WEBVTT - Part Two: Tech Bros Have Built A Cult Around AI 0:00:01.800 --> 0:00:09.240 Also media, Welcome back to Behind the Bastards, a podcast 0:00:09.360 --> 0:00:12.200 that again I watched it. 0:00:12.360 --> 0:00:14.360 My weird voice didn't work. I'm sorry. 0:00:14.400 --> 0:00:15.640 I didn't enjoy that at all. 0:00:15.800 --> 0:00:18.920 I know, I know that one was like more Dracula 0:00:19.400 --> 0:00:19.960 than the first. 0:00:20.560 --> 0:00:21.760 Yeah, draculsque. 0:00:21.880 --> 0:00:23.439 Sure he didn't sell it this time. 0:00:23.600 --> 0:00:28.560 Yeah, sorry, I apologize. What I don't apologize for is 0:00:28.600 --> 0:00:29.680 my guest, if. 0:00:29.560 --> 0:00:30.280 In' what he were? 0:00:33.040 --> 0:00:33.640 Is you boy? 0:00:34.159 --> 0:00:36.040 If working for. 0:00:35.920 --> 0:00:40.519 The drop out, the survivor's successors to college humor, who 0:00:40.600 --> 0:00:44.080 have who have blossomed like a phoenix from the ashes 0:00:44.120 --> 0:00:46.200 of the internet that Facebook killed? 0:00:47.280 --> 0:00:50.360 Yes, yes, still we rise. 0:00:50.520 --> 0:00:51.400 Still we rise. 0:00:51.840 --> 0:00:56.440 Speaking of rising my concerns about the cult dynamics within 0:00:56.520 --> 0:01:00.640 the AI sub cult, You're I don't know why I 0:01:00.640 --> 0:01:02.080 said it that way, So I. 0:01:02.560 --> 0:01:04.160 Don't know what you're on about right now. 0:01:04.920 --> 0:01:08.160 I don't know either. I don't know. I'm doing great. 0:01:08.240 --> 0:01:10.679 I'm doing great. This is this is me. 0:01:10.760 --> 0:01:13.600 I've been sober lately, so this is me living the 0:01:13.640 --> 0:01:16.160 sober life. I've just gotten worse. 0:01:16.400 --> 0:01:17.440 So I don't know. 0:01:17.440 --> 0:01:24.040 Everybody keep your kids on drugs, you know, or depends 0:01:24.080 --> 0:01:24.600 on the drugs. 0:01:24.640 --> 0:01:27.920 Drug Yeah, yeah. 0:01:26.840 --> 0:01:29.200 So perhaps the most amusing part of all of this 0:01:29.360 --> 0:01:32.119 is that a segment of the AI believing community has 0:01:32.160 --> 0:01:35.959 created not just a potential god, but a hell. And 0:01:36.000 --> 0:01:38.840 this is one of my favorite stories from these weirdos. 0:01:39.440 --> 0:01:42.440 One of the early online subcultures that influence the birth 0:01:42.440 --> 0:01:45.959 of EYAK are the Rationalists. And again the EAC people 0:01:45.959 --> 0:01:48.120 will say a lot of them don't like the rationalists, 0:01:48.160 --> 0:01:50.480 but they're they're related. They're like cousins in the same 0:01:50.520 --> 0:01:54.960 way Crackton College Humor are right. The Rationalists are a 0:01:55.000 --> 0:01:57.440 subculture that formed in the early aughts. They kind of 0:01:57.480 --> 0:02:00.720 came out of the online skeptic a the movement of 0:02:00.760 --> 0:02:03.400 the late nineties, and they formed in the early yachts 0:02:03.440 --> 0:02:06.120 around a series of blog posts by a man named 0:02:06.120 --> 0:02:11.120 Eleezer Yedkowski. Ydkowski fancies himself as something of a philosopher 0:02:11.200 --> 0:02:14.880 on AI, and his blog slash discussion board Less Wrong 0:02:15.040 --> 0:02:18.919 was an early hub of the broader AI subculture. Yeddkowski 0:02:19.240 --> 0:02:21.840 like he doesn't have a specific education, He just cames 0:02:21.880 --> 0:02:24.320 to be kind of an expert in AI and machine learning. 0:02:25.120 --> 0:02:30.200 He's a peculiar fellow, to say the least. The founding text, 0:02:30.280 --> 0:02:32.480 or at least one of them, of Rationalism is a 0:02:32.560 --> 0:02:35.760 six hundred and sixty thousand word Harry Potter fanfic. That 0:02:35.919 --> 0:02:39.320 is just just nonsense, it is. It's all about like 0:02:39.720 --> 0:02:43.120 rewriting Harry Potter. So his real magic is rational thinking. 0:02:43.320 --> 0:02:48.720 It's wild shit, he's like a psychopath. It's so such 0:02:48.760 --> 0:02:50.160 an odd choice. 0:02:49.919 --> 0:02:52.360 You know. It's just like the uh, what was it, 0:02:52.440 --> 0:02:56.079 fifty Shades of Gray? How yes, originally a Twilight fan thick, 0:02:56.440 --> 0:03:01.880 and there's going to be like a Cloud atlass esque. 0:03:00.560 --> 0:03:00.760 You know. 0:03:00.960 --> 0:03:03.840 But you know, the fifty shadsh Gray lady was not 0:03:03.919 --> 0:03:08.840 trying to create the new text for like a philosophical movement. 0:03:08.919 --> 0:03:12.320 She just wanted to get like people horny. And that's fine, 0:03:13.480 --> 0:03:17.360 that's perfectly acceptable. The most relevant thing about the six 0:03:17.440 --> 0:03:20.840 hundred and sixty thousand word Harry Potter fanfic is that 0:03:20.880 --> 0:03:23.520 it was the favorite book of Carolyn Ellison, this former 0:03:23.600 --> 0:03:29.000 CEO of FTX who recently testified against Sam Bankman freed 0:03:29.760 --> 0:03:32.560 or of Alimeter. Sorry she was the CEO of Alameda anyway. 0:03:32.800 --> 0:03:36.360 All these weird little subcultures rationalism and effective altruism are 0:03:36.400 --> 0:03:39.120 related to each other and influenced each other, even though 0:03:39.120 --> 0:03:41.920 again they often hate each other too. Yudkowski is seen 0:03:42.000 --> 0:03:45.160 as an object of ridicule by most YAK people. This 0:03:45.280 --> 0:03:47.280 is because he shares their view of AI as a 0:03:47.320 --> 0:03:51.280 potential deity, but he believes AGI will inevitably kill everyone. 0:03:51.560 --> 0:03:53.800 Thus we must bomb data centers. 0:03:53.520 --> 0:03:56.520 Which like look, he may have gotten to the right end. 0:03:56.680 --> 0:04:01.200 Play we get running like like forest Gum. He just 0:04:01.280 --> 0:04:01.840 kept running. 0:04:01.880 --> 0:04:04.200 We're like wait, wait, wait, wait, no, stop right there, 0:04:04.240 --> 0:04:05.840 stop right there. We may agree with you on this. 0:04:06.080 --> 0:04:09.800 Yeah, Yedkowski as a doomer now because he was surprised 0:04:09.800 --> 0:04:12.360 when Chad GBT came out. He was like horrified by 0:04:12.360 --> 0:04:14.280 how advanced it was and was like, oh my god, 0:04:14.440 --> 0:04:17.280 we're further along towards creating the AI that kills us all. 0:04:17.320 --> 0:04:19.920 We have to stop this now. And that made him 0:04:20.040 --> 0:04:22.240 He had been kind of flirted with a lot of 0:04:22.360 --> 0:04:25.800 like Silicon Valley people. He's rationalists are very much a 0:04:25.800 --> 0:04:28.320 Bay Area cult. He kind of has become increasingly a 0:04:28.360 --> 0:04:31.400 pariat of at least people with money in AI. But 0:04:31.600 --> 0:04:35.839 before that happened, his message board birthed something wondrous. In 0:04:35.880 --> 0:04:39.880 twenty ten, a less Wrong user named Rocko posted this question, 0:04:40.279 --> 0:04:43.280 what if an otherwise benevolent AI decided it had to 0:04:43.279 --> 0:04:45.640 torture any human who failed to work to bring it 0:04:45.680 --> 0:04:48.200 into existence. Right, what if we make an all powerful 0:04:48.240 --> 0:04:52.039 AI and its logical decision is that, well, I will 0:04:52.080 --> 0:04:54.480 have to punish all the human beings who were alive 0:04:54.560 --> 0:04:58.000 and who didn't try to further my existence, because that's 0:04:58.040 --> 0:05:00.720 the most real reasonable way to geary that I come 0:05:00.760 --> 0:05:04.320 into being. It's nonsense. This is a silly, silly thing 0:05:04.360 --> 0:05:07.159 to believe. It's all based on like the prisoner's dilemma, 0:05:07.200 --> 0:05:09.000 which is a concept in game theory, And it's not 0:05:09.640 --> 0:05:12.760 really worth explaining why, because the logic is it's the 0:05:12.839 --> 0:05:14.800 only the kind of thing that happens when people are 0:05:14.800 --> 0:05:19.120 two online and like completely get detached from reality. But 0:05:19.279 --> 0:05:22.240 Raco's conclusion here is that an AI who felt this 0:05:22.360 --> 0:05:25.599 way would punish its apast states for eternity by creating 0:05:25.640 --> 0:05:29.240 a virtual reality hell digitizing their consciousness and making them 0:05:29.240 --> 0:05:30.200 suffer for all time. 0:05:30.680 --> 0:05:31.680 Now, WHOA. 0:05:31.800 --> 0:05:33.960 You may have noticed Iffy number one. They're kind of 0:05:34.000 --> 0:05:38.960 ripping off our boy, Harlan Ellison, famed advocate of the 0:05:39.000 --> 0:05:42.480 writer's right to their work. But it's also just tech 0:05:42.520 --> 0:05:46.880 nerds recreating Pascal's wager, Like this is just Pascal's Rager 0:05:46.960 --> 0:05:49.880 with an AI like, you just stole again. 0:05:49.720 --> 0:05:53.919 These fucking plagiarists. You just stole from whoever past scal was, right. 0:05:53.839 --> 0:05:56.120 This is what happens when you are nerd and you 0:05:56.200 --> 0:05:59.160 refuse to read sci fi. You just you eventually just 0:05:59.440 --> 0:06:02.480 come up with these stories yourselves and think that you 0:06:02.520 --> 0:06:02.839 did it. 0:06:03.440 --> 0:06:03.680 Yeah. 0:06:03.680 --> 0:06:05.520 This, and if you're not familiar, folks, I think most 0:06:05.560 --> 0:06:07.760 people are Pascal's wagers. This kind of like concept from 0:06:07.800 --> 0:06:10.640 I think it's you'd call it a Christian apologetics. That's like, 0:06:11.960 --> 0:06:14.839 we may not know if hell is real or not, 0:06:15.120 --> 0:06:18.120 but because if it's real, the consequences are so dire 0:06:18.360 --> 0:06:22.279 and the cost of just saying yeah, I accept Jesus 0:06:22.440 --> 0:06:25.080 is so low, you should do that, right, Like or 0:06:25.160 --> 0:06:27.240 I think that's the basic idea, right, It's a lot 0:06:27.240 --> 0:06:29.680 of people interpret it. It's the whole idea behind like 0:06:29.720 --> 0:06:32.240 being a piece of shit and then converting on your deathbed. Basically, 0:06:33.080 --> 0:06:34.719 I don't know fully the history of it, but I 0:06:34.720 --> 0:06:37.960 know that they're basically aping it for a fucking Rocko's basilisk. 0:06:39.040 --> 0:06:41.400 And it's called basilisk because like a basilisk, if you 0:06:41.400 --> 0:06:44.279 look at it, it like enraptures your mind. You can't 0:06:44.279 --> 0:06:48.240 stop thinking about it. That comes from Reportedly, there's some 0:06:48.279 --> 0:06:50.479 debate over this when this went viral among like the 0:06:50.560 --> 0:06:53.400 less Wrong community. Yet Kowski had the banned discussion of 0:06:53.440 --> 0:06:56.000 it because it was like breaking people's minds. They were 0:06:56.000 --> 0:06:58.359 having nightmares. Am I working hard enough to make the 0:06:58.440 --> 0:07:00.720 AI reel? Is it gonna say me to hell? 0:07:02.440 --> 0:07:02.920 Yeah? 0:07:03.000 --> 0:07:06.320 It's unclear like how seriously people were because again there's 0:07:06.360 --> 0:07:09.440 just people talking on the internet for what it's worth. 0:07:09.520 --> 0:07:13.480 Yet Kowski didn't really like Rocko's basslist, but it's it's 0:07:13.560 --> 0:07:15.800 his place that birthed it. And for an idea of 0:07:15.800 --> 0:07:19.559 how influential this is, Elon Muskin Grimes met talking about 0:07:19.560 --> 0:07:23.040 the concept that was their meat cute was this was 0:07:23.080 --> 0:07:26.840 fucking AI Pascal's wager. Yeah, she like wrote a song 0:07:26.920 --> 0:07:31.960 about it. It's fucking ridiculous. These fucking people are such teks. 0:07:32.560 --> 0:07:37.960 Yeah. Wow, oh my god. Read Harlan Ellison. He did 0:07:38.000 --> 0:07:40.360 it better than you. God damn it. 0:07:41.400 --> 0:07:43.800 I will say reading this shit is the most I've 0:07:43.800 --> 0:07:46.239 ever felt like I have no mouth, but I must scream. 0:07:46.400 --> 0:07:51.800 So again, pour one out for the man. So this 0:07:51.960 --> 0:07:54.520 is all relevant. This AI hell some of these people 0:07:54.560 --> 0:07:56.840 have created because it's one more data point showing that 0:07:56.840 --> 0:08:00.000 the people who take AI very seriously as real intelligent 0:08:00.560 --> 0:08:02.640 always seem to turn it into religion. And this is 0:08:02.720 --> 0:08:05.120 kind of maybe the first schism, right, this is their 0:08:05.120 --> 0:08:08.760 Catholic Protestant split or their Catholic Orthodox split, because you've 0:08:08.800 --> 0:08:10.920 got a one side ed Kowski's people who are like, 0:08:11.000 --> 0:08:13.880 we will inevitably make a God and that God will 0:08:13.880 --> 0:08:16.960 destroy us, so we have to stop it, versus like, 0:08:17.000 --> 0:08:19.400 we will inevitably make a God and that God will 0:08:19.440 --> 0:08:22.280 take us to paradise along with Daddy Musk, We'll go 0:08:22.320 --> 0:08:24.640 to the stars. Right, those are the two. This is 0:08:24.680 --> 0:08:29.239 like the first heretical split within the Divine AI movement. 0:08:29.680 --> 0:08:33.040 And this stuff is relevant because so many of the 0:08:33.080 --> 0:08:36.400 fucking these subcultures and movements start out as a bunch 0:08:36.440 --> 0:08:40.120 of people arguing or discussing their ideas in online communities. 0:08:40.480 --> 0:08:42.400 And there is a reason for this. It's pretty well 0:08:42.440 --> 0:08:45.760 recognized that there are certain dynamics inherent to the kind 0:08:45.800 --> 0:08:49.640 of communities that start on the Internet that tend towards cultishness. 0:08:49.640 --> 0:08:51.520 This is part of why, like, we have a big 0:08:51.600 --> 0:08:54.640 subreddit for the podcast, it's like eighty something thousand people, 0:08:55.280 --> 0:08:57.280 which makes it in like the top one percent of Reddit, 0:08:57.600 --> 0:08:59.480 and I have been offered like to be able to 0:08:59.480 --> 0:09:02.040 moderate and like make policy there. I have nothing to 0:09:02.080 --> 0:09:04.640 do with the running of that subreddit because I'm like, 0:09:04.880 --> 0:09:06.959 that doesn't end well. I was on something offul as 0:09:06.960 --> 0:09:09.559 a kid. I know what happens when people make themselves 0:09:09.600 --> 0:09:13.120 mods of giant digital communities. They lose their fucking minds 0:09:15.080 --> 0:09:17.240 or all watching Elon Musk do it right now, it's 0:09:17.280 --> 0:09:19.960 the worst thing in the world for you. Thank you, 0:09:20.000 --> 0:09:22.400 by the way, to the people who do run that thing. Uh, 0:09:23.200 --> 0:09:26.560 because I am not going to. The skeptic community, which 0:09:26.640 --> 0:09:29.320 was huge through the late nineteen nineties and early two thousands, 0:09:29.400 --> 0:09:32.440 might be seen as the grandfather of all these little subcultures. 0:09:32.880 --> 0:09:35.959 After nine to eleven, prominent skeptics became vocally unhinged in 0:09:36.000 --> 0:09:38.400 their hatred of Islam, which brought them closer to different 0:09:38.440 --> 0:09:42.160 chunks of the nascent online far right. Weird shit started 0:09:42.200 --> 0:09:45.040 to crop up like a movement to rebrand skeptics as 0:09:45.120 --> 0:09:47.800 brights in light of the fact that their very clearly 0:09:47.840 --> 0:09:50.960 exceptional intelligence made them better than other people. And again 0:09:51.000 --> 0:09:52.679 you can see some similarity with this and the stuff 0:09:52.760 --> 0:09:55.559 Nick Land was talking about, only certain races will make 0:09:55.600 --> 0:09:58.280 it to space. I found a very old write up 0:09:58.320 --> 0:10:01.000 on plover dot net that described the method by which 0:10:01.000 --> 0:10:05.160 this kind of shit happens in digital communities. Quote online forums, 0:10:05.160 --> 0:10:07.960 whatever their subject, can be forbidding places for the newcomer. 0:10:08.160 --> 0:10:10.480 Over time, most of them tend to become dominated by 0:10:10.520 --> 0:10:12.920 small groups of snotty know it alls who stamp their 0:10:12.960 --> 0:10:16.360 personalities over the proceedings. But skeptic forums are uniquely meant 0:10:16.360 --> 0:10:20.160 for such people. A skeptic forum valorizes and in some cases, fetishizes, 0:10:20.360 --> 0:10:24.319 competitive geekery, gratuitous cleverness, macho displays of erudition. It's a 0:10:24.400 --> 0:10:27.640 gathering of rationalities, hard men thumping their chests, showing off 0:10:27.640 --> 0:10:31.199 their muscular logic, glancing sideways to compare their skeptical endowment 0:10:31.200 --> 0:10:33.560 with the next guy sniffing the air for signs of weakness. 0:10:33.720 --> 0:10:36.600 Together they create an oppressive, sweaty locker room atmosphere that 0:10:36.640 --> 0:10:40.200 helps keep uncomfortable demographics away. And that is where a 0:10:40.240 --> 0:10:42.000 lot of this shit is cropping up. 0:10:42.080 --> 0:10:42.320 Right. 0:10:42.520 --> 0:10:46.040 It is sweaty and uncomfortable, and there are mushrooms growing there, 0:10:46.080 --> 0:10:49.400 and some of those mushrooms are fucking fashions, and all 0:10:49.440 --> 0:10:52.200 of them want to take away the ability of artists 0:10:52.200 --> 0:10:53.640 to choose what happens to their art. 0:10:53.960 --> 0:10:56.600 Oh yeah, I feel like this is just so many 0:10:57.160 --> 0:11:00.920 parts of the Zeitgei's coming together because you know what 0:11:00.960 --> 0:11:04.520 it means to own media, you know. I feel like 0:11:04.559 --> 0:11:08.160 a very small microcosm of this is when people would 0:11:08.240 --> 0:11:12.280 like clip out stuff from YouTube videos or eight jokes 0:11:12.280 --> 0:11:16.240 from people who tweet, and when it goes you know, 0:11:16.520 --> 0:11:20.280 viral or in the original tweet is like, hey, you 0:11:20.360 --> 0:11:22.880 stole this for me, and it's either no, I didn't, 0:11:23.040 --> 0:11:25.760 or like, yeah, but you like put it on Twitter, 0:11:25.920 --> 0:11:30.000 so like I can just copy what you wrote. Yeah, 0:11:29.640 --> 0:11:32.840 And now it has evolved into yeah, we can just 0:11:32.920 --> 0:11:35.400 take from yours and let this machine learn how to 0:11:35.440 --> 0:11:36.920 do what you do so I can do it even 0:11:36.920 --> 0:11:39.560 though I don't have the talent to do it. 0:11:39.880 --> 0:11:40.280 Yeah. 0:11:40.320 --> 0:11:43.960 Absolutely, the reality of AI's promise is a lot more 0:11:44.000 --> 0:11:47.920 subdued than believers want to admit. In an article published 0:11:47.960 --> 0:11:52.280 by Frontiers in Ecology and Evolution, a peer reviewed research journal, 0:11:52.720 --> 0:11:56.320 doctor Andreas Roli and colleagues argue that AGI is not 0:11:56.559 --> 0:12:00.360 achievable in the current algorithmic frame of AI rese and 0:12:00.360 --> 0:12:03.360 this is a their claims are very stark that like 0:12:03.679 --> 0:12:06.640 the kind of way we make these these large language models, 0:12:06.640 --> 0:12:10.199 this algorithmic frame, cannot make an intelligence. That's their argument. 0:12:10.840 --> 0:12:13.800 One point they make is that intelligent organisms can both 0:12:13.880 --> 0:12:18.480 want things and improvise capabilities that no models have yet generated. 0:12:18.840 --> 0:12:21.400 They also argue, basically all of these things that individual 0:12:21.440 --> 0:12:26.480 AI type models can do, you know, recognize voice, recognize text, 0:12:26.720 --> 0:12:30.319 recognize faces, you know, this kind of stuff, those are 0:12:30.400 --> 0:12:33.280 pieces of what we would want from an artificial general intelligence. 0:12:33.320 --> 0:12:35.840 But they're not all combined in like the same thing 0:12:36.200 --> 0:12:39.960 that works seamlessly. And beyond that, it can't. It can't 0:12:40.040 --> 0:12:42.800 act based on anything internal right. 0:12:42.880 --> 0:12:43.760 It can only. 0:12:43.520 --> 0:12:46.559 Act based on prompts, And their argument is that algorithmic 0:12:46.600 --> 0:12:49.400 AI will not be able to make the jump to acting. Otherwise, 0:12:49.800 --> 0:12:53.040 what we call AI then lacks agency, the ability to 0:12:53.080 --> 0:12:56.480 make dynamic decisions of its own accord choices that are quote, 0:12:56.520 --> 0:13:00.719 not purely reactive, not entirely determined by environmental conditions. Mid 0:13:00.800 --> 0:13:02.880 journey can read a prompt and return with art it 0:13:02.920 --> 0:13:06.200 calculates will fit the criteria. Only a living artist can 0:13:06.280 --> 0:13:09.079 choose to seek out inspiration and technical knowledge and then 0:13:09.080 --> 0:13:12.880 produce the art that mid journey digests and regurgitates. Now, 0:13:12.920 --> 0:13:15.640 this paper is not going to be the last word 0:13:15.679 --> 0:13:17.959 on whether or not AGI is possible, or whether it's 0:13:17.960 --> 0:13:21.160 possible under our current algorithmic method of like making AIS. 0:13:21.360 --> 0:13:23.719 I'm not making myself a claim there. I'm saying these 0:13:23.760 --> 0:13:26.600 people are and I think their arguments are compelling. We 0:13:26.640 --> 0:13:29.520 don't know yet entirely. Again, this is not a settled 0:13:29.520 --> 0:13:32.400 field of research obviously, But my point is that the 0:13:32.440 --> 0:13:37.240 goals Andreasen and the Effective Accelerationist Crew champion right now 0:13:37.440 --> 0:13:39.840 are not based in fact. We don't know that what 0:13:39.840 --> 0:13:42.199 they're saying. That the most basic level of what they're 0:13:42.240 --> 0:13:45.400 saying is possible, and that means that their beliefs are 0:13:45.400 --> 0:13:46.360 based in faith. 0:13:46.960 --> 0:13:49.200 Right, How else can you look at that? Yeah? 0:13:49.400 --> 0:13:53.800 Yeah, Like this is a faith and again it's the 0:13:53.880 --> 0:13:56.680 kind of faith that, according to Andresen, makes you a 0:13:56.760 --> 0:13:59.200 murderer if you doubt it, which I don't think I 0:13:59.200 --> 0:14:04.359 need to draw your parallels two specific religions here, right, Yeah, Yeah. 0:14:04.080 --> 0:14:06.040 This is this is that point where when you're like 0:14:06.920 --> 0:14:10.880 Stone and you're watching those like you know, art time 0:14:10.960 --> 0:14:14.560 lapses and the picture is starting a form, and I'm like, Okay, 0:14:14.559 --> 0:14:17.120 I see what Robert's doing. I see the picture is coming. 0:14:17.400 --> 0:14:18.880 I was on your side from the jump. I just 0:14:18.920 --> 0:14:21.040 want to say, you know, I was, you know, I 0:14:21.120 --> 0:14:24.040 was like, yeah, no, I believe you. But now I'm 0:14:24.120 --> 0:14:27.120 watching the connections be made and yeah, I love it. 0:14:27.560 --> 0:14:28.440 Yeah now. 0:14:28.440 --> 0:14:32.080 Andresen's manifesto claims our enemies are not bad people, but 0:14:32.240 --> 0:14:34.560 rather bad ideas. And I have to wonder, doing all this, 0:14:34.840 --> 0:14:37.200 putting this episode out, where does that leave me in 0:14:37.240 --> 0:14:39.440 his eyes? Or doctor Rowley for that matter, and the 0:14:39.440 --> 0:14:41.920 other people who worked on that paper. We have seen 0:14:42.000 --> 0:14:43.960 many times in history what happens when members of a 0:14:43.960 --> 0:14:46.600 faith decide someone is their enemy and the enemy of 0:14:46.640 --> 0:14:50.120 their belief system. And right now, artists and copyright holders 0:14:50.120 --> 0:14:52.960 are the ones being treated as fair game by the 0:14:53.080 --> 0:14:57.320 AI industry. So my question is kind of first and foremost, 0:14:57.480 --> 0:15:01.960 who's going to be the next heretic? Right, Like, That's 0:15:02.280 --> 0:15:04.120 that's what I want to know. And I want to 0:15:04.160 --> 0:15:06.200 leave you all that thought before we go into some 0:15:06.280 --> 0:15:08.680 ads here, and then we will come back to talk 0:15:08.840 --> 0:15:12.160 about some people that I pissed off at CES. 0:15:12.440 --> 0:15:20.560 So that'll be fun. We're back. 0:15:21.040 --> 0:15:23.760 So one of the things I did was this panel 0:15:23.800 --> 0:15:27.280 on the AI driven restaurant and retail experience. 0:15:27.320 --> 0:15:28.120 I was very curious. 0:15:28.160 --> 0:15:30.840 I was AI going to change me getting some terrible 0:15:30.880 --> 0:15:32.440 food from McDonald's when I'm. 0:15:32.280 --> 0:15:33.000 On a road trip. 0:15:33.120 --> 0:15:37.240 Right, the host of that, Andy Hewles from Radius AI, 0:15:37.560 --> 0:15:40.600 asked the audience in relation to AI, raise your hand 0:15:40.640 --> 0:15:42.920 if you're a brand who feels like we've got this. 0:15:43.320 --> 0:15:47.520 That is how she phrased it. I hated it, but 0:15:47.560 --> 0:15:49.400 about a third of the room raised their hands. So 0:15:49.480 --> 0:15:51.280 next she asked for a shill of hands of the 0:15:51.360 --> 0:15:54.720 brands who identified with this statement. I'm not sure about this. 0:15:54.800 --> 0:15:57.480 I haven't tried it AI yet, but I want to, 0:15:57.520 --> 0:15:58.480 and that's why I'm here. 0:15:58.680 --> 0:15:58.840 Right. 0:15:59.360 --> 0:16:01.040 Most of the rest to the room raised their hands 0:16:01.040 --> 0:16:04.040 at that point, and she seems satisfied but said, and 0:16:04.080 --> 0:16:06.080 then I bet there's even some of you that are like, WHOA, 0:16:06.240 --> 0:16:08.120 I heard this is going to steal jobs, take away 0:16:08.200 --> 0:16:10.000 my privacy, affect the global economy. 0:16:10.040 --> 0:16:10.240 You know. 0:16:10.520 --> 0:16:12.560 AI is a little bit sketch in my mind, and 0:16:12.600 --> 0:16:14.600 I'm just worried about it, and I'm here to explore. 0:16:14.960 --> 0:16:18.440 Well that fit me, So I raised my hand. She 0:16:18.520 --> 0:16:20.760 didn't notice me at first, and so she like fakes 0:16:20.760 --> 0:16:22.680 a whisper and she's like, all right, good, there's none 0:16:22.680 --> 0:16:24.960 of you. And then she like looks over and sees 0:16:25.000 --> 0:16:27.600 me waving my hand and she says, louder and with 0:16:27.680 --> 0:16:31.400 evident disappointment, there's one. All right, you can ask questions 0:16:31.400 --> 0:16:31.960 at the end. 0:16:32.360 --> 0:16:32.960 So I did. 0:16:34.320 --> 0:16:36.840 I was very excited to get to do that. So 0:16:36.920 --> 0:16:40.000 the panel consisted of Bishad Mazati, a VP of engineering 0:16:40.000 --> 0:16:42.920 at Google, had mentioned during the panel that embracing AI 0:16:43.040 --> 0:16:45.680 could be the equivalent of adding a million employees to 0:16:45.760 --> 0:16:50.120 your company. The McDonald's representative, Michelle Gansel, claimed around the 0:16:50.160 --> 0:16:52.600 same time that her company used AI to prevent fifty 0:16:52.640 --> 0:16:55.160 million dollars in fraud attempts in just a single month. 0:16:55.640 --> 0:16:57.840 Now that's lovely, But I told her, you know, when 0:16:57.880 --> 0:17:00.040 I had my question, I was like, I'm going you 0:17:00.160 --> 0:17:04.040 assume most of those fraud attempts were AI generated, Right, So, yeah, 0:17:04.119 --> 0:17:06.400 you stopped a bunch of AI fraud, But that doesn't 0:17:06.440 --> 0:17:11.160 necessarily get me optimistic about AI's potential. And likewise, maybe 0:17:11.200 --> 0:17:13.679 Google gets the equivalent of a million employees, but so 0:17:13.760 --> 0:17:17.640 do all of the people committing fraud and disinformation on Google. Right, 0:17:17.880 --> 0:17:20.320 So again, how are we getting ahead? And I brought 0:17:20.400 --> 0:17:24.920 up this concept and evolutionary biology the red Queen hypothesis, 0:17:24.920 --> 0:17:27.760 which is kind of talking about the way that populations 0:17:27.800 --> 0:17:30.760 of animals evolve over time. Right where you've got an 0:17:30.760 --> 0:17:32.920 animal will evolve to be a better predator, so it's 0:17:32.960 --> 0:17:35.239 prey will evolve to be better at avoiding it. And 0:17:35.240 --> 0:17:37.640 it's kind of the reason it's the Red Queen dilemma 0:17:37.640 --> 0:17:39.639 is that, like, you've got to move as fast as 0:17:39.680 --> 0:17:42.479 you can just to stay in place. That's the red 0:17:42.720 --> 0:17:44.879 Queen dilemma, right, you got to move as fast as 0:17:44.920 --> 0:17:46.199 you can just to stay in one place. And I 0:17:46.240 --> 0:17:48.800 was like, is that not what we're going to wind 0:17:48.880 --> 0:17:52.080 up seeing with AI? Right, Yeah, we get better at 0:17:52.080 --> 0:17:55.159 a bunch of stuff, but it's eaten up to counter 0:17:55.320 --> 0:17:57.240 all of the things that get worse. And so I 0:17:57.240 --> 0:17:59.479 asked them, what are the odds that these gains are 0:17:59.480 --> 0:18:00.800 offset by the costs? 0:18:01.400 --> 0:18:01.640 Now? 0:18:01.680 --> 0:18:03.639 In the article that I wrote for Rolling Stone, I 0:18:03.640 --> 0:18:07.480 gave a significantly more condensed version of Bashad's answered, boiling 0:18:07.520 --> 0:18:10.240 out the ums and ohs and you knows, because that 0:18:10.240 --> 0:18:12.040 that would kind of make the case that he was 0:18:12.080 --> 0:18:16.840 absolutely unprepared for a vaguely critical question, a very basic one, 0:18:17.119 --> 0:18:19.960 and that he didn't really care enough to think about 0:18:19.960 --> 0:18:22.639 any of the security threats inherent to the technology. But 0:18:22.760 --> 0:18:26.800 actually that is what I think of him, and I'm 0:18:26.800 --> 0:18:30.639 gonna I'm gonna play you, audience. My concern is, like, 0:18:31.720 --> 0:18:34.560 what are the odds that a lot of these games 0:18:34.600 --> 0:18:35.920 that we get from AI are. 0:18:35.920 --> 0:18:37.320 Offset from the cost? 0:18:37.400 --> 0:18:40.800 You know, you noted Bashad that you get you know, 0:18:40.840 --> 0:18:43.480 a million extra workers by utilizing this, but so do 0:18:43.560 --> 0:18:46.399 the bad guy. So yeah, that's kind of where my 0:18:46.400 --> 0:18:47.720 my skepticism plays out. 0:18:48.440 --> 0:18:51.440 Yeah, certainly there will be you know, there are enough 0:18:51.560 --> 0:18:53.639 bad guys, I guess in the world which will use 0:18:53.640 --> 0:18:56.199 and and I forgot to use cases and and it's 0:18:56.280 --> 0:18:58.680 very important to also I have you know, be protected 0:18:58.760 --> 0:19:02.040 you know, against that against those and that's why you know, we. 0:19:02.040 --> 0:19:04.800 Take responsibility very serious, you know, also in terms of 0:19:04.840 --> 0:19:09.000 you know, what security aspects you know, brought fighting you know, 0:19:09.400 --> 0:19:11.359 and all of that for him, you know. 0:19:12.640 --> 0:19:13.359 And I think that's. 0:19:13.240 --> 0:19:15.280 Why I guess things should be regulated. And there's of 0:19:15.320 --> 0:19:19.439 course all these discussions out there, and I think, yeah. 0:19:18.920 --> 0:19:21.600 You may notice that that's not exactly a very like 0:19:22.359 --> 0:19:26.439 good response. I guess that's why this should be regulated. 0:19:26.680 --> 0:19:27.680 It's just like he. 0:19:27.680 --> 0:19:30.440 Starts talking so much faster with that, there's so much 0:19:30.440 --> 0:19:31.560 of like panic. 0:19:31.320 --> 0:19:36.200 Voice wavering too. I was like, man, yeah, it seems 0:19:36.200 --> 0:19:38.320 like he has that huge anime flop sweat. 0:19:38.600 --> 0:19:39.040 Yeah. 0:19:39.600 --> 0:19:41.880 One of the things about this cees as a trade 0:19:41.880 --> 0:19:43.840 show is that like a lot of people there do 0:19:43.920 --> 0:19:47.320