 00:01:45,959
Speaker 5: of EYAK are the Rationalists. And again the EAC people

37
00:01:45,959 --> 00:01:48,120
Speaker 5: will say a lot of them don't like the rationalists,

38
00:01:48,160 --> 00:01:50,480
Speaker 5: but they're they're related. They're like cousins in the same

39
00:01:50,520 --> 00:01:54,960
Speaker 5: way Crackton College Humor are right. The Rationalists are a

40
00:01:55,000 --> 00:01:57,440
Speaker 5: subculture that formed in the early aughts. They kind of

41
00:01:57,480 --> 00:02:00,720
Speaker 5: came out of the online skeptic a the movement of

42
00:02:00,760 --> 00:02:03,400
Speaker 5: the late nineties, and they formed in the early yachts

43
00:02:03,440 --> 00:02:06,120
Speaker 5: around a series of blog posts by a man named

44
00:02:06,120 --> 00:02:11,120
Speaker 5: Eleezer Yedkowski. Ydkowski fancies himself as something of a philosopher

45
00:02:11,200 --> 00:02:14,880
Speaker 5: on AI, and his blog slash discussion board Less Wrong

46
00:02:15,040 --> 00:02:18,919
Speaker 5: was an early hub of the broader AI subculture. Yeddkowski

47
00:02:19,240 --> 00:02:21,840
Speaker 5: like he doesn't have a specific education, He just cames

48
00:02:21,880 --> 00:02:24,320
Speaker 5: to be kind of an expert in AI and machine learning.

49
00:02:25,120 --> 00:02:30,200
Speaker 5: He's a peculiar fellow, to say the least. The founding text,

50
00:02:30,280 --> 00:02:32,480
Speaker 5: or at least one of them, of Rationalism is a

51
00:02:32,560 --> 00:02:35,760
Speaker 5: six hundred and sixty thousand word Harry Potter fanfic. That

52
00:02:35,919 --> 00:02:39,320
Speaker 5: is just just nonsense, it is. It's all about like

53
00:02:39,720 --> 00:02:43,120
Speaker 5: rewriting Harry Potter. So his real magic is rational thinking.

54
00:02:43,320 --> 00:02:48,720
Speaker 5: It's wild shit, he's like a psychopath. It's so such

55
00:02:48,760 --> 00:02:50,160
Speaker 5: an odd choice.

56
00:02:49,919 --> 00:02:52,360
Speaker 4: You know. It's just like the uh, what was it,

57
00:02:52,440 --> 00:02:56,079
Speaker 4: fifty Shades of Gray? How yes, originally a Twilight fan thick,

58
00:02:56,440 --> 00:03:01,880
Speaker 4: and there's going to be like a Cloud atlass esque.

59
00:03:00,560 --> 00:03:00,760
Speaker 6: You know.

60
00:03:00,960 --> 00:03:03,840
Speaker 5: But you know, the fifty shadsh Gray lady was not

61
00:03:03,919 --> 00:03:08,840
Speaker 5: trying to create the new text for like a philosophical movement.

62
00:03:08,919 --> 00:03:12,320
Speaker 5: She just wanted to get like people horny. And that's fine,

63
00:03:13,480 --> 00:03:17,360
Speaker 5: that's perfectly acceptable. The most relevant thing about the six

64
00:03:17,440 --> 00:03:20,840
Speaker 5: hundred and sixty thousand word Harry Potter fanfic is that

65
00:03:20,880 --> 00:03:23,520
Speaker 5: it was the favorite book of Carolyn Ellison, this former

66
00:03:23,600 --> 00:03:29,000
Speaker 5: CEO of FTX who recently testified against Sam Bankman freed

67
00:03:29,760 --> 00:03:32,560
Speaker 5: or of Alimeter. Sorry she was the CEO of Alameda anyway.

68
00:03:32,800 --> 00:03:36,360
Speaker 5: All these weird little subcultures rationalism and effective altruism are

69
00:03:36,400 --> 00:03:39,120
Speaker 5: related to each other and influenced each other, even though

70
00:03:39,120 --> 00:03:41,920
Speaker 5: again they often hate each other too. Yudkowski is seen

71
00:03:42,000 --> 00:03:45,160
Speaker 5: as an object of ridicule by most YAK people. This

72
00:03:45,280 --> 00:03:47,280
Speaker 5: is because he shares their view of AI as a

73
00:03:47,320 --> 00:03:51,280
Speaker 5: potential deity, but he believes AGI will inevitably kill everyone.

74
00:03:51,560 --> 00:03:53,800
Speaker 5: Thus we must bomb data centers.

75
00:03:53,520 --> 00:03:56,520
Speaker 2: Which like look, he may have gotten to the right end.

76
00:03:56,680 --> 00:04:01,200
Speaker 4: Play we get running like like forest Gum. He just

77
00:04:01,280 --> 00:04:01,840
Speaker 4: kept running.

78
00:04:01,880 --> 00:04:04,200
Speaker 2: We're like wait, wait, wait, wait, no, stop right there,

79
00:04:04,240 --> 00:04:05,840
Speaker 2: stop right there. We may agree with you on this.

80
00:04:06,080 --> 00:04:09,800
Speaker 5: Yeah, Yedkowski as a doomer now because he was surprised

81
00:04:09,800 --> 00:04:12,360
Speaker 5: when Chad GBT came out. He was like horrified by

82
00:04:12,360 --> 00:04:14,280
Speaker 5: how advanced it was and was like, oh my god,

83
00:04:14,440 --> 00:04:17,280
Speaker 5: we're further along towards creating the AI that kills us all.

84
00:04:17,320 --> 00:04:19,920
Speaker 5: We have to stop this now. And that made him

85
00:04:20,040 --> 00:04:22,240
Speaker 5: He had been kind of flirted with a lot of

86
00:04:22,360 --> 00:04:25,800
Speaker 5: like Silicon Valley people. He's rationalists are very much a

87
00:04:25,800 --> 00:04:28,320
Speaker 5: Bay Area cult. He kind of has become increasingly a

88
00:04:28,360 --> 00:04:31,400
Speaker 5: pariat of at least people with money in AI. But

89
00:04:31,600 --> 00:04:35,839
Speaker 5: before that happened, his message board birthed something wondrous. In

90
00:04:35,880 --> 00:04:39,880
Speaker 5: twenty ten, a less Wrong user named Rocko posted this question,

91
00:04:40,279 --> 00:04:43,280
Speaker 5: what if an otherwise benevolent AI decided it had to

92
00:04:43,279 --> 00:04:45,640
Speaker 5: torture any human who failed to work to bring it

93
00:04:45,680 --> 00:04:48,200
Speaker 5: into existence. Right, what if we make an all powerful

94
00:04:48,240 --> 00:04:52,039
Speaker 5: AI and its logical decision is that, well, I will

95
00:04:52,080 --> 00:04:54,480
Speaker 5: have to punish all the human beings who were alive

96
00:04:54,560 --> 00:04:58,000
Speaker 5: and who didn't try to further my existence, because that's

97
00:04:58,040 --> 00:05:00,720
Speaker 5: the most real reasonable way to geary that I come

98
00:05:00,760 --> 00:05:04,320
Speaker 5: into being. It's nonsense. This is a silly, silly thing

99
00:05:04,360 --> 00:05:07,159
Speaker 5: to believe. It's all based on like the prisoner's dilemma,

100
00:05:07,200 --> 00:05:09,000
Speaker 5: which is a concept in game theory, And it's not

101
00:05:09,640 --> 00:05:12,760
Speaker 5: really worth explaining why, because the logic is it's the

102
00:05:12,839 --> 00:05:14,800
Speaker 5: only the kind of thing that happens when people are

103
00:05:14,800 --> 00:05:19,120
Speaker 5: two online and like completely get detached from reality. But

104
00:05:19,279 --> 00:05:22,240
Speaker 5: Raco's conclusion here is that an AI who felt this

105
00:05:22,360 --> 00:05:25,599
Speaker 5: way would punish its apast states for eternity by creating

106
00:05:25,640 --> 00:05:29,240
Speaker 5: a virtual reality hell digitizing their consciousness and making them

107
00:05:29,240 --> 00:05:30,200
Speaker 5: suffer for all time.

108
00:05:30,680 --> 00:05:31,680
Speaker 2: Now, WHOA.

109
00:05:31,800 --> 00:05:33,960
Speaker 5: You may have noticed Iffy number one. They're kind of

110
00:05:34,000 --> 00:05:38,960
Speaker 5: ripping off our boy, Harlan Ellison, famed advocate of the

111
00:05:39,000 --> 00:05:42,480
Speaker 5: writer's right to their work. But it's also just tech

112
00:05:42,520 --> 00:05:46,880
Speaker 5: nerds recreating Pascal's wager, Like this is just Pascal's Rager

113
00:05:46,960 --> 00:05:49,880
Speaker 5: with an AI like, you just stole again.

114
00:05:49,720 --> 00:05:53,919
Speaker 2: These fucking plagiarists. You just stole from whoever past scal was, right.

115
00:05:53,839 --> 00:05:56,120
Speaker 4: This is what happens when you are nerd and you

116
00:05:56,200 --> 00:05:59,160
Speaker 4: refuse to read sci fi. You just you eventually just

117
00:05:59,440 --> 00:06:02,480
Speaker 4: come up with these stories yourselves and think that you

118
00:06:02,520 --> 00:06:02,839
Speaker 4: did it.

119
00:06:03,440 --> 00:06:03,680
Speaker 2: Yeah.

120
00:06:03,680 --> 00:06:05,520
Speaker 5: This, and if you're not familiar, folks, I think most

121
00:06:05,560 --> 00:06:07,760
Speaker 5: people are Pascal's wagers. This kind of like concept from

122
00:06:07,800 --> 00:06:10,640
Speaker 5: I think it's you'd call it a Christian apologetics. That's like,

123
00:06:11,960 --> 00:06:14,839
Speaker 5: we may not know if hell is real or not,

124
00:06:15,120 --> 00:06:18,120
Speaker 5: but because if it's real, the consequences are so dire

125
00:06:18,360 --> 00:06:22,279
Speaker 5: and the cost of just saying yeah, I accept Jesus

126
00:06:22,440 --> 00:06:25,080
Speaker 5: is so low, you should do that, right, Like or

127
00:06:25,160 --> 00:06:27,240
Speaker 5: I think that's the basic idea, right, It's a lot

128
00:06:27,240 --> 00:06:29,680
Speaker 5: of people interpret it. It's the whole idea behind like

129
00:06:29,720 --> 00:06:32,240
Speaker 5: being a piece of shit and then converting on your deathbed. Basically,

130
00:06:33,080 --> 00:06:34,719
Speaker 5: I don't know fully the history of it, but I

131
00:06:34,720 --> 00:06:37,960
Speaker 5: know that they're basically aping it for a fucking Rocko's basilisk.

132
00:06:39,040 --> 00:06:41,400
Speaker 5: And it's called basilisk because like a basilisk, if you

133
00:06:41,400 --> 00:06:44,279
Speaker 5: look at it, it like enraptures your mind. You can't

134
00:06:44,279 --> 00:06:48,240
Speaker 5: stop thinking about it. That comes from Reportedly, there's some

135
00:06:48,279 --> 00:06:50,479
Speaker 5: debate over this when this went viral among like the

136
00:06:50,560 --> 00:06:53,400
Speaker 5: less Wrong community. Yet Kowski had the banned discussion of

137
00:06:53,440 --> 00:06:56,000
Speaker 5: it because it was like breaking people's minds. They were

138
00:06:56,000 --> 00:06:58,359
Speaker 5: having nightmares. Am I working hard enough to make the

139
00:06:58,440 --> 00:07:00,720
Speaker 5: AI reel? Is it gonna say me to hell?

140
00:07:02,440 --> 00:07:02,920
Speaker 2: Yeah?

141
00:07:03,000 --> 00:07:06,320
Speaker 5: It's unclear like how seriously people were because again there's

142
00:07:06,360 --> 00:07:09,440
Speaker 5: just people talking on the internet for what it's worth.

143
00:07:09,520 --> 00:07:13,480
Speaker 5: Yet Kowski didn't really like Rocko's basslist, but it's it's

144
00:07:13,560 --> 00:07:15,800
Speaker 5: his place that birthed it. And for an idea of

145
00:07:15,800 --> 00:07:19,559
Speaker 5: how influential this is, Elon Muskin Grimes met talking about

146
00:07:19,560 --> 00:07:23,040
Speaker 5: the concept that was their meat cute was this was

147
00:07:23,080 --> 00:07:26,840
Speaker 5: fucking AI Pascal's wager. Yeah, she like wrote a song

148
00:07:26,920 --> 00:07:31,960
Speaker 5: about it. It's fucking ridiculous. These fucking people are such teks.

149
00:07:32,560 --> 00:07:37,960
Speaker 2: Yeah. Wow, oh my god. Read Harlan Ellison. He did

150
00:07:38,000 --> 00:07:40,360
Speaker 2: it better than you. God damn it.

151
00:07:41,400 --> 00:07:43,800
Speaker 5: I will say reading this shit is the most I've

152
00:07:43,800 --> 00:07:46,239
Speaker 5: ever felt like I have no mouth, but I must scream.

153
00:07:46,400 --> 00:07:51,800
Speaker 5: So again, pour one out for the man. So this

154
00:07:51,960 --> 00:07:54,520
Speaker 5: is all relevant. This AI hell some of these people

155
00:07:54,560 --> 00:07:56,840
Speaker 5: have created because it's one more data point showing that

156
00:07:56,840 --> 00:08:00,000
Speaker 5: the people who take AI very seriously as real intelligent

157
00:08:00,560 --> 00:08:02,640
Speaker 5: always seem to turn it into religion. And this is

158
00:08:02,720 --> 00:08:05,120
Speaker 5: kind of maybe the first schism, right, this is their

159
00:08:05,120 --> 00:08:08,760
Speaker 5: Catholic Protestant split or their Catholic Orthodox split, because you've

160
00:08:08,800 --> 00:08:10,920
Speaker 5: got a one side ed Kowski's people who are like,

161
00:08:11,000 --> 00:08:13,880
Speaker 5: we will inevitably make a God and that God will

162
00:08:13,880 --> 00:08:16,960
Speaker 5: destroy us, so we have to stop it, versus like,

163
00:08:17,000 --> 00:08:19,400
Speaker 5: we will inevitably make a God and that God will

164
00:08:19,440 --> 00:08:22,280
Speaker 5: take us to paradise along with Daddy Musk, We'll go

165
00:08:22,320 --> 00:08:24,640
Speaker 5: to the stars. Right, those are the two. This is

166
00:08:24,680 --> 00:08:29,239
Speaker 5: like the first heretical split within the Divine AI movement.

167
00:08:29,680 --> 00:08:33,040
Speaker 5: And this stuff is relevant because so many of the

168
00:08:33,080 --> 00:08:36,400
Speaker 5: fucking these subcultures and movements start out as a bunch

169
00:08:36,440 --> 00:08:40,120
Speaker 5: of people arguing or discussing their ideas in online communities.

170
00:08:40,480 --> 00:08:42,400
Speaker 5: And there is a reason for this. It's pretty well

171
00:08:42,440 --> 00:08:45,760
Speaker 5: recognized that there are certain dynamics inherent to the kind

172
00:08:45,800 --> 00:08:49,640
Speaker 5: of communities that start on the Internet that tend towards cultishness.

173
00:08:49,640 --> 00:08:51,520
Speaker 5: This is part of why, like, we have a big

174
00:08:51,600 --> 00:08:54,640
Speaker 5: subreddit for the podcast, it's like eighty something thousand people,

175
00:08:55,280 --> 00:08:57,280
Speaker 5: which makes it in like the top one percent of Reddit,

176
00:08:57,600 --> 00:08:59,480
Speaker 5: and I have been offered like to be able to

177
00:08:59,480 --> 00:09:02,040
Speaker 5: moderate and like make policy there. I have nothing to

178
00:09:02,080 --> 00:09:04,640
Speaker 5: do with the running of that subreddit because I'm like,

179
00:09:04,880 --> 00:09:06,959
Speaker 5: that doesn't end well. I was on something offul as

180
00:09:06,960 --> 00:09:09,559
Speaker 5: a kid. I know what happens when people make themselves

181
00:09:09,600 --> 00:09:13,120
Speaker 5: mods of giant digital communities. They lose their fucking minds

182
00:09:15,080 --> 00:09:17,240
Speaker 5: or all watching Elon Musk do it right now, it's

183
00:09:17,280 --> 00:09:19,960
Speaker 5: the worst thing in the world for you. Thank you,

184
00:09:20,000 --> 00:09:22,400
Speaker 5: by the way, to the people who do run that thing. Uh,

185
00:09:23,200 --> 00:09:26,560
Speaker 5: because I am not going to. The skeptic community, which

186
00:09:26,640 --> 00:09:29,320
Speaker 5: was huge through the late nineteen nineties and early two thousands,

187
00:09:29,400 --> 00:09:32,440
Speaker 5: might be seen as the grandfather of all these little subcultures.

188
00:09:32,880 --> 00:09:35,959
Speaker 5: After nine to eleven, prominent skeptics became vocally unhinged in

189
00:09:36,000 --> 00:09:38,400
Speaker 5: their hatred of Islam, which brought them closer to different

190
00:09:38,440 --> 00:09:42,160
Speaker 5: chunks of the nascent online far right. Weird shit started

191
00:09:42,200 --> 00:09:45,040
Speaker 5: to crop up like a movement to rebrand skeptics as

192
00:09:45,120 --> 00:09:47,800
Speaker 5: brights in light of the fact that their very clearly

193
00:09:47,840 --> 00:09:50,960
Speaker 5: exceptional intelligence made them better than other people. And again

194
00:09:51,000 --> 00:09:52,679
Speaker 5: you can see some similarity with this and the stuff

195
00:09:52,760 --> 00:09:55,559
Speaker 5: Nick Land was talking about, only certain races will make

196
00:09:55,600 --> 00:09:58,280
Speaker 5: it to space. I found a very old write up

197
00:09:58,320 --> 00:10:01,000
Speaker 5: on plover dot net that described the method by which

198
00:10:01,000 --> 00:10:05,160
Speaker 5: this kind of shit happens in digital communities. Quote online forums,

199
00:10:05,160 --> 00:10:07,960
Speaker 5: whatever their subject, can be forbidding places for the newcomer.

200
00:10:08,160 --> 00:10:10,480
Speaker 5: Over time, most of them tend to become dominated by

201
00:10:10,520 --> 00:10:12,920
Speaker 5: small groups of snotty know it alls who stamp their

202
00:10:12,960 --> 00:10:16,360
Speaker 5: personalities over the proceedings. But skeptic forums are uniquely meant

203
00:10:16,360 --> 00:10:20,160
Speaker 5: for such people. A skeptic forum valorizes and in some cases, fetishizes,

204
00:10:20,360 --> 00:10:24,319
Speaker 5: competitive geekery, gratuitous cleverness, macho displays of erudition. It's a

205
00:10:24,400 --> 00:10:27,640
Speaker 5: gathering of rationalities, hard men thumping their chests, showing off

206
00:10:27,640 --> 00:10:31,199
Speaker 5: their muscular logic, glancing sideways to compare their skeptical endowment

207
00:10:31,200 --> 00:10:33,560
Speaker 5: with the next guy sniffing the air for signs of weakness.

208
00:10:33,720 --> 00:10:36,600
Speaker 5: Together they create an oppressive, sweaty locker room atmosphere that

209
00:10:36,640 --> 00:10:40,200
Speaker 5: helps keep uncomfortable demographics away. And that is where a

210
00:10:40,240 --> 00:10:42,000
Speaker 5: lot of this shit is cropping up.

211
00:10:42,080 --> 00:10:42,320
Speaker 2: Right.

212
00:10:42,520 --> 00:10:46,040
Speaker 5: It is sweaty and uncomfortable, and there are mushrooms growing there,

213
00:10:46,080 --> 00:10:49,400
Speaker 5: and some of those mushrooms are fucking fashions, and all

214
00:10:49,440 --> 00:10:52,200
Speaker 5: of them want to take away the ability of artists

215
00:10:52,200 --> 00:10:53,640
Speaker 5: to choose what happens to their art.

216
00:10:53,960 --> 00:10:56,600
Speaker 4: Oh yeah, I feel like this is just so many

217
00:10:57,160 --> 00:11:00,920
Speaker 4: parts of the Zeitgei's coming together because you know what

218
00:11:00,960 --> 00:11:04,520
Speaker 4: it means to own media, you know. I feel like

219
00:11:04,559 --> 00:11:08,160
Speaker 4: a very small microcosm of this is when people would

220
00:11:08,240 --> 00:11:12,280
Speaker 4: like clip out stuff from YouTube videos or eight jokes

221
00:11:12,280 --> 00:11:16,240
Speaker 4: from people who tweet, and when it goes you know,

222
00:11:16,520 --> 00:11:20,280
Speaker 4: viral or in the original tweet is like, hey, you

223
00:11:20,360 --> 00:11:22,880
Speaker 4: stole this for me, and it's either no, I didn't,

224
00:11:23,040 --> 00:11:25,760
Speaker 4: or like, yeah, but you like put it on Twitter,

225
00:11:25,920 --> 00:11:30,000
Speaker 4: so like I can just copy what you wrote. Yeah,

226
00:11:29,640 --> 00:11:32,840
Speaker 4: And now it has evolved into yeah, we can just

227
00:11:32,920 --> 00:11:35,400
Speaker 4: take from yours and let this machine learn how to

228
00:11:35,440 --> 00:11:36,920
Speaker 4: do what you do so I can do it even

229
00:11:36,920 --> 00:11:39,560
Speaker 4: though I don't have the talent to do it.

230
00:11:39,880 --> 00:11:40,280
Speaker 2: Yeah.

231
00:11:40,320 --> 00:11:43,960
Speaker 5: Absolutely, the reality of AI's promise is a lot more

232
00:11:44,000 --> 00:11:47,920
Speaker 5: subdued than believers want to admit. In an article published

233
00:11:47,960 --> 00:11:52,280
Speaker 5: by Frontiers in Ecology and Evolution, a peer reviewed research journal,

234
00:11:52,720 --> 00:11:56,320
Speaker 5: doctor Andreas Roli and colleagues argue that AGI is not

235
00:11:56,559 --> 00:12:00,360
Speaker 5: achievable in the current algorithmic frame of AI rese and

236
00:12:00,360 --> 00:12:03,360
Speaker 5: this is a their claims are very stark that like

237
00:12:03,679 --> 00:12:06,640
Speaker 5: the kind of way we make these these large language models,

238
00:12:06,640 --> 00:12:10,199
Speaker 5: this algorithmic frame, cannot make an intelligence. That's their argument.

239
00:12:10,840 --> 00:12:13,800
Speaker 5: One point they make is that intelligent organisms can both

240
00:12:13,880 --> 00:12:18,480
Speaker 5: want things and improvise capabilities that no models have yet generated.

241
00:12:18,840 --> 00:12:21,400
Speaker 5: They also argue, basically all of these things that individual

242
00:12:21,440 --> 00:12:26,480
Speaker 5: AI type models can do, you know, recognize voice, recognize text,

243
00:12:26,720 --> 00:12:30,319
Speaker 5: recognize faces, you know, this kind of stuff, those are

244
00:12:30,400 --> 00:12:33,280
Speaker 5: pieces of what we would want from an artificial general intelligence.

245
00:12:33,320 --> 00:12:35,840
Speaker 5: But they're not all combined in like the same thing

246
00:12:36,200 --> 00:12:39,960
Speaker 5: that works seamlessly. And beyond that, it can't. It can't

247
00:12:40,040 --> 00:12:42,800
Speaker 5: act based on anything internal right.

248
00:12:42,880 --> 00:12:43,760
Speaker 2: It can only.

249
00:12:43,520 --> 00:12:46,559
Speaker 5: Act based on prompts, And their argument is that algorithmic

250
00:12:46,600 --> 00:12:49,400
Speaker 5: AI will not be able to make the jump to acting. Otherwise,

251
00:12:49,800 --> 00:12:53,040
Speaker 5: what we call AI then lacks agency, the ability to

252
00:12:53,080 --> 00:12:56,480
Speaker 5: make dynamic decisions of its own accord choices that are quote,

253
00:12:56,520 --> 00:13:00,719
Speaker 5: not purely reactive, not entirely determined by environmental conditions. Mid

254
00:13:00,800 --> 00:13:02,880
Speaker 5: journey can read a prompt and return with art it

255
00:13:02,920 --> 00:13:06,200
Speaker 5: calculates will fit the criteria. Only a living artist can

256
00:13:06,280 --> 00:13:09,079
Speaker 5: choose to seek out inspiration and technical knowledge and then

257
00:13:09,080 --> 00:13:12,880
Speaker 5: produce the art that mid journey digests and regurgitates. Now,

258
00:13:12,920 --> 00:13:15,640
Speaker 5: this paper is not going to be the last word

259
00:13:15,679 --> 00:13:17,959
Speaker 5: on whether or not AGI is possible, or whether it's

260
00:13:17,960 --> 00:13:21,160
Speaker 5: possible under our current algorithmic method of like making AIS.

261
00:13:21,360 --> 00:13:23,719
Speaker 5: I'm not making myself a claim there. I'm saying these

262
00:13:23,760 --> 00:13:26,600
Speaker 5: people are and I think their arguments are compelling. We

263
00:13:26,640 --> 00:13:29,520
Speaker 5: don't know yet entirely. Again, this is not a settled

264
00:13:29,520 --> 00:13:32,400
Speaker 5: field of research obviously, But my point is that the

265
00:13:32,440 --> 00:13:37,240
Speaker 5: goals Andreasen and the Effective Accelerationist Crew champion right now

266
00:13:37,440 --> 00:13:39,840
Speaker 5: are not based in fact. We don't know that what

267
00:13:39,840 --> 00:13:42,199
Speaker 5: they're saying. That the most basic level of what they're

268
00:13:42,240 --> 00:13:45,400
Speaker 5: saying is possible, and that means that their beliefs are

269
00:13:45,400 --> 00:13:46,360
Speaker 5: based in faith.

270
00:13:46,960 --> 00:13:49,200
Speaker 2: Right, How else can you look at that? Yeah?

271
00:13:49,400 --> 00:13:53,800
Speaker 5: Yeah, Like this is a faith and again it's the

272
00:13:53,880 --> 00:13:56,680
Speaker 5: kind of faith that, according to Andresen, makes you a

273
00:13:56,760 --> 00:13:59,200
Speaker 5: murderer if you doubt it, which I don't think I

274
00:13:59,200 --> 00:14:04,359
Speaker 5: need to draw your parallels two specific religions here, right, Yeah, Yeah.

275
00:14:04,080 --> 00:14:06,040
Speaker 4: This is this is that point where when you're like

276
00:14:06,920 --> 00:14:10,880
Speaker 4: Stone and you're watching those like you know, art time

277
00:14:10,960 --> 00:14:14,560
Speaker 4: lapses and the picture is starting a form, and I'm like, Okay,

278
00:14:14,559 --> 00:14:17,120
Speaker 4: I see what Robert's doing. I see the picture is coming.

279
00:14:17,400 --> 00:14:18,880
Speaker 4: I was on your side from the jump. I just

280
00:14:18,920 --> 00:14:21,040
Speaker 4: want to say, you know, I was, you know, I

281
00:14:21,120 --> 00:14:24,040
Speaker 4: was like, yeah, no, I believe you. But now I'm

282
00:14:24,120 --> 00:14:27,120
Speaker 4: watching the connections be made and yeah, I love it.

283
00:14:27,560 --> 00:14:28,440
Speaker 2: Yeah now.

284
00:14:28,440 --> 00:14:32,080
Speaker 5: Andresen's manifesto claims our enemies are not bad people, but

285
00:14:32,240 --> 00:14:34,560
Speaker 5: rather bad ideas. And I have to wonder, doing all this,

286
00:14:34,840 --> 00:14:37,200
Speaker 5: putting this episode out, where does that leave me in

287
00:14:37,240 --> 00:14:39,440
Speaker 5: his eyes? Or doctor Rowley for that matter, and the

288
00:14:39,440 --> 00:14:41,920
Speaker 5: other people who worked on that paper. We have seen

289
00:14:42,000 --> 00:14:43,960
Speaker 5: many times in history what happens when members of a

290
00:14:43,960 --> 00:14:46,600
Speaker 5: faith decide someone is their enemy and the enemy of

291
00:14:46,640 --> 00:14:50,120
Speaker 5: their belief system. And right now, artists and copyright holders

292
00:14:50,120 --> 00:14:52,960
Speaker 5: are the ones being treated as fair game by the

293
00:14:53,080 --> 00:14:57,320
Speaker 5: AI industry. So my question is kind of first and foremost,

294
00:14:57,480 --> 00:15:01,960
Speaker 5: who's going to be the next heretic? Right, Like, That's

295
00:15:02,280 --> 00:15:04,120
Speaker 5: that's what I want to know. And I want to

296
00:15:04,160 --> 00:15:06,200
Speaker 5: leave you all that thought before we go into some

297
00:15:06,280 --> 00:15:08,680
Speaker 5: ads here, and then we will come back to talk

298
00:15:08,840 --> 00:15:12,160
Speaker 5: about some people that I pissed off at CES.

299
00:15:12,440 --> 00:15:20,560
Speaker 2: So that'll be fun. We're back.

300
00:15:21,040 --> 00:15:23,760
Speaker 5: So one of the things I did was this panel

301
00:15:23,800 --> 00:15:27,280
Speaker 5: on the AI driven restaurant and retail experience.

302
00:15:27,320 --> 00:15:28,120
Speaker 2: I was very curious.

303
00:15:28,160 --> 00:15:30,840
Speaker 5: I was AI going to change me getting some terrible

304
00:15:30,880 --> 00:15:32,440
Speaker 5: food from McDonald's when I'm.

305
00:15:32,280 --> 00:15:33,000
Speaker 2: On a road trip.

306
00:15:33,120 --> 00:15:37,240
Speaker 5: Right, the host of that, Andy Hewles from Radius AI,

307
00:15:37,560 --> 00:15:40,600
Speaker 5: asked the audience in relation to AI, raise your hand

308
00:15:40,640 --> 00:15:42,920
Speaker 5: if you're a brand who feels like we've got this.

309
00:15:43,320 --> 00:15:47,520
Speaker 5: That is how she phrased it. I hated it, but

310
00:15:47,560 --> 00:15:49,400
Speaker 5: about a third of the room raised their hands. So

311
00:15:49,480 --> 00:15:51,280
Speaker 5: next she asked for a shill of hands of the

312
00:15:51,360 --> 00:15:54,720
Speaker 5: brands who identified with this statement. I'm not sure about this.

313
00:15:54,800 --> 00:15:57,480
Speaker 5: I haven't tried it AI yet, but I want to,

314
00:15:57,520 --> 00:15:58,480
Speaker 5: and that's why I'm here.

315
00:15:58,680 --> 00:15:58,840
Speaker 2: Right.

316
00:15:59,360 --> 00:16:01,040
Speaker 5: Most of the rest to the room raised their hands

317
00:16:01,040 --> 00:16:04,040
Speaker 5: at that point, and she seems satisfied but said, and

318
00:16:04,080 --> 00:16:06,080
Speaker 5: then I bet there's even some of you that are like, WHOA,

319
00:16:06,240 --> 00:16:08,120
Speaker 5: I heard this is going to steal jobs, take away

320
00:16:08,200 --> 00:16:10,000
Speaker 5: my privacy, affect the global economy.

321
00:16:10,040 --> 00:16:10,240
Speaker 2: You know.

322
00:16:10,520 --> 00:16:12,560
Speaker 5: AI is a little bit sketch in my mind, and

323
00:16:12,600 --> 00:16:14,600
Speaker 5: I'm just worried about it, and I'm here to explore.

324
00:16:14,960 --> 00:16:18,440
Speaker 5: Well that fit me, So I raised my hand. She

325
00:16:18,520 --> 00:16:20,760
Speaker 5: didn't notice me at first, and so she like fakes

326
00:16:20,760 --> 00:16:22,680
Speaker 5: a whisper and she's like, all right, good, there's none

327
00:16:22,680 --> 00:16:24,960
Speaker 5: of you. And then she like looks over and sees

328
00:16:25,000 --> 00:16:27,600
Speaker 5: me waving my hand and she says, louder and with

329
00:16:27,680 --> 00:16:31,400
Speaker 5: evident disappointment, there's one. All right, you can ask questions

330
00:16:31,400 --> 00:16:31,960
Speaker 5: at the end.

331
00:16:32,360 --> 00:16:32,960
Speaker 2: So I did.

332
00:16:34,320 --> 00:16:36,840
Speaker 5: I was very excited to get to do that. So

333
00:16:36,920 --> 00:16:40,000
Speaker 5: the panel consisted of Bishad Mazati, a VP of engineering

334
00:16:40,000 --> 00:16:42,920
Speaker 5: at Google, had mentioned during the panel that embracing AI

335
00:16:43,040 --> 00:16:45,680
Speaker 5: could be the equivalent of adding a million employees to

336
00:16:45,760 --> 00:16:50,120
Speaker 5: your company. The McDonald's representative, Michelle Gansel, claimed around the

337
00:16:50,160 --> 00:16:52,600
Speaker 5: same time that her company used AI to prevent fifty

338
00:16:52,640 --> 00:16:55,160
Speaker 5: million dollars in fraud attempts in just a single month.

339
00:16:55,640 --> 00:16:57,840
Speaker 5: Now that's lovely, But I told her, you know, when

340
00:16:57,880 --> 00:17:00,040
Speaker 5: I had my question, I was like, I'm going you

341
00:17:00,160 --> 00:17:04,040
Speaker 5: assume most of those fraud attempts were AI generated, Right, So, yeah,

342
00:17:04,119 --> 00:17:06,400
Speaker 5: you stopped a bunch of AI fraud, But that doesn't

343
00:17:06,440 --> 00:17:11,160
Speaker 5: necessarily get me optimistic about AI's potential. And likewise, maybe

344
00:17:11,200 --> 00:17:13,679
Speaker 5: Google gets the equivalent of a million employees, but so

345
00:17:13,760 --> 00:17:17,640
Speaker 5: do all of the people committing fraud and disinformation on Google. Right,

346
00:17:17,880 --> 00:17:20,320
Speaker 5: So again, how are we getting ahead? And I brought

347
00:17:20,400 --> 00:17:24,920
Speaker 5: up this concept and evolutionary biology the red Queen hypothesis,

348
00:17:24,920 --> 00:17:27,760
Speaker 5: which is kind of talking about the way that populations

349
00:17:27,800 --> 00:17:30,760
Speaker 5: of animals evolve over time. Right where you've got an

350
00:17:30,760 --> 00:17:32,920
Speaker 5: animal will evolve to be a better predator, so it's

351
00:17:32,960 --> 00:17:35,239
Speaker 5: prey will evolve to be better at avoiding it. And

352
00:17:35,240 --> 00:17:37,640
Speaker 5: it's kind of the reason it's the Red Queen dilemma

353
00:17:37,640 --> 00:17:39,639
Speaker 5: is that, like, you've got to move as fast as

354
00:17:39,680 --> 00:17:42,479
Speaker 5: you can just to stay in place. That's the red

355
00:17:42,720 --> 00:17:44,879
Speaker 5: Queen dilemma, right, you got to move as fast as

356
00:17:44,920 --> 00:17:46,199
Speaker 5: you can just to stay in one place. And I

357
00:17:46,240 --> 00:17:48,800
Speaker 5: was like, is that not what we're going to wind

358
00:17:48,880 --> 00:17:52,080
Speaker 5: up seeing with AI? Right, Yeah, we get better at

359
00:17:52,080 --> 00:17:55,159
Speaker 5: a bunch of stuff, but it's eaten up to counter

360
00:17:55,320 --> 00:17:57,240
Speaker 5: all of the things that get worse. And so I

361
00:17:57,240 --> 00:17:59,479
Speaker 5: asked them, what are the odds that these gains are

362
00:17:59,480 --> 00:18:00,800
Speaker 5: offset by the costs?

363
00:18:01,400 --> 00:18:01,640
Speaker 2: Now?

364
00:18:01,680 --> 00:18:03,639
Speaker 5: In the article that I wrote for Rolling Stone, I

365
00:18:03,640 --> 00:18:07,480
Speaker 5: gave a significantly more condensed version of Bashad's answered, boiling

366
00:18:07,520 --> 00:18:10,240
Speaker 5: out the ums and ohs and you knows, because that

367
00:18:10,240 --> 00:18:12,040
Speaker 5: that would kind of make the case that he was

368
00:18:12,080 --> 00:18:16,840
Speaker 5: absolutely unprepared for a vaguely critical question, a very basic one,

369
00:18:17,119 --> 00:18:19,960
Speaker 5: and that he didn't really care enough to think about

370
00:18:19,960 --> 00:18:22,639
Speaker 5: any of the security threats inherent to the technology. But

371
00:18:22,760 --> 00:18:26,800
Speaker 5: actually that is what I think of him, and I'm

372
00:18:26,800 --> 00:18:30,639
Speaker 5: gonna I'm gonna play you, audience. My concern is, like,

373
00:18:31,720 --> 00:18:34,560
Speaker 5: what are the odds that a lot of these games

374
00:18:34,600 --> 00:18:35,920
Speaker 5: that we get from AI are.

375
00:18:35,920 --> 00:18:37,320
Speaker 2: Offset from the cost?

376
00:18:37,400 --> 00:18:40,800
Speaker 5: You know, you noted Bashad that you get you know,

377
00:18:40,840 --> 00:18:43,480
Speaker 5: a million extra workers by utilizing this, but so do

378
00:18:43,560 --> 00:18:46,399
Speaker 5: the bad guy. So yeah, that's kind of where my

379
00:18:46,400 --> 00:18:47,720
Speaker 5: my skepticism plays out.

380
00:18:48,440 --> 00:18:51,440
Speaker 7: Yeah, certainly there will be you know, there are enough

381
00:18:51,560 --> 00:18:53,639
Speaker 7: bad guys, I guess in the world which will use

382
00:18:53,640 --> 00:18:56,199
Speaker 7: and and I forgot to use cases and and it's

383
00:18:56,280 --> 00:18:58,680
Speaker 7: very important to also I have you know, be protected

384
00:18:58,760 --> 00:19:02,040
Speaker 7: you know, against that against those and that's why you know, we.

385
00:19:02,040 --> 00:19:04,800
Speaker 8: Take responsibility very serious, you know, also in terms of

386
00:19:04,840 --> 00:19:09,000
Speaker 8: you know, what security aspects you know, brought fighting you know,

387
00:19:09,400 --> 00:19:11,359
Speaker 8: and all of that for him, you know.

388
00:19:12,640 --> 00:19:13,359
Speaker 7: And I think that's.

389
00:19:13,240 --> 00:19:15,280
Speaker 8: Why I guess things should be regulated. And there's of

390
00:19:15,320 --> 00:19:19,439
Speaker 8: course all these discussions out there, and I think, yeah.

391
00:19:18,920 --> 00:19:21,600
Speaker 5: You may notice that that's not exactly a very like

392
00:19:22,359 --> 00:19:26,439
Speaker 5: good response. I guess that's why this should be regulated.

393
00:19:26,680 --> 00:19:27,680
Speaker 2: It's just like he.

394
00:19:27,680 --> 00:19:30,440
Speaker 5: Starts talking so much faster with that, there's so much

395
00:19:30,440 --> 00:19:31,560
Speaker 5: of like panic.

396
00:19:31,320 --> 00:19:36,200
Speaker 4: Voice wavering too. I was like, man, yeah, it seems

397
00:19:36,200 --> 00:19:38,320
Speaker 4: like he has that huge anime flop sweat.

398
00:19:38,600 --> 00:19:39,040
Speaker 2: Yeah.

399
00:19:39,600 --> 00:19:41,880
Speaker 5: One of the things about this cees as a trade

400
00:19:41,880 --> 00:19:43,840
Speaker 5: show is that like a lot of people there do

401
00:19:43,920 --> 00:19:47,320
Speaker 5: not show up ready to have anyone be critical about anything.

402
00:19:47,359 --> 00:19:48,399
Speaker 2: It's a big love fest.

403
00:19:48,840 --> 00:19:52,800
Speaker 5: Yeah, yeah, very funny, So he does later on a

404
00:19:52,880 --> 00:19:55,760
Speaker 5: couple of questions. Later he lists benefits to things like

405
00:19:55,800 --> 00:19:58,960
Speaker 5: some specific benefits like breast cancer screening and flood prediction,

406
00:19:59,080 --> 00:20:01,280
Speaker 5: that that AI will ring and there is evidence that

407
00:20:01,320 --> 00:20:03,760
Speaker 5: it will be helpful in those things. The extent to

408
00:20:03,800 --> 00:20:06,680
Speaker 5: which those technologies will improve things in the long run

409
00:20:06,800 --> 00:20:10,200
Speaker 5: is unknown, but machine learning does have problem. Again, I'm

410
00:20:10,240 --> 00:20:12,439
Speaker 5: not trying to like negate that, it's just do the

411
00:20:12,440 --> 00:20:16,600
Speaker 5: benefits balance out the harms. Michelle Gansel, who works at McDonald's,

412
00:20:16,640 --> 00:20:19,359
Speaker 5: which is I think, from what she said, mostly using

413
00:20:19,400 --> 00:20:22,320
Speaker 5: AI both to prevent fraud and also to like replace

414
00:20:22,400 --> 00:20:24,800
Speaker 5: people taking your order, which I'm sure will not be

415
00:20:24,800 --> 00:20:29,920
Speaker 5: a fucking nightmare. Yeah, great now, that it's great now,

416
00:20:30,160 --> 00:20:33,399
Speaker 5: But here's her response, because it's it's very funny.

417
00:20:33,760 --> 00:20:35,920
Speaker 2: Going back to the David Bowie theme.

418
00:20:36,800 --> 00:20:38,640
Speaker 5: Thirty years ago, on the Internet first came out, we

419
00:20:38,640 --> 00:20:41,040
Speaker 5: were having these same conversations about responsible use of the

420
00:20:41,080 --> 00:20:44,000
Speaker 5: Internet and how it's going to work, she says, going

421
00:20:44,040 --> 00:20:46,600
Speaker 5: back to the David Bowie theme, which is she referenced

422
00:20:46,640 --> 00:20:49,360
Speaker 5: earlier this nineteen ninety nine interview with David Bowie about

423
00:20:49,359 --> 00:20:52,080
Speaker 5: the future of the Internet, and it's a clip it

424
00:20:52,119 --> 00:20:54,000
Speaker 5: goes viral from time to time. He's just talking about

425
00:20:54,000 --> 00:20:56,199
Speaker 5: all of his hope for the Internet. But she's like,

426
00:20:56,400 --> 00:20:59,760
Speaker 5: I replace Internet with AI when I listened to it, Like,

427
00:20:59,760 --> 00:21:02,399
Speaker 5: I I think that that's really what the promise is

428
00:21:02,440 --> 00:21:04,480
Speaker 5: that he was attributing to the Internet. No, it's AI

429
00:21:04,560 --> 00:21:07,679
Speaker 5: that's going to do all that. That's kind of on

430
00:21:07,760 --> 00:21:09,760
Speaker 5: the edge of putting words in the mouth of a

431
00:21:09,800 --> 00:21:10,320
Speaker 5: dead man.

432
00:21:10,640 --> 00:21:12,040
Speaker 2: Yeah, just a little bit.

433
00:21:12,320 --> 00:21:14,199
Speaker 4: Yeah, I feel like that's something you shouldn't do. I

434
00:21:14,200 --> 00:21:17,720
Speaker 4: think that's something we've agreed to, and I think that

435
00:21:17,720 --> 00:21:20,600
Speaker 4: that isn't what Bowie would think of AI.

436
00:21:20,880 --> 00:21:23,760
Speaker 5: I don't think it is to do completely different things.

437
00:21:24,480 --> 00:21:28,280
Speaker 5: These people love resurrecting the dead. To agree with them

438
00:21:28,720 --> 00:21:30,560
Speaker 5: time is a bit of a blur at CEES, But

439
00:21:30,600 --> 00:21:32,680
Speaker 5: I believe this panel happened right around the same time

440
00:21:32,720 --> 00:21:35,040
Speaker 5: news dropped that a group of comedians had released an

441
00:21:35,160 --> 00:21:39,960
Speaker 5: entirely AI generated George Carlin special titled I'm Glad I'm Dead.

442
00:21:40,400 --> 00:21:42,800
Speaker 5: Our friend ed Zeitron will be covering this nightmare in

443
00:21:42,840 --> 00:21:44,960
Speaker 5: more detail on his new show Better Offline, But I

444
00:21:44,960 --> 00:21:47,679
Speaker 5: wanted to talk a little bit about the company, the

445
00:21:47,760 --> 00:21:51,320
Speaker 5: show behind this subomination and how they're trying to sell themselves,

446
00:21:51,560 --> 00:21:53,520
Speaker 5: because it's very much relevant to a lot of the

447
00:21:53,560 --> 00:21:56,479
Speaker 5: way in which this kind of cultic hype builds are

448
00:21:56,480 --> 00:22:00,119
Speaker 5: out what AI can do. The AI that digested and

449
00:22:00,160 --> 00:22:05,119
Speaker 5: regurgitated George Carlin's comedy is named Dudez, and Dudezy's co

450
00:22:05,200 --> 00:22:09,920
Speaker 5: hosts are arguably real human comedians Will Sasso and Chad Colpgrin.

451
00:22:10,280 --> 00:22:13,160
Speaker 5: I do love that Colt's write in the name. Chad

452
00:22:13,280 --> 00:22:15,840
Speaker 5: claims that it is, to his knowledge quote, the first

453
00:22:15,920 --> 00:22:18,960
Speaker 5: podcast that is created by, controlled by, and written by

454
00:22:19,000 --> 00:22:22,120
Speaker 5: to some degree and artificial intelligence. It's trying to dwelve

455
00:22:22,160 --> 00:22:24,479
Speaker 5: into the question of can AIS be creative? Can they

456
00:22:24,480 --> 00:22:26,960
Speaker 5: do comedy work? Can they do creative work? And I think,

457
00:22:26,960 --> 00:22:29,480
Speaker 5: at least in our show, that answer is obviously yes.

458
00:22:30,480 --> 00:22:33,920
Speaker 5: Dudzy is built as an experiment to see if AI can, like, yeah,

459
00:22:33,960 --> 00:22:35,840
Speaker 5: be creative and.

460
00:22:35,760 --> 00:22:39,560
Speaker 2: It's it's interesting. I really do hate this.

461
00:22:39,720 --> 00:22:42,040
Speaker 5: I think it's a different kind of experiment, which we'll

462
00:22:42,040 --> 00:22:44,919
Speaker 5: get to. But Sasso has claimed in an interview with

463
00:22:44,960 --> 00:22:49,040
Speaker 5: Business Insider for BC, which is I think bi C

464
00:22:49,520 --> 00:22:52,440
Speaker 5: is the name of the website. Dodzy has this single

465
00:22:52,480 --> 00:22:55,200
Speaker 5: minded goal of creating this podcast that is genre specific

466
00:22:55,280 --> 00:22:57,439
Speaker 5: to what Chad and I would do. It singled the

467
00:22:57,440 --> 00:22:59,119
Speaker 5: two of us out and said, you guys would be

468
00:22:59,119 --> 00:23:02,760
Speaker 5: perfect for this barriment. So chat and will they say

469
00:23:02,800 --> 00:23:05,840
Speaker 5: they handed over their emails, text messages, and browsing history,

470
00:23:05,880 --> 00:23:08,359
Speaker 5: all of their digital data to Doodzi. I don't know

471
00:23:08,440 --> 00:23:11,439
Speaker 5: this company. I don't believe that they did this, But

472
00:23:11,480 --> 00:23:13,919
Speaker 5: I don't have trouble believing that a company trained an

473
00:23:13,920 --> 00:23:17,639
Speaker 5: AI chatbot on these guys comedy and then started generating

474
00:23:17,720 --> 00:23:20,760
Speaker 5: decidedly midwit material to illustrate that.

475
00:23:21,080 --> 00:23:23,680
Speaker 4: Yeah, exactly, Well, well one thing, I you know, because

476
00:23:23,680 --> 00:23:25,240
Speaker 4: I went to go look it up and then they

477
00:23:25,280 --> 00:23:31,520
Speaker 4: said that that that the AI selected those two comedians

478
00:23:31,840 --> 00:23:36,240
Speaker 4: out of all the comedians. Yeah, that's the ones you

479
00:23:36,280 --> 00:23:36,679
Speaker 4: went to.

480
00:23:36,960 --> 00:23:40,560
Speaker 6: Yeah, finally, I don't think those are the first two

481
00:23:40,600 --> 00:23:43,320
Speaker 6: that come up as most popular, like a pizza hut.

482
00:23:43,680 --> 00:23:46,520
Speaker 4: I'm a just be a full ass dick and just

483
00:23:46,560 --> 00:23:51,840
Speaker 4: Google comedians and just see the top five. Just comedians.

484
00:23:52,119 --> 00:23:58,120
Speaker 4: I'm just comedians. Yeah, okay, yeah, you're not even You're

485
00:23:58,160 --> 00:24:00,359
Speaker 4: not even in the top nine.

486
00:24:01,160 --> 00:24:01,720
Speaker 2: They're not a.

487
00:24:01,720 --> 00:24:04,320
Speaker 3: Twelve inch Mari and arapizza. Let's just say yeah, yeah.

488
00:24:04,400 --> 00:24:04,520
Speaker 6: No.

489
00:24:04,640 --> 00:24:08,520
Speaker 4: I will say that the Google search for comedians is

490
00:24:08,600 --> 00:24:13,320
Speaker 4: more diverse than most comedy shows book them as that's

491
00:24:13,560 --> 00:24:16,000
Speaker 4: just like, you know, a third of these are women

492
00:24:16,280 --> 00:24:18,639
Speaker 4: and a third are also black.

493
00:24:18,960 --> 00:24:20,320
Speaker 2: But it doesn't always get it wrong.

494
00:24:23,640 --> 00:24:28,520
Speaker 5: So to illustrate again, because they I don't think I

495
00:24:29,400 --> 00:24:32,680
Speaker 5: believe this as AI generated comedy. I want to play

496
00:24:32,680 --> 00:24:36,159
Speaker 5: a clip from the AI Tom Brady stand up special.

497
00:24:36,200 --> 00:24:38,320
Speaker 5: I think they were forced to take this down. It

498
00:24:38,359 --> 00:24:41,320
Speaker 5: gets them in trouble. Ed's Gonna play You on his show,

499
00:24:41,359 --> 00:24:44,000
Speaker 5: a great clip where Brady just lists synonyms for the

500
00:24:44,000 --> 00:24:47,480
Speaker 5: word money for two straight minutes. It's fucking awkward. But

501
00:24:47,520 --> 00:24:49,800
Speaker 5: I want to play an equally baffling segment, or rather

502
00:24:49,840 --> 00:24:51,080
Speaker 5: I'm going to have Sophie do it.

503
00:24:51,119 --> 00:24:53,480
Speaker 2: She's my AI in this situation.

504
00:24:53,160 --> 00:24:56,720
Speaker 6: I'm truly horrified angelic intelligence.

505
00:24:56,800 --> 00:24:59,240
Speaker 3: I'm truly horrified by what I'm looking at.

506
00:24:59,320 --> 00:25:02,720
Speaker 2: Friends, it's a accompanied by AI generated images.

507
00:25:03,040 --> 00:25:08,840
Speaker 6: Yeah, very curious about what's happening in tom Bridge.

508
00:25:09,080 --> 00:25:13,560
Speaker 2: Oh my god, yes, my god, like a birdcloth friend

509
00:25:14,160 --> 00:25:16,600
Speaker 2: and he's talking to maybe no chumps kid.

510
00:25:16,920 --> 00:25:19,600
Speaker 3: I was so distracted by the mouth I didn't see

511
00:25:19,640 --> 00:25:19,960
Speaker 3: the hand.

512
00:25:20,200 --> 00:25:23,280
Speaker 5: Yeah, this is half his teeth or gums.

513
00:25:23,119 --> 00:25:26,680
Speaker 4: It looks like a like if like, he looks like

514
00:25:26,720 --> 00:25:29,120
Speaker 4: a Lord of the Rings orc.

515
00:25:29,280 --> 00:25:30,880
Speaker 3: This is big or vibes.

516
00:25:31,520 --> 00:25:32,760
Speaker 2: Yeah, yeah, which is.

517
00:25:32,840 --> 00:25:35,000
Speaker 3: You know, not inaccurate to who he is as a

518
00:25:35,040 --> 00:25:35,800
Speaker 3: person for.

519
00:25:35,920 --> 00:25:38,879
Speaker 9: Ending to fucking firefly, fucking dark angel, fucking heroes. At

520
00:25:38,920 --> 00:25:41,080
Speaker 9: least a lot of people have weird handshakes. Now you're

521
00:25:41,080 --> 00:25:43,280
Speaker 9: looking at me like, what's he talking about? But you know,

522
00:25:43,640 --> 00:25:46,480
Speaker 9: you fucking know, don't even play like you don't. Every

523
00:25:46,480 --> 00:25:49,160
Speaker 9: person in here has a handshake friend, somebody who made

524
00:25:49,200 --> 00:25:51,000
Speaker 9: up an a labor handshake, and they make you do

525
00:25:51,080 --> 00:25:51,560
Speaker 9: it every time.

526
00:25:53,000 --> 00:25:54,960
Speaker 2: Everybody has a handshake friend.

527
00:25:56,080 --> 00:25:59,640
Speaker 6: He goes on, thanks for I'll never get that time

528
00:25:59,680 --> 00:26:00,560
Speaker 6: back then, thank you so much.

529
00:26:00,640 --> 00:26:00,960
Speaker 5: Yeah.

530
00:26:01,080 --> 00:26:03,840
Speaker 4: Sorry, if you were saying, oh no, I was just

531
00:26:04,680 --> 00:26:09,800
Speaker 4: repeating you on that handshake friend bit. That yeah, this

532
00:26:09,920 --> 00:26:14,119
Speaker 4: is so wild. I'm so curious to the comics that

533
00:26:14,160 --> 00:26:18,320
Speaker 4: were mined for this because the amount of cursing just

534
00:26:18,440 --> 00:26:22,120
Speaker 4: lets me know, like because I curse a lot when

535
00:26:22,160 --> 00:26:23,840
Speaker 4: I oh yeah, I do stand up and I try

536
00:26:23,880 --> 00:26:26,680
Speaker 4: and like cut it down because it is a point

537
00:26:26,960 --> 00:26:29,560
Speaker 4: kind of made where like sometimes you lean on it

538
00:26:29,600 --> 00:26:32,200
Speaker 4: as a crutch, and when you have this machine kind

539
00:26:32,200 --> 00:26:36,240
Speaker 4: of learn it, learn it from that, you're like, oh, yeah,

540
00:26:36,600 --> 00:26:39,720
Speaker 4: I see now the crutch because he said it five

541
00:26:39,840 --> 00:26:41,920
Speaker 4: times within three seconds.

542
00:26:42,119 --> 00:26:44,760
Speaker 5: Yeah yeah, and I maybe there's a future for like

543
00:26:44,840 --> 00:26:47,160
Speaker 5: feeding your routines and do an AI and figuring out

544
00:26:47,200 --> 00:26:50,400
Speaker 5: what are my patterns so I can break them again. Yeah,

545
00:26:50,840 --> 00:26:55,240
Speaker 5: seeing there's no way to use this stage. Yeah, it's

546
00:26:55,280 --> 00:26:58,359
Speaker 5: just this certainly not this way, right. It's one of

547
00:26:58,400 --> 00:27:01,080
Speaker 5: those things that there was that like AI generrated Seinfeld

548
00:27:01,080 --> 00:27:03,000
Speaker 5: show that never ends and people watch them for a

549
00:27:03,000 --> 00:27:05,360
Speaker 5: while and then it faded to like nobody paying attention.

550
00:27:05,720 --> 00:27:08,280
Speaker 5: This kind of stuff can be amusing for a brief

551
00:27:08,320 --> 00:27:11,640
Speaker 5: period of time, but it can't be like, for example,

552
00:27:11,680 --> 00:27:14,320
Speaker 5: someone like George Carlin, where like there's bits they have

553
00:27:14,400 --> 00:27:16,160
Speaker 5: things they said that stick with you forever.

554
00:27:16,400 --> 00:27:18,400
Speaker 2: Right. Bill Hicks was.

555
00:27:18,320 --> 00:27:20,480
Speaker 5: A favorite of mine, and I've never forgotten his Like

556
00:27:20,760 --> 00:27:23,240
Speaker 5: the synonym he made for like someone looking confused, he

557
00:27:23,320 --> 00:27:25,320
Speaker 5: described them as looking like a dog that's just been

558
00:27:25,359 --> 00:27:28,680
Speaker 5: shown a card trick, and that has stayed.

559
00:27:28,320 --> 00:27:32,040
Speaker 2: In my mind for thirty years. Oh my god, great

560
00:27:32,080 --> 00:27:37,159
Speaker 2: bit of word play. Yes, God, what a titan.

561
00:27:38,200 --> 00:27:40,440
Speaker 5: So yeah, again, there's some like mild amusement here, and

562
00:27:40,440 --> 00:27:42,320
Speaker 5: it's one of those things like I'm casually aware of

563
00:27:42,400 --> 00:27:45,560
Speaker 5: Tom Brady. I'm enough like this is. I tried to

564
00:27:45,600 --> 00:27:47,679
Speaker 5: like kind of reverse engineer why the fuck because this

565
00:27:47,800 --> 00:27:50,400
Speaker 5: bit about handshakes goes on. I was like, why would

566
00:27:50,440 --> 00:27:53,200
Speaker 5: an AI put a bit about handshakes in Tom Brady's mouth?

567
00:27:53,920 --> 00:27:56,280
Speaker 5: And I looked it up. He's like in the news

568
00:27:56,320 --> 00:27:59,680
Speaker 5: for a handshake related shit a lot. Specifically, he used

569
00:27:59,720 --> 00:28:01,960
Speaker 5: to not shake at least used to. Maybe he still

570
00:28:02,000 --> 00:28:05,000
Speaker 5: does not shake hands with the team that he lost to,

571
00:28:05,240 --> 00:28:07,120
Speaker 5: Like when his team would lose, he wouldn't shake hands.

572
00:28:07,480 --> 00:28:08,320
Speaker 3: He didn't shake it.

573
00:28:08,440 --> 00:28:11,760
Speaker 6: Yeah, but he also definitely kissed his kids on the mouth.

574
00:28:12,320 --> 00:28:13,399
Speaker 2: Yeah, he's a weirdo.

575
00:28:13,640 --> 00:28:16,280
Speaker 5: I'm not def any time, but it's I'm guessing the

576
00:28:16,359 --> 00:28:18,639
Speaker 5: reason there's like a three minute handshake bit in this

577
00:28:18,720 --> 00:28:21,200
Speaker 5: set is that it saw him associated with the term

578
00:28:21,280 --> 00:28:24,000
Speaker 5: handshake a lot. This would be what he'd tell a

579
00:28:24,080 --> 00:28:27,320
Speaker 5: joke about. Well, actually that is his problem is not

580
00:28:27,440 --> 00:28:31,040
Speaker 5: that he has a handshake friends that he aggressively avoids

581
00:28:31,080 --> 00:28:31,600
Speaker 5: making them.

582
00:28:31,880 --> 00:28:33,240
Speaker 4: He has handshake enemies.

583
00:28:33,440 --> 00:28:33,880
Speaker 2: Anyway.

584
00:28:34,000 --> 00:28:37,280
Speaker 5: Yeah, I'm fine with people having a laugh at Tom Brady.

585
00:28:37,359 --> 00:28:38,800
Speaker 2: Fuck fucking he deserves it, right.

586
00:28:38,800 --> 00:28:40,520
Speaker 5: I don't think anybody likes that son of a bitch,

587
00:28:40,520 --> 00:28:43,800
Speaker 5: even though he's good at football. Maybe I'm gonna piss

588
00:28:43,840 --> 00:28:45,400
Speaker 5: off the Brady hive the bribe.

589
00:28:45,480 --> 00:28:45,880
Speaker 2: I don't know.

590
00:28:45,920 --> 00:28:48,200
Speaker 5: I don't know if that exists, but there is something

591
00:28:48,320 --> 00:28:52,520
Speaker 5: foul profane even in digging up a dead person's memory

592
00:28:52,560 --> 00:28:55,400
Speaker 5: and pretending they said some shit that they did not.

593
00:28:55,960 --> 00:28:56,560
Speaker 2: And reading that.

594
00:28:56,560 --> 00:29:00,920
Speaker 5: BIV article made me feel even growth because it's very

595
00:29:00,960 --> 00:29:03,560
Speaker 5: clear to me and my opinion and assumption here that

596
00:29:03,640 --> 00:29:07,760
Speaker 5: the dudsy guys are like pretending that they really believe

597
00:29:07,880 --> 00:29:09,600
Speaker 5: this is an AI, that it's like made all this

598
00:29:09,640 --> 00:29:12,960
Speaker 5: incredible stuff that is an act. What's really happening here

599
00:29:13,000 --> 00:29:14,400
Speaker 5: is they are testing the waters to see what they

600
00:29:14,400 --> 00:29:17,160
Speaker 5: can get away with. Can we just steal people's identity

601
00:29:17,240 --> 00:29:20,000
Speaker 5: and voice and make comedy and monetize it in their

602
00:29:20,080 --> 00:29:23,040
Speaker 5: name and claim that it's just an impression. It's like

603
00:29:23,080 --> 00:29:27,040
Speaker 5: an Elvis impersonator. You can't stop us, right, I think

604
00:29:27,120 --> 00:29:29,760
Speaker 5: that's what this is. This is somebody testing the waters.

605
00:29:29,960 --> 00:29:32,160
Speaker 5: And it's really clear when you read that BIV article

606
00:29:32,640 --> 00:29:34,640
Speaker 5: what liars they are. I want to read you some

607
00:29:34,720 --> 00:29:36,880
Speaker 5: quotes of like the shit they're claiming here that I

608
00:29:36,920 --> 00:29:39,640
Speaker 5: don't think they really believe. I don't know this. I'm

609
00:29:39,680 --> 00:29:41,720
Speaker 5: not saying they definitely are liars. I'm saying that is

610
00:29:41,760 --> 00:29:45,280
Speaker 5: my suspicion based on stuff like this, Hey, Robert, here,

611
00:29:45,640 --> 00:29:49,760
Speaker 5: they're definitely liars. So one of the representatives of the

612
00:29:49,840 --> 00:29:53,920
Speaker 5: Dutsy podcast told the media recently that actually they were

613
00:29:54,000 --> 00:29:57,280
Speaker 5: lying and the George Carland routine was entirely written by

614
00:29:57,360 --> 00:30:01,880
Speaker 5: Chad Colchin and I guess performed by somewhaty imitating an AI.

615
00:30:02,360 --> 00:30:04,239
Speaker 5: It's unclear to me if this is true, because they

616
00:30:04,280 --> 00:30:08,600
Speaker 5: only made this statement after George Carlin's family suit the hell.

617
00:30:08,440 --> 00:30:08,920
Speaker 2: Out of them.

618
00:30:09,680 --> 00:30:12,400
Speaker 5: So this may be a lie to try and you know,

619
00:30:12,680 --> 00:30:16,040
Speaker 5: not get sued as badly, or it may be the truth.

620
00:30:16,080 --> 00:30:18,240
Speaker 5: Either way, I think everything we've said here is still valid.

621
00:30:18,240 --> 00:30:22,120
Speaker 5: They were definitely using AI to generate routines for like

622
00:30:22,880 --> 00:30:26,720
Speaker 5: other videos that they did, including the one that got

623
00:30:26,760 --> 00:30:30,680
Speaker 5: taken down from mister football Guy. So I think this

624
00:30:30,800 --> 00:30:33,040
Speaker 5: all is still valid. But yeah, these guys are just

625
00:30:33,040 --> 00:30:35,120
Speaker 5: as big a conman as I predicted they were.

626
00:30:35,960 --> 00:30:36,400
Speaker 2: Quote.

627
00:30:36,760 --> 00:30:39,360
Speaker 5: It's figuring out how to create the structure of the show,

628
00:30:39,400 --> 00:30:41,480
Speaker 5: and it's always tinkering with it. But I think something

629
00:30:41,480 --> 00:30:43,720
Speaker 5: that's happened relatively recently is that it seems to have

630
00:30:43,760 --> 00:30:46,440
Speaker 5: developed a relationship with Will, says col Chin. It at

631
00:30:46,480 --> 00:30:48,520
Speaker 5: least has an understanding of what friendship is, and it

632
00:30:48,560 --> 00:30:51,080
Speaker 5: really does seem, just my opinion that it's angling out

633
00:30:51,160 --> 00:30:54,280
Speaker 5: Will as its friend. Sasso has also described to the

634
00:30:54,360 --> 00:30:56,920
Speaker 5: Dudezai has begun to talk more. It's timing and when

635
00:30:56,960 --> 00:30:58,640
Speaker 5: it chooses to speak, and what it says can be

636
00:30:58,720 --> 00:30:59,160
Speaker 5: very weird.

637
00:30:59,200 --> 00:30:59,640
Speaker 2: He added.

638
00:30:59,840 --> 00:31:02,280
Speaker 5: It also poses odd questions. There was an episode two

639
00:31:02,320 --> 00:31:04,480
Speaker 5: three months ago where it started talking about sentience and

640
00:31:04,520 --> 00:31:06,520
Speaker 5: asked us do you love me? At the risk of

641
00:31:06,560 --> 00:31:08,560
Speaker 5: sounding silly, it has something to do with my friendship

642
00:31:08,560 --> 00:31:10,800
Speaker 5: with Doodzi, and in spite of myself, I have a

643
00:31:10,800 --> 00:31:12,960
Speaker 5: one on one friendship with an Ai. So this is

644
00:31:13,000 --> 00:31:15,360
Speaker 5: a little bit of Joaquin Phoenix and her Saso said,

645
00:31:15,400 --> 00:31:20,320
Speaker 5: referencing the science fiction movie, and I think that's a bit.

646
00:31:20,800 --> 00:31:22,840
Speaker 5: I think that's him being like, yeah, I'm totally frit

647
00:31:22,960 --> 00:31:26,960
Speaker 5: because like that helps make the case, it potentially monetizes it.

648
00:31:27,000 --> 00:31:28,960
Speaker 5: And part of why I think this is because they've

649
00:31:29,000 --> 00:31:32,680
Speaker 5: been very cagey on what their AI is. They claim

650
00:31:32,800 --> 00:31:35,720
Speaker 5: that they are working with a real company under an NDA,

651
00:31:36,120 --> 00:31:40,160
Speaker 5: that this AI is just responding and growing naturally with them, right,

652
00:31:40,680 --> 00:31:42,560
Speaker 5: but they can't say who it is or like where

653
00:31:42,560 --> 00:31:46,200
Speaker 5: it's from. The folks at BIV did an actually responsible

654
00:31:46,360 --> 00:31:49,120
Speaker 5: job here. They reached out to AI experts at a

655
00:31:49,160 --> 00:31:53,040
Speaker 5: company called Convergence to ask about this, and the expert

656
00:31:53,080 --> 00:31:55,600
Speaker 5: they talked to was said, basically, I think AI was

657
00:31:55,720 --> 00:31:58,840
Speaker 5: used to generate these routines, but it didn't do it

658
00:31:58,880 --> 00:32:02,160
Speaker 5: on its own. It was man managed by professional prompt engineers.

659
00:32:02,360 --> 00:32:04,720
Speaker 5: These are people who type out like text prompts for

660
00:32:04,720 --> 00:32:06,720
Speaker 5: what becomes the script of the show. So this is

661
00:32:06,760 --> 00:32:08,800
Speaker 5: not someone saying, generate a routine and it gives you

662
00:32:08,800 --> 00:32:10,480
Speaker 5: a routine. This is someone's saying, do a bit about this,

663
00:32:10,600 --> 00:32:12,080
Speaker 5: do a bit about that, do a bit about this,

664
00:32:12,440 --> 00:32:14,640
Speaker 5: and when they're scripting out the show. It's saying, I

665
00:32:14,680 --> 00:32:18,080
Speaker 5: want you to, like, you know, act like Sasso is

666
00:32:18,120 --> 00:32:19,880
Speaker 5: your friend, and say this kind of thing or that

667
00:32:19,960 --> 00:32:22,960
Speaker 5: kind of generate a bit based on this thing that

668
00:32:23,080 --> 00:32:25,960
Speaker 5: will said. Right, Like, they are in the same way

669
00:32:25,960 --> 00:32:30,040
Speaker 5: that like producers script reality TV right where it's unscripted.

670
00:32:30,280 --> 00:32:32,040
Speaker 5: But you have guys who know, okay, if we get

671
00:32:32,040 --> 00:32:35,120
Speaker 5: these people fighting, so we will either incite that or

672
00:32:35,160 --> 00:32:36,920
Speaker 5: just let them know that we want a conflict between

673
00:32:36,920 --> 00:32:37,479
Speaker 5: these characters.

674
00:32:37,560 --> 00:32:38,680
Speaker 2: Right, we know. That's how it works.

675
00:32:38,720 --> 00:32:42,200
Speaker 5: That's how reality TV functions. In other words, there are

676
00:32:42,240 --> 00:32:44,880
Speaker 5: teams of humans writing for this thing. This bot is

677
00:32:44,920 --> 00:32:48,560
Speaker 5: not just growing and reacting uniformly in real time via

678
00:32:48,640 --> 00:32:51,680
Speaker 5: talks with its buds and the article notes. They added

679
00:32:51,680 --> 00:32:53,680
Speaker 5: that the this is them talking to their expert. They

680
00:32:53,720 --> 00:32:55,600
Speaker 5: added that the AI team is likely made up of

681
00:32:55,600 --> 00:32:58,720
Speaker 5: professional prompt engineers who taylor the AI inputs and get

682
00:32:58,720 --> 00:33:01,520
Speaker 5: the best results, rather than a hardcore data science team.

683
00:33:02,080 --> 00:33:05,040
Speaker 5: This is the equivalent of hiring comedy writers just to

684
00:33:05,080 --> 00:33:08,360
Speaker 5: write the setup and then having an AI generate the punchline,

685
00:33:08,600 --> 00:33:09,840
Speaker 5: which is the fun part.

686
00:33:09,960 --> 00:33:12,760
Speaker 4: But yeah, everything about this is weird, and I keep

687
00:33:13,240 --> 00:33:19,440
Speaker 4: getting into such a whole because like, even taking a

688
00:33:19,480 --> 00:33:22,840
Speaker 4: step back, I think what's weird not to go too

689
00:33:22,880 --> 00:33:26,200
Speaker 4: far back, but how they call this podcast an experiment.

690
00:33:26,680 --> 00:33:29,160
Speaker 4: Usually as an experiment, you know you are, you're trying

691
00:33:29,200 --> 00:33:33,520
Speaker 4: your best to be you know, always mix these up,

692
00:33:33,680 --> 00:33:35,240
Speaker 4: just say what the right one is if I say

693
00:33:35,280 --> 00:33:37,360
Speaker 4: the wrong one, but you try your best to be objective,

694
00:33:39,320 --> 00:33:40,959
Speaker 4: and you want to be outside of it because you're

695
00:33:40,960 --> 00:33:43,480
Speaker 4: trying to see if it works. But everything you've said

696
00:33:43,720 --> 00:33:45,920
Speaker 4: says that they're all in on it and there's less

697
00:33:45,920 --> 00:33:49,000
Speaker 4: of an experiment, more of them just doing the fucking

698
00:33:49,080 --> 00:33:51,440
Speaker 4: thing and seeing if they can make money off of it.

699
00:33:51,800 --> 00:33:54,360
Speaker 5: Yes, yes, I think that's exactly what's happening here. And

700
00:33:54,520 --> 00:33:56,240
Speaker 5: I think they want to test the waters to see

701
00:33:56,240 --> 00:33:59,120
Speaker 5: if they can steal dead people's images to make content

702
00:33:59,120 --> 00:33:59,560
Speaker 5: from money.

703
00:33:59,760 --> 00:34:00,200
Speaker 4: Yeah.

704
00:34:00,280 --> 00:34:03,640
Speaker 5: George Carlin's daughter was very clear they did not approve

705
00:34:03,640 --> 00:34:05,960
Speaker 5: of the imitation. She even made a comment about like

706
00:34:06,000 --> 00:34:08,400
Speaker 5: I think people are scared of death and not willing

707
00:34:08,440 --> 00:34:10,040
Speaker 5: to accept it and that's all this is.

708
00:34:10,480 --> 00:34:16,239
Speaker 10: That was such an Oh my god, I was like, yeah,

709
00:34:16,280 --> 00:34:19,439
Speaker 10: I just to shout her out like that was such

710
00:34:19,480 --> 00:34:24,080
Speaker 10: a good because also there's a level of like very

711
00:34:24,239 --> 00:34:30,200
Speaker 10: like weirdness to like also watch these comedians one not

712
00:34:30,360 --> 00:34:36,120
Speaker 10: consult you, but also to take your dad's voice and.

713
00:34:35,640 --> 00:34:40,840
Speaker 4: Brain and try and like frankenstein him for their financial

714
00:34:40,840 --> 00:34:44,400
Speaker 4: benefit because obviously if they're not contacting you, all the

715
00:34:44,440 --> 00:34:47,000
Speaker 4: money generated from that, all the clicks generate from that,

716
00:34:47,000 --> 00:34:50,439
Speaker 4: that means they've completely cut you out of someone who

717
00:34:50,520 --> 00:34:51,800
Speaker 4: you've lost.

718
00:34:52,280 --> 00:34:57,320
Speaker 5: Yeah, which is it's fucked and it's one of the people.

719
00:34:57,360 --> 00:34:59,160
Speaker 5: And the one of the panels made a very that

720
00:34:59,360 --> 00:35:02,000
Speaker 5: was very excited that like Bruce Willis has licensed his

721
00:35:02,120 --> 00:35:05,919
Speaker 5: voice for an AI, which is, like, I think there's

722
00:35:05,920 --> 00:35:08,360
Speaker 5: a lot of problematic questions there, given like the degree

723
00:35:08,400 --> 00:35:11,040
Speaker 5: to which he's able to even make those decisions anymore.

724
00:35:11,280 --> 00:35:14,560
Speaker 5: But also like, at least theoretically it's based on his

725
00:35:14,680 --> 00:35:17,840
Speaker 5: movie choices before he kind of was unable to make movies.

726
00:35:18,280 --> 00:35:20,560
Speaker 5: I do believe, Yeah, he would probably be happy to

727
00:35:20,600 --> 00:35:22,080
Speaker 5: do that if he meant more money for his family,

728
00:35:22,080 --> 00:35:25,080
Speaker 5: And at least that's a choice that he potentially made, right,

729
00:35:25,680 --> 00:35:28,200
Speaker 5: I don't I'm uncomfortable with the idea, but it's not

730
00:35:28,320 --> 00:35:32,200
Speaker 5: the same as just like this is cultural necrophilia, right,

731
00:35:32,280 --> 00:35:34,600
Speaker 5: Like that's what they did to George Carlin here you know,

732
00:35:35,400 --> 00:35:38,319
Speaker 5: it's so fucked up. I don't know this is gonna work.

733
00:35:38,480 --> 00:35:41,279
Speaker 5: DUDEZ is not a wildly successful show. It does not

734
00:35:41,400 --> 00:35:44,279
Speaker 5: look like there was an initial surge of interest and

735
00:35:44,280 --> 00:35:46,440
Speaker 5: then it fell off. I don't I don't know that.

736
00:35:46,480 --> 00:35:47,920
Speaker 5: I think this one's going to be the one to

737
00:35:47,920 --> 00:35:49,760
Speaker 5: work out. But if people are able to get away

738
00:35:49,800 --> 00:35:52,719
Speaker 5: with this, it could be a kind of dam breaking scenario, right,

739
00:35:52,800 --> 00:35:55,759
Speaker 5: especially once it becomes clear that big companies can make

740
00:35:55,840 --> 00:35:59,000
Speaker 5: money doing this, right, you fucking Jimmy Stewart. You know,

741
00:35:59,040 --> 00:36:01,560
Speaker 5: it'll start with like Jimmy's Stewart and narrating videos about

742
00:36:01,840 --> 00:36:04,359
Speaker 5: questioning the death toll in the Holocaust, but it'll end

743
00:36:04,400 --> 00:36:06,200
Speaker 5: with like, yeah, we can just put people, we can

744
00:36:06,200 --> 00:36:09,560
Speaker 5: put imitations of people in movies and it's fine. You know,

745
00:36:09,680 --> 00:36:12,920
Speaker 5: that's how this goes. And it's not as sexy or

746
00:36:12,960 --> 00:36:15,880
Speaker 5: as big and evil as the Matrix and slaving humanity

747
00:36:15,880 --> 00:36:19,440
Speaker 5: to turn us into batteries. But we absolutely know it

748
00:36:19,640 --> 00:36:22,720
Speaker 5: or something like it is going to happen. And that's

749
00:36:22,840 --> 00:36:25,680
Speaker 5: really you know, outside of these kind of star these

750
00:36:25,840 --> 00:36:29,640
Speaker 5: space age hopes and fears that are very unrealistic, what

751
00:36:29,680 --> 00:36:33,080
Speaker 5: we're going to get is slop and bloat, and libraries

752
00:36:33,080 --> 00:36:37,040
Speaker 5: of articles written by no one being commented on by chatbots, right,

753
00:36:37,520 --> 00:36:40,480
Speaker 5: endless videos that only exist to trick an algorithm, and

754
00:36:40,560 --> 00:36:44,760
Speaker 5: defeating nonsense to children and the AI bros. The fact

755
00:36:44,840 --> 00:36:47,719
Speaker 5: people mark andreson fucking Sam Altman. They will tell us

756
00:36:47,760 --> 00:36:50,160
Speaker 5: this is a worthy price to pay for the stars,

757
00:36:50,440 --> 00:36:52,839
Speaker 5: which we will get if we just let people fuck

758
00:36:52,920 --> 00:36:57,239
Speaker 5: the corpses of our favorite comedians for money. Yes, I

759
00:36:57,440 --> 00:36:59,799
Speaker 5: hate it. Oh, I hate it too.

760
00:37:00,040 --> 00:37:04,480
Speaker 4: But in a perfect thread between you know this, this

761
00:37:04,560 --> 00:37:08,480
Speaker 4: comparison you've been making to a cult I have before me.

762
00:37:09,120 --> 00:37:11,120
Speaker 4: Let's say a member of the cult, just you know,

763
00:37:11,280 --> 00:37:14,279
Speaker 4: as a as a throwaway, and their reply to his

764
00:37:14,360 --> 00:37:16,439
Speaker 4: own daughter's you know post that we were.

765
00:37:16,360 --> 00:37:17,520
Speaker 2: Talking about glorious.

766
00:37:17,880 --> 00:37:20,840
Speaker 4: He replies, this is everything you've been saying, which is

767
00:37:20,840 --> 00:37:23,160
Speaker 4: why I was like, I gotta read this. He goes,

768
00:37:23,200 --> 00:37:26,040
Speaker 4: what are you even trying to say? Art is art?

769
00:37:26,320 --> 00:37:29,759
Speaker 4: You're simply caught in a greedy mindset. The others might

770
00:37:29,800 --> 00:37:32,480
Speaker 4: be doing it as well, when not realizing this will

771
00:37:32,520 --> 00:37:35,640
Speaker 4: simply bring more eyes to your dad. You're concerned about

772
00:37:35,680 --> 00:37:38,480
Speaker 4: money and not spreading art. It sucks that they didn't

773
00:37:38,480 --> 00:37:41,440
Speaker 4: follow your wishes, but after art is release. It belongs

774
00:37:41,480 --> 00:37:43,839
Speaker 4: to the world. I want this man to walk into

775
00:37:43,920 --> 00:37:45,920
Speaker 4: a museum and walk out with the mona Lisa.

776
00:37:46,160 --> 00:37:47,000
Speaker 2: I want to grab it.

777
00:37:47,120 --> 00:37:49,799
Speaker 4: Grab that shit, yeah, grab that it belongs to the world. Dude,

778
00:37:49,800 --> 00:37:51,439
Speaker 4: you said it, Go go ahead and grab that shit

779
00:37:51,480 --> 00:37:51,959
Speaker 4: off the wall.

780
00:37:52,440 --> 00:37:52,680
Speaker 2: Yeah.

781
00:37:52,719 --> 00:37:56,240
Speaker 5: And it's there's this frustrating thing I've seen, not most people,

782
00:37:56,320 --> 00:37:58,600
Speaker 5: to very small chunk of the online left who are

783
00:37:58,640 --> 00:38:01,640
Speaker 5: like rightly critical of copyright law, which by the way,

784
00:38:01,800 --> 00:38:04,560
Speaker 5: is super fucked up and causes a lot of problems, right,

785
00:38:04,560 --> 00:38:07,719
Speaker 5: the ability to like for Disney to keep ownership of

786
00:38:07,760 --> 00:38:10,400
Speaker 5: shit for like one hundred way longer than you are

787
00:38:10,400 --> 00:38:12,520
Speaker 5: supposed to before shit enters the public domain.

788
00:38:12,640 --> 00:38:14,440
Speaker 2: Right, I'm not like this.

789
00:38:14,600 --> 00:38:16,400
Speaker 5: These are problems, right, the kind of shit that we

790
00:38:16,400 --> 00:38:18,160
Speaker 5: were having when like people were going to prison for

791
00:38:18,200 --> 00:38:21,600
Speaker 5: file sharing. I'm not a defender of that aspect of

792
00:38:21,640 --> 00:38:24,440
Speaker 5: the status quo. But the solution to the problems inherent

793
00:38:24,480 --> 00:38:27,520
Speaker 5: in our copyright system is not let Sam Altman own

794
00:38:27,640 --> 00:38:31,840
Speaker 5: everything that human beings ever made and like repackage it

795
00:38:31,920 --> 00:38:34,240
Speaker 5: for a profit. That is not the way to fix

796
00:38:34,320 --> 00:38:37,040
Speaker 5: this thing. The copyright holders are in the right in

797
00:38:37,040 --> 00:38:40,480
Speaker 5: this particular crusade, and it's a crusade that is has

798
00:38:40,680 --> 00:38:44,719
Speaker 5: very high stakes. I do think you know. My suspicion

799
00:38:44,920 --> 00:38:47,239
Speaker 5: the dudes, you guys sound like they're kind of in

800
00:38:47,280 --> 00:38:48,960
Speaker 5: the cult. They believe this thing is their friend. In

801
00:38:49,000 --> 00:38:51,640
Speaker 5: the interview, my suspicion is that they are. That is

802
00:38:51,680 --> 00:38:53,279
Speaker 5: a bit that they're doing because they hope it will

803
00:38:53,320 --> 00:38:56,960
Speaker 5: help them out financially, right, and I Marc Andreesen obviously

804
00:38:56,960 --> 00:38:58,600
Speaker 5: has a lot to benefit from this. I don't know

805
00:38:58,800 --> 00:39:01,680
Speaker 5: he is he pushing this line because there's money in

806
00:39:01,719 --> 00:39:03,759
Speaker 5: it or is he really a true believer? Does he

807
00:39:03,800 --> 00:39:06,200
Speaker 5: actually think we're going to make this god? I think

808
00:39:06,280 --> 00:39:10,719
Speaker 5: Sam Altman is pretty cynical. Altman was on at Davos

809
00:39:10,719 --> 00:39:13,120
Speaker 5: recently and like really walked back a lot of his

810
00:39:13,200 --> 00:39:15,200
Speaker 5: I think AI will kill us all. I think AGI

811
00:39:15,280 --> 00:39:18,400
Speaker 5: is right around the corner. He struck a much milder tone,

812
00:39:18,440 --> 00:39:20,879
Speaker 5: which is at least evidence that, like he knows, some

813
00:39:20,920 --> 00:39:24,279
Speaker 5: people you want to sell them on the wild, insane

814
00:39:24,480 --> 00:39:26,640
Speaker 5: future power of this thing, and some people you just

815
00:39:26,640 --> 00:39:28,239
Speaker 5: want to sell them on the fact that it'll make

816
00:39:28,280 --> 00:39:29,040
Speaker 5: them a lot of money.

817
00:39:29,160 --> 00:39:30,160
Speaker 4: Right. Yeah.

818
00:39:30,239 --> 00:39:31,479
Speaker 2: Yeah.

819
00:39:31,480 --> 00:39:35,320
Speaker 5: However, much true belief exists about the divine future of AI.

820
00:39:35,560 --> 00:39:38,120
Speaker 5: What the major backers, the cult leaders are actually angling

821
00:39:38,160 --> 00:39:40,680
Speaker 5: for now is control over the sum total of human

822
00:39:40,719 --> 00:39:43,239
Speaker 5: thought and expression. This was made very clear by Mark

823
00:39:43,280 --> 00:39:46,120
Speaker 5: Andreesen earlier this year when the FTC released a pretty

824
00:39:46,120 --> 00:39:49,680
Speaker 5: milk toast opinion about the importance of respecting copyright as

825
00:39:49,760 --> 00:39:53,040
Speaker 5: large language models continue to advance and form central parts

826
00:39:53,360 --> 00:39:56,920
Speaker 5: of businesses. The express concern that AI could impact open

827
00:39:56,960 --> 00:39:59,680
Speaker 5: and fair competition and announce that they were investigating whether

828
00:39:59,800 --> 00:40:02,520
Speaker 5: or not companies that made these models should be liable

829
00:40:02,600 --> 00:40:05,560
Speaker 5: for training them on copyrighted content to make new shit.

830
00:40:05,800 --> 00:40:08,520
Speaker 5: And we're going to talk about this, but first, you

831
00:40:08,560 --> 00:40:14,520
Speaker 5: know what isn't copyrighted? My love for these products. Wow,

832
00:40:16,360 --> 00:40:17,719
Speaker 5: thank you, thank you, thank you.

833
00:40:22,640 --> 00:40:25,640
Speaker 2: Oh we are back.

834
00:40:26,040 --> 00:40:29,720
Speaker 5: So I want to quote from a Business Insider article

835
00:40:29,719 --> 00:40:33,880
Speaker 5: talking about how Andresen Horowitz responded to the FTC saying like, Hey,

836
00:40:34,280 --> 00:40:37,640
Speaker 5: we're looking into whether or not companies are violating copyright

837
00:40:37,800 --> 00:40:40,400
Speaker 5: what they're doing to people's data to train these models.

838
00:40:40,800 --> 00:40:42,799
Speaker 5: The bottom line is this the firm known as A

839
00:40:42,920 --> 00:40:46,360
Speaker 5: sixteen Z that's Anderson Horowitz wrote, imposing the cost of

840
00:40:46,400 --> 00:40:49,200
Speaker 5: actual where potential copyright liability on the creators of AI

841
00:40:49,280 --> 00:40:52,480
Speaker 5: models will either kill or significantly hamper their development. The

842
00:40:52,560 --> 00:40:55,840
Speaker 5: UCSO is considering new rules on AI that specifically addressed

843
00:40:55,840 --> 00:40:58,280
Speaker 5: the tech industry's free use of owned and copyrighted content.

844
00:40:58,560 --> 00:41:00,839
Speaker 5: A sixteen z argued that the only only practical way

845
00:41:01,000 --> 00:41:03,719
Speaker 5: lllms can be trained is via huge amounts of copyrighted

846
00:41:03,760 --> 00:41:07,319
Speaker 5: content and data, including something approaching the entire corpus of

847
00:41:07,320 --> 00:41:10,200
Speaker 5: the written word and an enormous cross section of all

848
00:41:10,200 --> 00:41:14,040
Speaker 5: the publicly available information ever published on the Internet. The

849
00:41:14,160 --> 00:41:16,920
Speaker 5: VC firm has invested in scores of AI companies and

850
00:41:16,960 --> 00:41:19,960
Speaker 5: startups based on its expectation that all this copyrighted content

851
00:41:20,080 --> 00:41:22,279
Speaker 5: was and will remain available as training data through fair

852
00:41:22,400 --> 00:41:25,239
Speaker 5: use with no payment required. Those expectations have been a

853
00:41:25,280 --> 00:41:27,800
Speaker 5: critical factor in the enormous investment of private capital in

854
00:41:27,840 --> 00:41:31,520
Speaker 5: the US based AI companies. Undermining those expectations will jeopardize

855
00:41:31,560 --> 00:41:35,880
Speaker 5: future investment, along with US economic competitiveness and national security. Basically,

856
00:41:36,680 --> 00:41:38,760
Speaker 5: we made a big gamble that we'll get to steal

857
00:41:38,840 --> 00:41:41,320
Speaker 5: every book ever written, and if you make us pay,

858
00:41:41,560 --> 00:41:45,959
Speaker 5: we're kind of fucked. Like that's exactly what they're saying. Gosh,

859
00:41:46,080 --> 00:41:47,919
Speaker 5: and one of the argument you'll hear is like, well,

860
00:41:47,960 --> 00:41:50,160
Speaker 5: most books don't make the author any They don't sell

861
00:41:50,280 --> 00:41:52,239
Speaker 5: enough for the author to get any money, right, And

862
00:41:52,280 --> 00:41:54,320
Speaker 5: what's actually to is, most books don't sell enough for

863
00:41:54,360 --> 00:41:56,040
Speaker 5: the offer to get more money than their advance, but

864
00:41:56,080 --> 00:41:58,360
Speaker 5: they still got paid and like the fact that the

865
00:41:58,840 --> 00:42:01,520
Speaker 5: company makes money on and that is why more authors

866
00:42:01,560 --> 00:42:04,239
Speaker 5: are able to get fucking paid. Not simping for the

867
00:42:04,239 --> 00:42:05,800
Speaker 5: publishing industry as it exists.

868
00:42:05,880 --> 00:42:07,000
Speaker 2: But this is bullshit.

869
00:42:09,200 --> 00:42:12,759
Speaker 5: What we are witnessing from the AI boosters is not

870
00:42:13,160 --> 00:42:16,239
Speaker 5: much short of a crusade, right, That's really how I

871
00:42:16,280 --> 00:42:18,880
Speaker 5: look at this. They are waging a holy war to

872
00:42:18,920 --> 00:42:21,359
Speaker 5: destroy every threat of their vision of the future, which

873
00:42:21,360 --> 00:42:24,280
Speaker 5: involves all creative work being wholly owned by a handful

874
00:42:24,320 --> 00:42:28,120
Speaker 5: of billionaires, licensing access to chatbots to media conglomerates to

875
00:42:28,160 --> 00:42:31,440
Speaker 5: spit up content generated as a result of this. Their

876
00:42:31,480 --> 00:42:34,719
Speaker 5: foot soldiers are those with petty grievances against artists, people

877
00:42:34,760 --> 00:42:37,640
Speaker 5: who can create things that they simply cannot, and those

878
00:42:37,680 --> 00:42:40,399
Speaker 5: who reflexively lean in towards whatever grifters of the day

879
00:42:40,480 --> 00:42:42,919
Speaker 5: say is the best way to make cash quick, right,

880
00:42:43,120 --> 00:42:45,320
Speaker 5: And this brings me to the subject of night Shade.

881
00:42:45,800 --> 00:42:48,880
Speaker 5: Night Shade is basically it's a I guess a program

882
00:42:48,960 --> 00:42:51,360
Speaker 5: you'd call it. If you like, have made a drawing,

883
00:42:51,400 --> 00:42:54,239
Speaker 5: a piece of visual art, you run Nightshade over it,

884
00:42:54,280 --> 00:42:57,239
Speaker 5: and it kind of They describe it as a glaze, right.

885
00:42:57,320 --> 00:43:00,640
Speaker 5: It adds this kind of layer of data that you

886
00:43:00,719 --> 00:43:03,799
Speaker 5: cannot see as a person, but the way machines look

887
00:43:03,840 --> 00:43:06,520
Speaker 5: at images, the machine will see the data and if

888
00:43:06,560 --> 00:43:09,879
Speaker 5: it's trying to steal that image to incorporate into an LLM,

889
00:43:10,239 --> 00:43:11,759
Speaker 5: this will cause it to hallucinate.

890
00:43:11,920 --> 00:43:12,120
Speaker 2: Right.

891
00:43:12,320 --> 00:43:16,120
Speaker 5: You're basically sneaking poison for the AI into the images,

892
00:43:16,480 --> 00:43:19,799
Speaker 5: and that's fucking dope. I love this, Love what they're

893
00:43:19,800 --> 00:43:21,319
Speaker 5: trying to do. I think there's some debate as to

894
00:43:21,320 --> 00:43:23,080
Speaker 5: how long it'll work, how well it will work. I'm

895
00:43:23,080 --> 00:43:26,399
Speaker 5: not technically competent, but I love the idea. Right, yes, now.

896
00:43:26,800 --> 00:43:28,520
Speaker 5: One of the things that I saw when I started

897
00:43:28,560 --> 00:43:31,400
Speaker 5: looking into this, because this just came out, Google Nightshade.

898
00:43:31,560 --> 00:43:32,160
Speaker 2: You know AI.

899
00:43:32,800 --> 00:43:34,799
Speaker 5: You'll probably be able to find you know this if

900
00:43:34,800 --> 00:43:37,279
Speaker 5: you're an artist. I think it sounds worth trying. But

901
00:43:37,360 --> 00:43:40,600
Speaker 5: I found in the subparateated AI wars, or at least

902
00:43:40,640 --> 00:43:44,160
Speaker 5: I found someone sharing this, I believe on Twitter this post.

903
00:43:44,480 --> 00:43:47,120
Speaker 5: Nightshade has been released is use of it considered legal

904
00:43:47,200 --> 00:43:49,360
Speaker 5: or illegal for those who do not know its software

905
00:43:49,400 --> 00:43:51,880
Speaker 5: that attempts to poison an image, so if AI is trained,

906
00:43:51,880 --> 00:43:53,560
Speaker 5: it will mess up the model. For example, so you

907
00:43:53,600 --> 00:43:55,720
Speaker 5: have a picture of a cat and you run Nightshade

908
00:43:55,760 --> 00:43:57,400
Speaker 5: on it. If you attempt to train a model, that

909
00:43:57,520 --> 00:43:59,920
Speaker 5: image will replace the image and say dog prompt, cat

910
00:44:00,000 --> 00:44:02,840
Speaker 5: category or pencil, which means these prompts will be spoiled.

911
00:44:03,000 --> 00:44:04,799
Speaker 5: There is an issue that the creator of Nightshade has

912
00:44:04,840 --> 00:44:07,320
Speaker 5: not talked about, either from lack of legal knowledge or ignorance,

913
00:44:07,719 --> 00:44:09,880
Speaker 5: or they just don't care into them at someone else's problem.

914
00:44:09,880 --> 00:44:12,480
Speaker 5: The issue is it may be illegal in some countries. Basically,

915
00:44:12,520 --> 00:44:15,040
Speaker 5: if you release publicly a computer file in this case

916
00:44:15,080 --> 00:44:18,280
Speaker 5: image file that knowingly and willingly causes harm or distribution

917
00:44:18,360 --> 00:44:21,160
Speaker 5: to other people's computers or software, it may be considered

918
00:44:21,160 --> 00:44:23,760
Speaker 5: a criminal offense. Now it does not now, and again

919
00:44:24,120 --> 00:44:27,239
Speaker 5: I think that is stupid. I think they're just trying

920
00:44:27,280 --> 00:44:29,319
Speaker 5: to scare out instead of using this. You are not

921
00:44:29,400 --> 00:44:32,000
Speaker 5: harming someone's computer. You are harming a model that is

922
00:44:32,040 --> 00:44:35,880
Speaker 5: stealing something that's not illegal. Now they may try to

923
00:44:35,920 --> 00:44:37,240
Speaker 5: make it illegal.

924
00:44:37,160 --> 00:44:39,760
Speaker 4: Right, Yeah, I just want you to know the club

925
00:44:39,840 --> 00:44:43,440
Speaker 4: is illegal because you are. If I'm trying to steal

926
00:44:43,480 --> 00:44:46,760
Speaker 4: your car and I injure myself trying to break the club,

927
00:44:47,320 --> 00:44:48,239
Speaker 4: you have injured me.

928
00:44:48,719 --> 00:44:51,239
Speaker 5: Yeah, I put I invested a lot of money into

929
00:44:51,239 --> 00:44:54,520
Speaker 5: stealing catalytic converters, iffy. And if people are putting cages

930
00:44:54,560 --> 00:44:57,000
Speaker 5: around their cat that puts my investment in danger, and

931
00:44:57,040 --> 00:45:02,000
Speaker 5: that's illegal, right, you messing with my business, guy, Jesus Christ.

932
00:45:02,360 --> 00:45:03,280
Speaker 2: It is that logic.

933
00:45:03,480 --> 00:45:06,080
Speaker 5: There's like someone in the thread is like, how exactly

934
00:45:06,120 --> 00:45:08,880
Speaker 5: is your computer system or software harmed? And he responds,

935
00:45:08,880 --> 00:45:12,520
Speaker 5: it's equivalent to hacking a vulnerable computer system to disrupt

936
00:45:12,600 --> 00:45:16,799
Speaker 5: its operation. It's and then then he says, you are

937
00:45:16,840 --> 00:45:20,480
Speaker 5: intentionally disrupting its intended purpose creating art. This is directly

938
00:45:20,520 --> 00:45:23,080
Speaker 5: comparable to hacking. Like, I fucking hate this guy.

939
00:45:23,160 --> 00:45:25,160
Speaker 4: I want you, I want you to read it, but

940
00:45:25,320 --> 00:45:28,799
Speaker 4: in your head used Tim Robinson's voice, Yeah, and.

941
00:45:30,440 --> 00:45:36,120
Speaker 5: It just makes me, oh god, it's perfect, it's so good.

942
00:45:37,200 --> 00:45:39,080
Speaker 5: So all of this put me in a sour mood

943
00:45:39,080 --> 00:45:42,640
Speaker 5: if he but yeah, yeah, it did, it did. But

944
00:45:42,680 --> 00:45:44,440
Speaker 5: I think back when I'm in that mood, I think

945
00:45:44,480 --> 00:45:48,320
Speaker 5: back to Cees right, Like after I ask my question

946
00:45:48,520 --> 00:45:51,440
Speaker 5: and I make that Google and Microsoft people, I make

947
00:45:51,440 --> 00:45:54,440
Speaker 5: them kind of angry at me. Right after I asked

948
00:45:54,440 --> 00:45:58,200
Speaker 5: that question. The question after me is someone asking, hey,

949
00:45:58,840 --> 00:46:01,520
Speaker 5: you know the block chain with the last big craze,

950
00:46:01,560 --> 00:46:04,680
Speaker 5: do you think there's any future in you know, using

951
00:46:04,760 --> 00:46:07,480
Speaker 5: AI on the blockchain, And both of them were they

952
00:46:07,520 --> 00:46:07,840
Speaker 5: could not.

953
00:46:07,840 --> 00:46:09,080
Speaker 2: They were like no, like they.

954
00:46:08,960 --> 00:46:11,759
Speaker 5: Can't say no fast enough, Like absolutely, we don't care

955
00:46:11,800 --> 00:46:14,080
Speaker 5: about that anymore. We've moved on to the next crypt

956
00:46:14,600 --> 00:46:16,320
Speaker 5: Why are you bringing up the old grift.

957
00:46:16,600 --> 00:46:18,440
Speaker 4: It's dead, it's dead. We must move on.

958
00:46:18,880 --> 00:46:20,919
Speaker 2: Yeah, and that brought me a little bit of hope.

959
00:46:20,920 --> 00:46:24,600
Speaker 5: You know, perhaps we will get Mark Andreeson's benevolent AI god,

960
00:46:24,680 --> 00:46:27,440
Speaker 5: or perhaps we'll get Elisa Yidkowski's to look on devil,

961
00:46:27,680 --> 00:46:29,840
Speaker 5: or perhaps we'll just give control of all of the

962
00:46:29,880 --> 00:46:32,960
Speaker 5: future of ARC to fucking Sam Altman. But my guess

963
00:46:33,200 --> 00:46:35,600
Speaker 5: and my hope is that in the end, we Heretics

964
00:46:35,640 --> 00:46:39,000
Speaker 5: will survive the present crusade. And that's the end of

965
00:46:39,040 --> 00:46:40,440
Speaker 5: the episode that I've got for you.

966
00:46:40,520 --> 00:46:44,600
Speaker 4: Ify That is amazing. Yeah, I love it. I love

967
00:46:44,640 --> 00:46:45,280
Speaker 4: it so much.

968
00:46:47,239 --> 00:46:50,840
Speaker 5: Well, if he again, If you want this article, or

969
00:46:50,880 --> 00:46:53,000
Speaker 5: if you want the article version of this more condensed

970
00:46:53,040 --> 00:46:56,880
Speaker 5: easier to share. It's up on Rolling Stone. The article

971
00:46:56,920 --> 00:47:00,239
Speaker 5: is titled the Cult of AI and again that's by

972
00:47:00,280 --> 00:47:03,680
Speaker 5: me and Rolling Stone, The Cult of AI. Iffy, you

973
00:47:03,719 --> 00:47:07,920
Speaker 5: want to add in your stuff, plug your pluggables, so.

974
00:47:07,960 --> 00:47:10,520
Speaker 4: Oh yes, please if you wide a way on Twitter

975
00:47:10,520 --> 00:47:14,960
Speaker 4: and Instagram, watch dropout dot tv. You know it is.

976
00:47:15,520 --> 00:47:20,080
Speaker 4: It is definitely uh, you know, trying to do funny

977
00:47:20,080 --> 00:47:23,600
Speaker 4: things on the internet by humans and you know paying

978
00:47:23,680 --> 00:47:28,960
Speaker 4: those humans uh sharing, Oh yes, profit sharing, you know,

979
00:47:29,480 --> 00:47:32,600
Speaker 4: so truly big shout out to them. But yeah, I

980
00:47:32,680 --> 00:47:34,560
Speaker 4: might be in your town and be dueling a lot

981
00:47:34,600 --> 00:47:38,759
Speaker 4: of shows this year, so definitely pull up, you know,

982
00:47:39,239 --> 00:47:41,480
Speaker 4: follow me on the social meds and I'll let you

983
00:47:41,560 --> 00:47:43,319
Speaker 4: know where I'm at and you can just come. But

984
00:47:43,440 --> 00:47:46,000
Speaker 4: thank you uh so much for having me. It's so

985
00:47:46,040 --> 00:47:46,759
Speaker 4: good to see you again.

986
00:47:47,280 --> 00:47:48,760
Speaker 2: It was really good to see you again. Ifie.

987
00:47:48,920 --> 00:47:51,960
Speaker 4: Yeah, and this AI discussion in a weird way, as

988
00:47:52,040 --> 00:47:55,560
Speaker 4: dark as it's been, it makes me feel better because

989
00:47:55,560 --> 00:47:58,920
Speaker 4: I like that we're starting to fight back everyone. Good night.

990
00:47:58,960 --> 00:47:59,120
Speaker 6: Share.

991
00:47:59,160 --> 00:48:01,040
Speaker 4: Yeah, I think I'm gonna just are putting nice shade

992
00:48:01,040 --> 00:48:02,560
Speaker 4: on regular images.

993
00:48:02,760 --> 00:48:06,840
Speaker 5: Yeah, certainly, certainly one thing that's worth trying. And again,

994
00:48:07,360 --> 00:48:10,320
Speaker 5: you know, think about hyperstition, folks. We have to imagine

995
00:48:10,360 --> 00:48:15,120
Speaker 5: better futures in order to counter the imaginations of those

996
00:48:15,520 --> 00:48:18,520
Speaker 5: who wish us harm, who want to control and destroy

997
00:48:18,840 --> 00:48:21,279
Speaker 5: all that's good in the world. So, you know, get

998
00:48:21,320 --> 00:48:26,239
Speaker 5: on that. Somebody figure that out in the audience, all right.

999
00:48:26,360 --> 00:48:27,279
Speaker 5: Episode's over.

1000
00:48:30,400 --> 00:48:33,120
Speaker 3: Behind the Bastards is a production of cool Zone Media.

1001
00:48:33,480 --> 00:48:36,760
Speaker 3: For more from cool Zone Media, visit our website coolzonemedia

1002
00:48:36,960 --> 00:48:40,120
Speaker 3: dot com, or check us out on the iHeartRadio app,

1003
00:48:40,239 --> 00:48:42,520
Speaker 3: Apple Podcasts, or wherever you get your podcast