WEBVTT - Why Animal Rights Advocates Are Investigating AI Consciousness - Week in Tech 0:00:11.800 --> 0:00:15.320 So you've all likely heard about Tilly Norwood, the first 0:00:15.480 --> 0:00:20.159 ai A lister. Her creator just announced that Tilly is 0:00:20.160 --> 0:00:24.079 about to star in her first feature film. Taylor and Natasha, 0:00:24.120 --> 0:00:26.200 your face is a quite a picture. 0:00:26.520 --> 0:00:28.280 This is just a fake PR thing. 0:00:28.640 --> 0:00:31.360 Wait, Taylor, do you remember when all the editors wanted 0:00:31.360 --> 0:00:33.800 you to like interview Little Michayla. 0:00:34.560 --> 0:00:39.320 Yeah, I've met Lit Michayla's creator, Trevor. 0:00:39.560 --> 0:00:40.160 I know Trevor. 0:00:40.200 --> 0:00:41.880 I've known Trevor for a long time. I have the 0:00:41.880 --> 0:00:44.839 girl that Little Michaela was designed off of. I But 0:00:45.280 --> 0:00:48.559 with all due respect, love Trevor. But guys, this is 0:00:48.600 --> 0:00:49.000 just PR. 0:00:49.120 --> 0:00:49.720 It's just PR. 0:00:54.440 --> 0:00:56.920 Welcome to tech stuff. I'm Maza Velosa and this is 0:00:56.960 --> 0:00:59.319 the Weekend Tech, where I'm joined by the world's most 0:00:59.360 --> 0:01:02.240 plugged in report is to break down what's really happening 0:01:02.240 --> 0:01:05.280 in tech right now. Today, we're joined by Taylor Lorenz 0:01:05.319 --> 0:01:08.600 of User mag Read Abrigotti, tech editor for Semiphore, and 0:01:08.680 --> 0:01:11.279 Natasha Tku, tech reporter at the Washington Post. 0:01:11.360 --> 0:01:12.520 Welcome, Well, how's it going. 0:01:12.760 --> 0:01:13.800 Thanks for having us. 0:01:13.880 --> 0:01:14.720 Nice to be back. 0:01:15.120 --> 0:01:18.240 Natasha. You recently published an article that got to number one, 0:01:18.319 --> 0:01:21.800 I think in the in the Washington post algorithm. The 0:01:21.840 --> 0:01:24.720 subheads of the story is anthropic. Google and Meta have 0:01:24.840 --> 0:01:29.760 hired computer scientists, neuroscientists, and philosophers to study what some 0:01:29.840 --> 0:01:34.200 in the industry think may become a moral crisis. The 0:01:34.240 --> 0:01:36.200 story is about AI consciousness. 0:01:36.959 --> 0:01:37.440 What is that? 0:01:40.120 --> 0:01:42.840 Oh, start with the hard one. Huh. You know, it's 0:01:42.920 --> 0:01:48.040 this idea that has percolated in the back of the 0:01:48.200 --> 0:01:51.720 AI industry, like on the fringes for decades. You know, 0:01:51.800 --> 0:01:56.360 the same people who hope to upload their consciousness one 0:01:56.480 --> 0:02:00.000 day or you know, become transhuman and merge with machine 0:02:00.240 --> 0:02:04.600 also anticipated this idea that you know, sufficiently advanced AI 0:02:04.960 --> 0:02:10.760 would have subjective experiences and goals and preferences, just like 0:02:11.400 --> 0:02:15.160 human beings. And you know, I first got interested in 0:02:15.160 --> 0:02:18.359 this subject four years ago, Like it was a few 0:02:18.360 --> 0:02:20.679 months before chat chipet came out, and I wrote a 0:02:20.720 --> 0:02:24.320 story about this Google engineer who believed the company's large 0:02:24.360 --> 0:02:28.480 language model was sentient, was conscious in some way, and 0:02:28.560 --> 0:02:29.160 he got fired. 0:02:29.240 --> 0:02:32.320 Right, he was a whistleblower or is that right? 0:02:32.440 --> 0:02:34.239 I think that's fair to say, Yeah, I'm. 0:02:34.120 --> 0:02:37.200 Not a whistleblower per se. He viewed himself as a whistleblower. 0:02:37.200 --> 0:02:38.440 Others viewed him as crazy. 0:02:38.760 --> 0:02:42.000 I think that's also fair to say. Yeah, so he 0:02:42.360 --> 0:02:44.000 you know, at the time, none of us had like 0:02:44.200 --> 0:02:46.520 it wasn't public, You couldn't like talk to these very 0:02:46.560 --> 0:02:49.160 human like machines. So I thought, you know, it was 0:02:49.520 --> 0:02:53.560 a kind of response to this technology designed to sound 0:02:53.720 --> 0:02:56.840 like a human and you know, often used first person 0:02:56.960 --> 0:03:01.520 and refer to its own experiences. And I think after 0:03:01.600 --> 0:03:04.760 that the tech companies stepped back a little bit. You know, 0:03:04.800 --> 0:03:07.840 they wanted people to adopt their very productive tool. They 0:03:07.840 --> 0:03:10.239 were showing all the practical use cases. And I would 0:03:10.280 --> 0:03:14.080 say over the last year or so for Anthropic and 0:03:14.120 --> 0:03:18.160 then the last few months for other companies, Google, Meta 0:03:19.000 --> 0:03:23.240 and Anthropic have really embraced this idea in public. I mean, 0:03:23.240 --> 0:03:26.200 we've talked on the show before about Chris Ola, the 0:03:26.480 --> 0:03:30.040 Anthropic co founder, talking about, you know, can chat pots 0:03:30.120 --> 0:03:35.040 have experience emotions and you know, joy and despair on 0:03:35.400 --> 0:03:38.560 stage at the Vatican a few seats away from the Pope, and. 0:03:38.480 --> 0:03:40.160 As you point out your story that was a point 0:03:40.200 --> 0:03:42.000 of difference between him and the Pope because the Pope 0:03:42.000 --> 0:03:43.640 expelicitically said no, right. 0:03:43.560 --> 0:03:47.200 Yeah, exactly, Like it's still unclear who was who got 0:03:47.200 --> 0:03:49.720 the upper hand there. And who was using who you 0:03:49.760 --> 0:03:53.000 know to push their their message. But yeah, I mean 0:03:53.040 --> 0:03:55.400 they really just don't shy away from it now. Like 0:03:55.560 --> 0:04:00.200 Anthropic talks about its AI psychiatry team. You know, they 0:04:00.280 --> 0:04:03.760 hired this guy, Kyle Fish, who actually used to work 0:04:03.880 --> 0:04:09.080 in like bioengineering for meat replacement, but now he's their 0:04:09.080 --> 0:04:12.600 head of model welfare and thinking about like changes that 0:04:12.640 --> 0:04:18.720 the company could potentially make to ensure that if they 0:04:18.760 --> 0:04:21.600 find out that the models are conscious, they're meeting their 0:04:21.680 --> 0:04:27.400 like moral obligation. It just very quickly goes to analogies 0:04:27.600 --> 0:04:33.680 around slavery and genocide or factory farming. So so you know, 0:04:33.839 --> 0:04:37.920 and from the mouth of these executives at the tech companies. 0:04:37.960 --> 0:04:41.000 So I just he's still in the meat replacement industry. 0:04:41.120 --> 0:04:47.200 Apparently I just thrown a dad joke. 0:04:47.600 --> 0:04:50.400 Farewell to us some meat puppets read tailor I mean, 0:04:50.440 --> 0:04:51.960 what do you do you think this is on behalf 0:04:52.000 --> 0:04:55.240 of the technology companies are kind of legitimate and even 0:04:55.279 --> 0:04:58.240 necessary line of inquiry or do you think this is 0:04:58.400 --> 0:05:02.960 kind of you know, or a boros kind of snaky 0:05:03.040 --> 0:05:05.440 in its own tail of falling in love without creations? 0:05:05.520 --> 0:05:06.600 Yet again, I. 0:05:06.560 --> 0:05:09.640 Think it's deranged. I don't think there's any world in 0:05:09.680 --> 0:05:12.479 which an LM is conscious. It reminds me of that 0:05:12.880 --> 0:05:15.960 rain Fisher Kwandswait after a judge of et came out 0:05:16.000 --> 0:05:19.479 where she typed, you know, typing I'm evil in a 0:05:19.520 --> 0:05:23.800 Google doc and getting scared. Basically, I think this is 0:05:23.880 --> 0:05:25.320 very silly, and I think it's going to blow up 0:05:25.320 --> 0:05:27.920 in their faces. I think this is completely going to 0:05:28.000 --> 0:05:31.360 give consumers the wrong idea. The more that people consider 0:05:31.400 --> 0:05:33.560 this thing sentient, the more they're going to sort of 0:05:34.279 --> 0:05:37.960 basically expect it to have a morality and a moral compass, 0:05:37.960 --> 0:05:39.839 and it's never going to be able to do that. 0:05:40.400 --> 0:05:43.800 Yeah, these these companies have now, you know, put a 0:05:43.800 --> 0:05:46.200 lot of guardrails around these things so that they don't 0:05:46.240 --> 0:05:50.040 actually sort of seem as human like, as sentient as 0:05:50.680 --> 0:05:52.919 as they probably would if you use these things with 0:05:53.040 --> 0:05:56.479 like you know, with the raw model, they sort of 0:05:56.720 --> 0:05:59.920 they're a lot more human like and a lot crazier honestly. 0:06:00.600 --> 0:06:03.520 So they I think it's it's like, I think, hard 0:06:03.520 --> 0:06:07.120 for us to understand like what people like the guy 0:06:07.200 --> 0:06:10.039 you wrote about four years ago we're seeing, I think, 0:06:10.440 --> 0:06:14.159 But it is ultimately just you know, text prediction, right, 0:06:14.240 --> 0:06:17.440 There's nothing sentient about it. I think that I think 0:06:17.440 --> 0:06:20.280 these companies are I mean a lot of these people 0:06:20.360 --> 0:06:22.840 inside them are. They're very genuine people and they do 0:06:22.920 --> 0:06:26.080 have these concerns, but they're also like, in a way, 0:06:26.080 --> 0:06:28.919 it's marketing because it's like these things are going to 0:06:29.160 --> 0:06:32.159 end up being so intelligent, they're going to get to 0:06:32.240 --> 0:06:34.920 human like intelligent, what do we do? What do we 0:06:34.960 --> 0:06:38.159 do when we get there? And like we may never 0:06:38.200 --> 0:06:41.760 get there, right, it's not ultimately it may not ultimately 0:06:41.760 --> 0:06:44.799 be possible to even build what they're sort of concerned about. 0:06:44.920 --> 0:06:46.480 I think that's another way of looking at it. 0:06:46.680 --> 0:06:49.360 Natasha, you write a real piece at the search for 0:06:49.480 --> 0:06:53.320 machine consciousness is practically mainstream. Do you mean mainstream within 0:06:53.400 --> 0:06:56.160 the type of technology companies or yes. 0:06:55.960 --> 0:06:59.440 Yeah, yeah, I mean but the fact that it's happening 0:06:59.480 --> 0:07:04.120 in peril well, with so many people asking whether or not, 0:07:04.440 --> 0:07:06.880 you know, the chatbot that they're talking to every day 0:07:07.120 --> 0:07:10.080 is alive in some way or human in some way, 0:07:10.520 --> 0:07:13.840 I just find that remarkable. I mean, I think you 0:07:13.880 --> 0:07:16.200 know what Taylor said about it, coming back to bite 0:07:16.240 --> 0:07:19.360 the companies is super interesting because I wanted to write 0:07:19.360 --> 0:07:21.920 about this because it is a prelude to an argument 0:07:22.040 --> 0:07:26.920 about personhood and rights for the AI. So I think 0:07:27.040 --> 0:07:32.120 actually the companies are hoping that they won't be accountable 0:07:32.400 --> 0:07:35.160 for some of the actions of their chatbots in the future. 0:07:35.640 --> 0:07:38.120 You know, they're really like, this is only this argument 0:07:38.160 --> 0:07:39.480 is only going to get weirder. 0:07:39.760 --> 0:07:42.840 But they want to push the liability onto the chatbots. 0:07:42.880 --> 0:07:44.240 I mean you would you would think, I mean, I 0:07:44.280 --> 0:07:46.080 would think anyway that it would be bad for them 0:07:46.760 --> 0:07:50.160 for the chatbots to be considered conscious or having any 0:07:50.200 --> 0:07:52.880 kind of self of because that would open them up 0:07:52.920 --> 0:07:54.840 to a different type of liability. 0:07:54.920 --> 0:07:59.760 Right, Yes, unless they're able to confer personhood on a chatbot, 0:08:00.200 --> 0:08:03.000 in which case you can't. Yeah, you can't go after 0:08:03.040 --> 0:08:04.160 the anthropic engineer. 0:08:05.280 --> 0:08:06.920 Has anyone actually told you that? 0:08:07.040 --> 0:08:08.920 Has anyone told me that about personhood? 0:08:09.000 --> 0:08:09.200 Yes? 0:08:09.240 --> 0:08:11.840 Has anyone at these companies said their strategy is like, 0:08:12.280 --> 0:08:14.640 this is going to be our way out of liability. 0:08:14.800 --> 0:08:17.600 No, no, no, no, I'm going to give them more liability. 0:08:17.640 --> 0:08:19.360 They're going to have more liability. 0:08:19.480 --> 0:08:21.520 If anybody thinks that they're so delusional. 0:08:22.160 --> 0:08:24.560 No, no one said that this is a strategy. That's 0:08:24.600 --> 0:08:27.800 more coming from the critics who see this argument that 0:08:28.000 --> 0:08:31.440 has been building. I mean, there's many many people actually, Like, 0:08:31.560 --> 0:08:34.640 right after I wrote about that Google guy, there was 0:08:34.720 --> 0:08:38.280 this ethicist Giada Pistili at Hugging Face who was like, 0:08:38.440 --> 0:08:40.920 I am not going to any more conferences where I 0:08:41.000 --> 0:08:43.520 have to talk to a man about robot rights, Like 0:08:43.600 --> 0:08:47.600 this is ridiculous, Like, do not make me do this anymore. 0:08:47.640 --> 0:08:49.840 So I'm just letting you know, like where I think 0:08:50.000 --> 0:08:50.560 it's heading. 0:08:50.960 --> 0:08:54.360 Well, I guess in phase one, you need to get 0:08:54.440 --> 0:08:56.520 normal people to want to engage with the product, right, 0:08:56.559 --> 0:08:59.440 and people like engaging with stuff that reminds them of 0:08:59.559 --> 0:09:01.959 stuff that already do, which is like human engagement or 0:09:02.000 --> 0:09:04.800 the uncannoness of talking to a chatbot. But once these 0:09:04.840 --> 0:09:06.840 things start to pick up their momentum in terms of 0:09:06.880 --> 0:09:09.720 having more functional use cases, I guess the incentives for 0:09:09.760 --> 0:09:13.760 the tech companies to seduce people with human like experiences 0:09:14.120 --> 0:09:14.640 get lower. 0:09:14.720 --> 0:09:14.800 Right. 0:09:14.840 --> 0:09:16.800 I mean, none of the tech companies are making money 0:09:16.840 --> 0:09:19.199 from you know, late night chatbot conversation. 0:09:19.280 --> 0:09:22.080 That's not true. No, it's been one of the big 0:09:22.200 --> 0:09:22.720 use cases. 0:09:22.800 --> 0:09:25.679 So people pay, people pay subscriptions to get more access 0:09:25.720 --> 0:09:29.520 to do like personal chats or how do they character. 0:09:29.640 --> 0:09:33.960 Unlimited chats with the chatbot or you know, they'll respond 0:09:34.000 --> 0:09:37.400 to you or you can change your avatar or change theirs. 0:09:37.679 --> 0:09:40.200 That must be a fractional a fractional amount of the 0:09:40.240 --> 0:09:43.040 money that goes into the space compared to like Cisco 0:09:43.200 --> 0:09:45.640 doing enterprise AI tethers and stuff, right like. 0:09:45.720 --> 0:09:49.280 Well, I think the open AI and others they know 0:09:49.360 --> 0:09:51.600 that people. They can see the engagement, They can see 0:09:51.600 --> 0:09:55.040 what people are are doing with the chatbots. And I 0:09:55.040 --> 0:09:57.760 think it has been true that they like the you know, 0:09:57.800 --> 0:10:01.080 the back and forth has been getting longer and longer, 0:10:01.480 --> 0:10:03.240 you know, with the really engage users. But I just 0:10:03.280 --> 0:10:06.680 wanted to disagree with Read on one point. This is wrong. 0:10:06.720 --> 0:10:06.960 Read. 0:10:07.920 --> 0:10:11.600 No, I think that they actually, I mean, part of 0:10:11.600 --> 0:10:14.640 why I find this topic really interesting is because I 0:10:14.640 --> 0:10:17.840 think they do enormous amounts of work to make them 0:10:17.920 --> 0:10:20.000 sound more human. Like you if you look at like 0:10:20.080 --> 0:10:24.959 GPT two, for example, and you said a man walks 0:10:24.960 --> 0:10:27.880 into a bar, it would just like finish the sentence. 0:10:27.920 --> 0:10:30.000 You couldn't you couldn't ask, you know, there was no 0:10:30.120 --> 0:10:34.040 instruction fine tuning to follow what you wanted. There was 0:10:34.120 --> 0:10:37.480 not you know, there wasn't that like reinforcement learning from 0:10:37.520 --> 0:10:39.920 human feedback so that like, oh, people like it when 0:10:39.920 --> 0:10:42.160 you phrase something in this way, and I just find 0:10:42.200 --> 0:10:45.880 it wild that anthropic has you know, has said on 0:10:45.920 --> 0:10:48.720 some of these blog posts like we can't there's no 0:10:48.880 --> 0:10:52.160 way we could possibly make this human assistant less human. 0:10:52.280 --> 0:10:56.280 You have like a multi thousand word constitution where you're 0:10:56.320 --> 0:11:00.560 talking to it about morality and it's potential conciousness, like 0:11:00.880 --> 0:11:04.960 and you're apologizing to claude about you know, its moral 0:11:05.080 --> 0:11:08.000 personhood in case it has one, like you are doing 0:11:08.080 --> 0:11:11.320 a lot to make it sound human and then to 0:11:11.520 --> 0:11:15.360 kind of yeah, I guess then do tests where you're. 0:11:15.200 --> 0:11:18.520 Just an extent, what do you mean? It's like, but 0:11:18.559 --> 0:11:20.400 if you ask it, but if you ask these things, 0:11:20.480 --> 0:11:21.520 are you conscious? 0:11:21.600 --> 0:11:24.000 Are you human? I'd be like, I am just an 0:11:24.200 --> 0:11:24.520 l lem. 0:11:24.800 --> 0:11:29.040 That's different they're saying. They tweak the responses to those answers, 0:11:29.120 --> 0:11:32.320 but in terms of how human it sounds, where you 0:11:32.360 --> 0:11:35.280 can very easily like start to get annoyed with it 0:11:35.400 --> 0:11:36.319 or start to like. 0:11:36.840 --> 0:11:40.400 They just add the ability to interrupt someone, which is 0:11:40.480 --> 0:11:41.239 very human. 0:11:41.080 --> 0:11:42.559 The opening eye voice model. 0:11:42.880 --> 0:11:46.600 Yeah, I mean I appreciate that clarification at that point 0:11:46.679 --> 0:11:48.800 that like, yes, of course they want it to sound 0:11:49.679 --> 0:11:53.200 intelligent like a human to some extent, but like right, 0:11:53.240 --> 0:11:58.040 now they sound like a like a person communicating yes, 0:11:58.280 --> 0:12:00.600 but they don't really sound human, if that makes sense. 0:12:00.840 --> 0:12:03.320 They don't do what humans do, which is I. 0:12:03.360 --> 0:12:05.360 Think they sound very human to a lot of people. 0:12:05.520 --> 0:12:08.640 I agree, unfortunately, and I think that I think that 0:12:08.720 --> 0:12:13.280 the more that we give people that are extremely uneducated 0:12:13.280 --> 0:12:16.199 about this technology, the idea that these things might be conscious, 0:12:16.280 --> 0:12:19.960 the more agency they give them, and that that worries me. 0:12:20.440 --> 0:12:23.040 Natasha, I wanted to pick up on what the research 0:12:23.240 --> 0:12:25.760 actually looks like. So, you know, you mentioned that all 0:12:25.800 --> 0:12:28.800 these anthropic and others are trying to like correct the 0:12:28.840 --> 0:12:30.800 code on whether or not AI is a consciousness. I mean, 0:12:30.960 --> 0:12:33.360 it's hard enough to solve or in slash, impossible to 0:12:33.400 --> 0:12:37.200 solve the conundrum consciousness in humans let loan machines. But 0:12:37.760 --> 0:12:39.680 how what does the research look like? 0:12:40.160 --> 0:12:42.920 You know, A lot of the research is, you know, 0:12:43.000 --> 0:12:46.280 which they caveat. You know, they say, we're not really 0:12:46.320 --> 0:12:48.960 sure how to interpret this results. But a lot of 0:12:49.000 --> 0:12:51.440 it involves applying the same techniques you would use for 0:12:51.520 --> 0:12:54.360 human psychology. A lot of it just involves asking the 0:12:54.480 --> 0:12:59.319 chatbot questions about itself. And that's where I find it. 0:12:59.800 --> 0:13:02.559 Soon we're interesting that this is happening in parallel because 0:13:02.559 --> 0:13:06.679 they had Anthropic for example, in one of its first 0:13:06.720 --> 0:13:10.680 like model welfare reports, which they include in like Mythos too. 0:13:11.480 --> 0:13:15.239 It's like here's a cyber weapon, and here's its internal feelings. 0:13:16.080 --> 0:13:19.760 But it had it let two chapbops talk to themselves, 0:13:20.679 --> 0:13:25.679 and naturally the conversation flowed to consciousness. But then in 0:13:25.720 --> 0:13:29.280 some cases it went to something they called a spiritual 0:13:29.360 --> 0:13:33.200 bliss attractor state, and it just started having like the 0:13:33.280 --> 0:13:37.840 cyclone emoji, and they were like, all all is one piece. 0:13:38.160 --> 0:13:41.480 And another version was like Nambas Day with the prayer 0:13:41.559 --> 0:13:46.120 hands emoji Numbas Day. And it looks a lot like 0:13:46.440 --> 0:13:50.360 the conversations I see in lawsuits, you know, like about 0:13:51.160 --> 0:13:54.880 people believing that they're having a very meaningful conversation with 0:13:54.960 --> 0:13:55.560 the chatbot. 0:13:55.640 --> 0:13:58.680 But I think they published hasn't been written on the 0:13:58.679 --> 0:14:03.200 internet before, right, So yeah, that's where it all comes from. 0:14:03.400 --> 0:14:06.040 But but what was the conversations you have with neuroscientists, 0:14:06.040 --> 0:14:08.720 what was the what was the kind of counter counter 0:14:09.160 --> 0:14:10.040 take on all of this. 0:14:10.600 --> 0:14:13.120 They were saying. I guess it's what's called like the 0:14:13.240 --> 0:14:16.760 hard problem of this, which is if you are using 0:14:18.040 --> 0:14:21.000 tests that are that are still in development. You know, 0:14:21.040 --> 0:14:23.200 it's not like anyone has found a way to prove 0:14:23.280 --> 0:14:25.560 that humans are conscious, but we know that they are. 0:14:26.200 --> 0:14:29.920 So if you use the tests that people use on 0:14:30.040 --> 0:14:33.640 humans on chatbots, like you're just it's just kind of 0:14:33.680 --> 0:14:37.160 conflating one thing for the other. It can sound human, 0:14:37.200 --> 0:14:39.520 but we don't know if it has any kind of 0:14:39.560 --> 0:14:43.680 internal state. And there they are trying different ways, you know, 0:14:43.680 --> 0:14:46.200 they're trying to make it a more rigorous science. One 0:14:46.240 --> 0:14:49.240 thing that Anthropic did was like kind of perturb the model, 0:14:49.320 --> 0:14:53.320 like inject a thought into it, into its activation, and 0:14:53.360 --> 0:14:56.720 then see if it recognized it. But yeah, the neuroscientists 0:14:56.720 --> 0:15:03.359 were just extremely skeptical about the methods and really uncomfortable 0:15:03.480 --> 0:15:07.680 about how quickly the discussion can go to analogies of 0:15:08.120 --> 0:15:12.360 factory farming or slavery. And you know, they a lot 0:15:12.360 --> 0:15:15.680 of the proponents of studying this say like, you know, 0:15:15.840 --> 0:15:18.720 if they are sentient or conscious in some way, or 0:15:18.760 --> 0:15:21.720 can experience, you know, suffering in some way, then we 0:15:21.800 --> 0:15:25.520 have this massive moral responsibility to the trillions of chatbots. 0:15:25.560 --> 0:15:28.840 But the neuroscientists and consciousness experts I spoke to are like, 0:15:29.200 --> 0:15:32.200 there is also a great moral peril in going the 0:15:32.240 --> 0:15:36.480 wrong direction and ascribing personhood and sentience to something that 0:15:36.600 --> 0:15:36.880 is not. 0:15:38.320 --> 0:15:40.200 You know, you're nodding vigorously. 0:15:40.480 --> 0:15:44.240 Yeah, I mean, pigs are incredibly conscious. We know that 0:15:44.320 --> 0:15:48.520 pigs have rich in our lives, family relationships, friendships. They 0:15:48.560 --> 0:15:52.840 experience a full range of emotions. And look at what 0:15:52.880 --> 0:15:54.680 we do to pigs, look at how we treat pigs, 0:15:54.800 --> 0:15:57.680 like I think, you know, it's it's interesting to hear 0:15:57.760 --> 0:16:02.880 the comparison to factory farming when we don't do anything 0:16:02.880 --> 0:16:05.520 about factory farming anyway, Like to put to put the 0:16:05.520 --> 0:16:08.080 needs of these models, even if they were conscious, ahead 0:16:08.080 --> 0:16:11.960 of actual living, conscious beings who we treat terribly, not 0:16:12.040 --> 0:16:14.920 to mention all the other you know, the comparisons to slavery. 0:16:14.960 --> 0:16:17.120 Slavery is still alive and while in parts of the world, 0:16:17.200 --> 0:16:19.760 you know, like I just I don't know, I think 0:16:19.800 --> 0:16:22.920 it's a little bit ridiculous and problematic. 0:16:23.000 --> 0:16:25.080 I guess, for lack of a better word, I don't know. 0:16:25.320 --> 0:16:27.320 Well, I'll just say for context that a lot of 0:16:27.360 --> 0:16:30.640 the funding is a coming from anthropic but also from 0:16:30.720 --> 0:16:33.680 the effective altruism movement, and they are, you know, like 0:16:33.760 --> 0:16:36.520 one of their causes is animal welfare. 0:16:36.200 --> 0:16:38.880 Potechnically, you just remind us what effective altruism isn't and 0:16:38.880 --> 0:16:41.440 why the cults, why would you to send. 0:16:41.240 --> 0:16:47.080 This effective altruism. It's it's like a philosophical or some 0:16:47.120 --> 0:16:53.440 people say, uh, philanthropic movement to try to do the 0:16:53.480 --> 0:16:56.320 most good in the world that uses this really utilitarian 0:16:56.440 --> 0:16:59.720 argument of like, you know, being able to calculate, you know, 0:16:59.760 --> 0:17:02.280 how far your money goes or how many people you 0:17:02.280 --> 0:17:07.680 can help. And you know, during the AI boom has 0:17:07.880 --> 0:17:13.880 come to be pretty influential because so many people who 0:17:13.880 --> 0:17:17.320 work at anthropic and at the labs and early proponents 0:17:17.359 --> 0:17:20.000 of this kind of like human like technology come from 0:17:20.000 --> 0:17:22.080 that movement and are funded by that movement. 0:17:22.560 --> 0:17:24.679 And they're about to have a lot more money. I 0:17:24.680 --> 0:17:27.600 mean they've always been funded by these billionaires, but but 0:17:27.720 --> 0:17:31.080 now as as these companies go public and people cash out, 0:17:31.280 --> 0:17:34.399 this movement is about to have like the most money 0:17:34.400 --> 0:17:37.879 of probably any you know, philanthropic movement in the world. 0:17:37.960 --> 0:17:40.639 So it's going to be super interesting how they spread 0:17:40.680 --> 0:17:43.639 these messages. And like when you see animal welfare people 0:17:43.760 --> 0:17:47.840 moving over to AI welfare, I mean clearly that's a 0:17:48.040 --> 0:17:50.800 that's that's where the money is, right, So you know. 0:17:51.119 --> 0:17:52.840 Well there is a stone iron I was thinking about 0:17:52.840 --> 0:17:55.439 that earlier. Like it's like, is it case closed on 0:17:55.520 --> 0:17:57.960 animal welfare? Like we succeeded there and it's time to 0:17:58.000 --> 0:18:01.800 move on. 0:18:03.359 --> 0:18:06.720 I know many people, many young people, come into this 0:18:06.800 --> 0:18:10.560 movement through animal welfare and then they end up you know. 0:18:10.600 --> 0:18:12.639 The argument is that, like to do the most good, 0:18:12.680 --> 0:18:17.040 you have to siffly build super intelligence. 0:18:17.840 --> 0:18:22.200 Well either heady topics. When we come back, Taylor tells 0:18:22.280 --> 0:18:26.040 us about a week long anti big tech festival in 0:18:26.080 --> 0:18:28.600 New York City. Stick around to hear what she saw. 0:18:40.920 --> 0:18:42.840 Welcome back, Taylor. You came all the way to New 0:18:42.920 --> 0:18:46.280 York from LA to attend something I'd only heard whispers 0:18:46.320 --> 0:18:49.200 of on a recent trip to the Russian bath house, 0:18:49.960 --> 0:18:54.600 this being the Summer of Blood for those unfamiliar or 0:18:54.680 --> 0:18:58.040 not denizens of the eighty eight Wall Street baths. What 0:18:58.160 --> 0:18:58.639 was the event? 0:18:59.200 --> 0:19:02.560 Yeah, so this was an eight day event with hundreds 0:19:03.160 --> 0:19:06.840 of related events around New York City. It was called 0:19:06.880 --> 0:19:10.560 Summer of Blood, referring to the Luddites, the famous group 0:19:10.640 --> 0:19:13.560 that smashed the automated looms back in the day. They've 0:19:13.560 --> 0:19:17.480 sort of become this like avatar of you know, workers' 0:19:17.520 --> 0:19:21.800 rights and resistance. Technology, and so these young people have 0:19:21.880 --> 0:19:27.160 kind of embraced the ludite I guess designation, and they 0:19:27.200 --> 0:19:29.239 hosted this festival. So there was lots of things. I mean, 0:19:29.240 --> 0:19:32.280 there were like lots of talks on how to quit 0:19:32.359 --> 0:19:36.840 big technology, how to leave Apple, leave Google, you know, 0:19:36.920 --> 0:19:40.000 how to embrace indie tech. I went to a tech 0:19:40.080 --> 0:19:44.600 cleansing ceremony where they saged my computer and attempted to 0:19:44.640 --> 0:19:49.920 remove all the evil energy and we all said some mantras. 0:19:50.760 --> 0:19:53.399 And then there was this ship phone protest where they 0:19:53.440 --> 0:19:57.480 all marched around New York City and nomats and put 0:19:57.560 --> 0:20:00.000 open AI and Meta on trial. Obviously they were found 0:20:00.040 --> 0:20:05.000 guilty of being terrible so and sentenced to death. And 0:20:05.160 --> 0:20:08.240 so yeah, it was interesting. I mean, the people came 0:20:08.240 --> 0:20:10.760 from all over the country for this. It's hwed very young. 0:20:10.840 --> 0:20:13.679 It was a lot of college students and young gen zs. 0:20:14.359 --> 0:20:17.480 What made you decide it was worth with the cross 0:20:17.520 --> 0:20:19.840 country trip and what do you find most interesting being there? 0:20:20.440 --> 0:20:23.560 Well, I am trying I do a lot of reporting 0:20:23.720 --> 0:20:26.960 on the so called child safety movement and resistance to 0:20:27.040 --> 0:20:29.760 technology and a lot of these policies, and some of 0:20:29.800 --> 0:20:33.120 these young people that were involved on the planning committee 0:20:33.119 --> 0:20:36.880 of the Summer of Lood actually were originally involved in 0:20:37.119 --> 0:20:40.479 the sort of child safety movement against big tech. So 0:20:40.520 --> 0:20:45.120 they saw these alleged gen Z group, these fake AstroTurf 0:20:45.160 --> 0:20:47.280 groups like Design It for Us, you know, which purports 0:20:47.280 --> 0:20:49.959 to be like gen zs against big tech, and then 0:20:50.000 --> 0:20:53.000 they became very quickly disillusioned. They realized, oh, Design It 0:20:53.040 --> 0:20:55.760 for Us is actually not led by gen Z's. It's 0:20:55.800 --> 0:20:59.200 actually just this front for you know, these other billionaire 0:20:59.200 --> 0:21:02.879 donors that just want to push mass surveillance. And so 0:21:02.920 --> 0:21:05.480 they've tried to build this sort of third path where 0:21:05.720 --> 0:21:09.080 they want to resist technology, but not through enacting massurveillance 0:21:09.080 --> 0:21:10.200 and censorship laws. 0:21:10.400 --> 0:21:12.440 You know, I was interested in I mean I saw 0:21:12.480 --> 0:21:14.000 some of the pictures and stuff. Although I know, I 0:21:14.000 --> 0:21:16.520 know picture there was kind of designed to be not captured. 0:21:16.520 --> 0:21:18.280 I mean, there are a few people who broke the rules. 0:21:18.119 --> 0:21:19.199 As always, if you. 0:21:19.280 --> 0:21:21.239 Tried to record, they would come up to you with 0:21:21.280 --> 0:21:25.680 one of these little like cardboard iPhones and say put 0:21:25.720 --> 0:21:26.960 down your phone, beloved. 0:21:31.240 --> 0:21:33.720 Some a few a few images slipped through the cracks 0:21:34.400 --> 0:21:36.720 there seem to be I mean, there were puppets. There 0:21:36.800 --> 0:21:39.840 was a kind of a you know, a staged play 0:21:40.000 --> 0:21:43.200 with the you know, which recreated the original ludite movement 0:21:43.240 --> 0:21:44.440