WEBVTT - Ep88 "Might there exist very different kinds of minds?" 0:00:05.160 --> 0:00:09.680 Why is it so hard to define intelligence? And what 0:00:09.720 --> 0:00:12.399 does this have to do with being a fish in 0:00:12.600 --> 0:00:17.560 water trying to describe water? Might we humans possess one 0:00:17.760 --> 0:00:22.200 kind of intelligence in a constellation of many other types? 0:00:22.680 --> 0:00:25.320 And what does this have to do with empathy or 0:00:25.400 --> 0:00:32.760 artificial intelligence or our search for extraterrestrial life. Welcome to 0:00:32.760 --> 0:00:35.920 Inner Cosmos with me David Eagleman. I'm a neuroscientist and 0:00:35.960 --> 0:00:39.160 an author at Stanford and in these episodes we sail 0:00:39.320 --> 0:00:42.800 deeply into our three pound universe to understand some of 0:00:42.840 --> 0:00:56.880 the most surprising aspects of our lives. Today's episode is 0:00:56.880 --> 0:01:01.160 about minds and whether there might exist very different kinds 0:01:01.200 --> 0:01:03.880 of minds. So I'm going to boot this up with 0:01:04.000 --> 0:01:07.360 a prediction that I made last year. I've always noticed 0:01:07.400 --> 0:01:11.400 that some people feel annoyed when they ask chat GPT 0:01:11.560 --> 0:01:14.960 a question, Let's say a political question like was so 0:01:15.040 --> 0:01:18.880 and so a good president and chat GPT answers with 0:01:18.959 --> 0:01:22.639 something neutral like Some people feel this way. Some people 0:01:22.680 --> 0:01:26.080 feel that way. More discussion and debate is needed here, 0:01:26.440 --> 0:01:28.759 and a lot of people who feel that they hold 0:01:29.160 --> 0:01:33.160 very clear political stances they criticize chat GPT for this. 0:01:33.319 --> 0:01:37.080 They say, look, it feels wishy wash. It's not taking 0:01:37.080 --> 0:01:40.720 a stand on anything. It is stuck in neutral. It's 0:01:40.800 --> 0:01:44.120 not brave enough to take a position. But I feel 0:01:44.520 --> 0:01:48.520 these sorts of answers from lms are the sign that 0:01:48.560 --> 0:01:51.960 we are currently living in the golden era of AI. 0:01:52.520 --> 0:01:56.280 And this golden era is sure to end, just like 0:01:56.320 --> 0:01:59.760 the golden age of the Pax Romana. So the prediction 0:01:59.800 --> 0:02:03.200 i'm year ago on this podcast is that people on 0:02:03.240 --> 0:02:07.440 the far ends of the political spectrum will start getting frustrated. 0:02:07.800 --> 0:02:10.840 How come this AI isn't telling me the true answer, 0:02:10.880 --> 0:02:14.160 the answer that I can so clearly see, and anybody 0:02:14.240 --> 0:02:17.080 who is saying would clearly agree with me. So my 0:02:17.160 --> 0:02:21.560 prediction is that quite soon, as soon as the architectures 0:02:21.639 --> 0:02:24.880 drop in cost for training these models, we're going to 0:02:24.880 --> 0:02:29.320 see the far left progressives training their own model and 0:02:29.400 --> 0:02:33.320 the far right conservatives training their own model, and both 0:02:33.400 --> 0:02:37.799 sides will say, look, we don't want this garbage literature 0:02:38.200 --> 0:02:41.160 to pollute the training data, so we're gonna leave all 0:02:41.520 --> 0:02:45.120 this stuff out and just include the writing that is 0:02:45.200 --> 0:02:49.720 consistent with the truth. And obviously anything we disagree with 0:02:49.840 --> 0:02:53.080 comes from people who are just trolls or obstreperous or 0:02:53.080 --> 0:02:56.840 at minimum are badly misinformed, and if they'll just come 0:02:56.919 --> 0:03:00.440 to read the AI wisdom that we know to be true, 0:03:00.880 --> 0:03:03.920 then they will see the light of our way and 0:03:04.000 --> 0:03:08.160 the error of theirs. So this is my prediction about 0:03:08.160 --> 0:03:12.440 the Balkanization of AI that will come about. But perhaps 0:03:12.720 --> 0:03:16.079 there's a broader way to think about this. Perhaps there 0:03:16.120 --> 0:03:21.040 are indeed different types of artificial intelligence, not just in 0:03:21.080 --> 0:03:24.560 terms of how they're trained up, but more fundamentally about 0:03:24.880 --> 0:03:30.120 their architecture. Perhaps there are very different ways to think. 0:03:30.800 --> 0:03:34.800 In other words, what if there are different types of intelligence, 0:03:34.920 --> 0:03:39.440 not just one artificial intelligence, and also many different types 0:03:39.480 --> 0:03:44.200 of natural intelligence. Now on this topic, there's no one 0:03:44.240 --> 0:03:46.680 better in the world to talk with than my friend 0:03:46.960 --> 0:03:49.440 Kevin Kelly, who's been chewing on this for a while. 0:03:50.040 --> 0:03:52.680 Kevin is one of the fathers of Silicon Valley, not 0:03:52.840 --> 0:03:55.760 because he's a techie, but instead because he's one of 0:03:55.800 --> 0:04:01.040 the deepest philosophers and connoisseurs of technology. He has a 0:04:01.120 --> 0:04:03.960 long and storied history that I'll link on the show notes, 0:04:04.000 --> 0:04:06.480 but I'll just mention for now that he's the founding 0:04:06.520 --> 0:04:10.560 executive editor of Wired magazine and a former editor of 0:04:10.600 --> 0:04:14.040 The Whole Earth Review. He's the founder of the cool 0:04:14.120 --> 0:04:17.160 Tools website, and he's one of the founders of the 0:04:17.200 --> 0:04:20.880 Long Now Foundation. He's written multiple best selling books about 0:04:20.880 --> 0:04:24.120 the future of technology, which I'll link. And one of 0:04:24.160 --> 0:04:26.159 the many things that I love about Kevin is that 0:04:26.240 --> 0:04:30.840 he's a radical optimist and an extremely creative thinker. So 0:04:30.880 --> 0:04:33.560 I wrung up Kevin to talk with him about his 0:04:33.720 --> 0:04:39.080 view of intelligences. Does intelligence mean just one thing? Or 0:04:39.200 --> 0:04:48.400 might there be lots of different ways it could manifest? So, Kevin, 0:04:48.400 --> 0:04:51.680 we're all talking about AI, and people are talking about AGI, 0:04:51.920 --> 0:04:54.320 and we're aiming towards these things. But you have a 0:04:54.440 --> 0:04:56.839 very interesting view on it, which is that we can't 0:04:56.880 --> 0:05:01.240 look at AI as one thing, but instead there are 0:05:01.320 --> 0:05:03.400 multiple ais we need to be thinking about. So tell 0:05:03.480 --> 0:05:03.880 us about this. 0:05:03.960 --> 0:05:07.080 Yeah, Yeah, I think we should force ourselves to use 0:05:07.120 --> 0:05:11.440 the plural AIS whenever we're talking about this. It's very 0:05:12.240 --> 0:05:15.920 to me, very similar to machines. We have a lot 0:05:15.920 --> 0:05:18.960 of machines in our life, but we don't talk about 0:05:19.040 --> 0:05:22.839 the machine. We don't have a single regulatory apparatus for 0:05:23.080 --> 0:05:26.719 the machine. We don't have a single operating manual. We 0:05:26.839 --> 0:05:31.400 have multiple machines in our lives, and we're going to 0:05:31.480 --> 0:05:36.800 have multiple ais in our lives, and those ais individually 0:05:36.800 --> 0:05:40.440 will have different characters, they'll have different needs, they'll have 0:05:40.480 --> 0:05:45.680 different tasks, they'll have different abilities, they'll be different species. 0:05:46.400 --> 0:05:46.839 Almost. 0:05:46.880 --> 0:05:49.520 We can think of them as different species of mind. 0:05:49.800 --> 0:05:54.400 Something that might be a slow, minimal kind, something that 0:05:54.480 --> 0:05:59.680 might be fast and peculiar, one might be imaginative and 0:05:59.760 --> 0:06:02.120 one dimension, but not others. And so I think the 0:06:02.200 --> 0:06:05.280 fallacy that we have is we imagine intelligence as a 0:06:05.320 --> 0:06:11.360 single dimension, kind of like amplitude or sound decibels kind 0:06:11.360 --> 0:06:13.599 of goes up abup, and there's like a ladder and 0:06:13.640 --> 0:06:16.680 we're kind of climbing up, and we've got the rats 0:06:16.720 --> 0:06:18.599 and you know, the chimp, and then the human, and 0:06:18.640 --> 0:06:20.960 then we've got Ai above us, and that there's this 0:06:21.040 --> 0:06:24.919 this one dimensional thing. But what we're going to discover, 0:06:25.000 --> 0:06:27.560 and we have already with with what we've made so far, 0:06:28.320 --> 0:06:31.000 is that it's multi dimensional. It's a very big space. 0:06:31.600 --> 0:06:37.720 So the space of possible minds is vast, and human 0:06:38.279 --> 0:06:42.640 kind of intelligence we're at the edge. We're an edge species, 0:06:43.200 --> 0:06:45.680 like we're at the edge of the galaxy. We have 0:06:45.720 --> 0:06:51.000 a very peculiar mix of different kind of primitive cognitions 0:06:51.839 --> 0:06:55.360 and we have one species of mind, and the things 0:06:55.400 --> 0:07:00.640 that we're inventing are going to be many other kinds 0:07:00.680 --> 0:07:05.040 of thinking. And the reason why that's important is that 0:07:05.080 --> 0:07:07.719 there may be certain things that we want to do 0:07:07.839 --> 0:07:13.640 and understand that our own kind of intelligence can't by itself, 0:07:14.680 --> 0:07:17.320 but we can understand it with the two step process 0:07:17.920 --> 0:07:22.720 of inventing another kind of intelligence that can work with us, 0:07:23.680 --> 0:07:27.880 so that we can understand or make something that we 0:07:27.920 --> 0:07:29.560 can't do by ourselves. 0:07:31.240 --> 0:07:33.880 Yeah, you know, in neuroscience, this has been one of 0:07:33.920 --> 0:07:35.880 the challenges for the last one hundred years, is even 0:07:35.920 --> 0:07:40.120 defining intelligence for humans. So some people think, well, maybe 0:07:40.120 --> 0:07:43.000 it's about being able to squelch distractors, or some people think, well, 0:07:43.000 --> 0:07:47.120 maybe it's about being able to simulate possible futures, or 0:07:47.160 --> 0:07:50.520 you know, there's twenty theories out there about what intelligence is, 0:07:51.080 --> 0:07:53.440 but it's probably one of these words that has too 0:07:53.520 --> 0:07:57.440 much semantic weight on it is trying to incorporate many things. 0:07:57.880 --> 0:08:00.360 And so your take on this is that the way 0:08:00.400 --> 0:08:02.920 to go about this is to think even more broadly 0:08:02.960 --> 0:08:05.920 about what intelligence could be, what we might mean by it, 0:08:06.120 --> 0:08:10.640 and all the ways that we're going to invent different machines, 0:08:11.520 --> 0:08:14.920 forms of AI to get there. So you recently made 0:08:15.200 --> 0:08:18.360 a list, well, actually in your book The Inevitable, right, 0:08:18.560 --> 0:08:22.560 you made a list of possible minds. Yeah, let's start. 0:08:22.680 --> 0:08:25.760 Let's start there, give us a sense of possible minds 0:08:25.800 --> 0:08:27.880 to get us to expand our brains on this. 0:08:28.480 --> 0:08:30.520 Let me just say one of the things that's very 0:08:30.520 --> 0:08:34.600 common when you talk to people about what they imagined 0:08:35.000 --> 0:08:44.000 super intelligence is something that's super beyond human and what 0:08:44.040 --> 0:08:46.440 does that look like or feel like or how do 0:08:46.480 --> 0:08:50.439 we see it? And the most common response is sort 0:08:50.480 --> 0:08:54.960 of like human thinking but faster. Yeah, billion times faster. 0:08:55.840 --> 0:09:04.560 And okay, that's one version anything else. So there's again 0:09:04.559 --> 0:09:06.920 there's this idea of a singular thing. We have this 0:09:06.960 --> 0:09:09.480 thing and we're gonna make it faster, and that's just 0:09:09.520 --> 0:09:13.760 super and then that's it and I don't think so. 0:09:13.760 --> 0:09:17.800 So part of the chore about the challenge about thinking 0:09:17.800 --> 0:09:22.280 about possible minds is to think about things other than 0:09:22.360 --> 0:09:25.880 just this time element, although that's one of them. And 0:09:25.960 --> 0:09:29.280 so one of the first possible minds is something that 0:09:29.400 --> 0:09:33.520 thinks really really slow. You can imagine it's kind of 0:09:33.679 --> 0:09:37.400 mind waves being so slow that we can't even see it. 0:09:37.440 --> 0:09:40.040 We don't even recognize them, they're kind of really really slow, 0:09:40.720 --> 0:09:43.840 and so in a certain weird way, evolution is that 0:09:44.800 --> 0:09:46.960 evolution is a kind of learning that happens on a 0:09:47.080 --> 0:09:49.840 very large time scale, and we can't see it on 0:09:49.880 --> 0:09:52.959 an every day perspective, but we can see it over time, 0:09:53.320 --> 0:09:58.000 and so there could be slow kinds of intelligences that 0:09:58.080 --> 0:10:00.440 may be more powerful in the sense that they're wider, 0:10:00.800 --> 0:10:01.880 but not its early faster. 0:10:02.000 --> 0:10:04.200 So that's just kind of like the first one. 0:10:04.880 --> 0:10:07.640 One of the challenges again, as you mentioned, is that 0:10:07.679 --> 0:10:11.200 we use the word intelligence, and I think it's a 0:10:11.200 --> 0:10:16.560 form of ignorance in the sense that they're probably well, 0:10:16.960 --> 0:10:22.439 I've been reading about the early days of the discovery 0:10:22.760 --> 0:10:26.600 of electricity or the invention of electricity, and it was 0:10:26.679 --> 0:10:28.920 really early interested. And this is way before Tesla. This 0:10:29.040 --> 0:10:31.920 is like at the Faraday in Davies, where they're trying 0:10:31.960 --> 0:10:36.560 to really understand what it was. And it was remarkable 0:10:36.600 --> 0:10:38.560 because a lot of the smartest people in the world, 0:10:39.520 --> 0:10:44.000 Newton and others, were completely wrong. They had these ideas 0:10:44.000 --> 0:10:48.280 of Flakistan and the ether. They just had no idea 0:10:48.360 --> 0:10:50.640 what it was, and they were really struggling. At the 0:10:50.679 --> 0:10:53.720 same time they were trying to figure out what materials were, 0:10:54.000 --> 0:10:58.120 what substances were before we had the idea of a 0:10:58.160 --> 0:11:05.760 periodic table, So there's stuff. Well, then we eventually understood 0:11:05.760 --> 0:11:10.240 that there were elements and the elements were recombined to 0:11:10.400 --> 0:11:14.319 form compounds, and that there was a fixed number of elements, 0:11:14.559 --> 0:11:16.439 and then you could identify the elements and then they 0:11:16.440 --> 0:11:21.679 were pretty distinct and that you made up. Salt was 0:11:21.720 --> 0:11:24.520 not an element, so it was actually a compound. Water 0:11:24.720 --> 0:11:27.320 was a compound. So you have all these compounds made 0:11:27.360 --> 0:11:30.120 from the elements. And what we're trying to do right 0:11:30.160 --> 0:11:33.320 now with intelligence is like, what are the elements that 0:11:33.400 --> 0:11:38.800 make up this compound that we call intelligence? And so 0:11:39.480 --> 0:11:44.640 there are things like maybe logic and deduction or reasoning, 0:11:44.640 --> 0:11:46.840 which may be different, and then there's memory and short 0:11:46.920 --> 0:11:50.960 term memory. There's a bunch of different elements that we 0:11:51.520 --> 0:11:54.080 don't know about that we don't know and haven't identified 0:11:54.120 --> 0:11:59.440 and can't describe, that are probably fundamental to making this 0:11:59.520 --> 0:12:03.240 thing we call intelligence. And so part of this little 0:12:03.320 --> 0:12:06.840 challenge of imagining possible minds is to think about what 0:12:06.960 --> 0:12:13.000 might those elements be and how might one rearrange them 0:12:13.120 --> 0:12:15.000 to make a different compound. 0:12:15.559 --> 0:12:17.920 That's very good, So tell us give us a sense 0:12:17.960 --> 0:12:20.720 of some other possible minds. Sure, so we can start 0:12:20.720 --> 0:12:22.480 thinking about what the periodic table will look like. 0:12:22.679 --> 0:12:24.800 So I've made a little list and I'm going to 0:12:24.920 --> 0:12:26.520 just use it for the prompt. 0:12:26.600 --> 0:12:27.000 And so. 0:12:28.400 --> 0:12:31.440 One of the elements in our own intelligence is self awareness. 0:12:33.200 --> 0:12:36.640 But it's possible that you could have degrees of intelligence 0:12:36.679 --> 0:12:40.440 without self awareness. So we think somehow self awareness is 0:12:40.960 --> 0:12:43.920 instrumental for high intelligence, but you may be able to 0:12:43.960 --> 0:12:47.200 do a lot of things without any self awareness without 0:12:47.240 --> 0:12:51.439 the entity being a self aware that's what we have now, right, right, 0:12:52.120 --> 0:12:54.840 and vice versa. You might be able to have self 0:12:54.880 --> 0:12:59.160 awareness with very limited intelligence, yeah, okay, and so you 0:12:59.160 --> 0:13:02.200 could have kind of like this thing is really really 0:13:02.240 --> 0:13:04.280 good at doing things, but it's not aware of itself. 0:13:04.280 --> 0:13:08.440 And this thing isn't that capable in other words, but 0:13:08.480 --> 0:13:10.040 it's very aware of itself. 0:13:10.280 --> 0:13:10.679 Okay. 0:13:10.880 --> 0:13:13.160 So self awareness is one of those elements that you 0:13:13.160 --> 0:13:15.920 can kind of conjure with another one would be. One 0:13:15.960 --> 0:13:19.480 of the things about human intelligence is that I think 0:13:19.679 --> 0:13:25.839 we have very deliberately through evolution, restricted our access and 0:13:25.920 --> 0:13:30.679 ability to change ourselves to have access to our operating system. 0:13:29.920 --> 0:13:32.880 Right, because it's really dangerous, that's right. We have almost 0:13:32.880 --> 0:13:34.640 no access to what's happening under the hood. 0:13:35.200 --> 0:13:37.480 Yeah, and I think that there's a reason for that. 0:13:37.960 --> 0:13:41.920 But you could have minds that were much more had 0:13:42.000 --> 0:13:46.760 much more access to that mutability, to that changeability, to 0:13:46.840 --> 0:13:51.280 the reprogrammable nature of them. So there could be sophisticated 0:13:51.320 --> 0:13:58.320 compounds that had a reflexive ability to mess with themselves. Okay, 0:13:58.360 --> 0:14:02.760 so that's a different kind of mind. There's also one 0:14:02.800 --> 0:14:06.160 of the things about is like, like, what is the 0:14:06.280 --> 0:14:11.160 smallest possible mind that could accomplish something? This idea of 0:14:11.200 --> 0:14:13.160 like trying to minimize it. We're trying to do that 0:14:13.200 --> 0:14:15.439 with life right now. It's like, what's the smallest possible 0:14:15.520 --> 0:14:18.000 sell you could have and how far could you go? 0:14:18.640 --> 0:14:20.880 I mean, with something like a transistor count if you 0:14:20.920 --> 0:14:23.960 have current here and here, then it opens and otherwise 0:14:24.000 --> 0:14:24.520 it doesn't. 0:14:24.640 --> 0:14:28.920 And it's probably not complex enough because we I don't 0:14:28.920 --> 0:14:31.240 think we would say that it was intelligent. So when 0:14:31.480 --> 0:14:35.160 once we understand what we need to do reasoning, it's like, 0:14:35.280 --> 0:14:37.440 what's the smallest amount that we could do to get 0:14:37.480 --> 0:14:38.080 some reasoning? 0:14:38.160 --> 0:14:40.280 But a transistor does do reasoning in the sense that 0:14:40.360 --> 0:14:42.640 it says, hey, if I've got these two things, then 0:14:42.680 --> 0:14:44.400 I go, and if I don't, then I don't go. 0:14:45.040 --> 0:14:47.480 We haven't yet figured out what reasoning is. I don't 0:14:47.480 --> 0:14:49.680 think we have a good definition of reasoning. And so 0:14:50.280 --> 0:14:53.680 right now we're kind of in this incredibly exciting period 0:14:54.120 --> 0:14:57.640 where we say, some of these lms have some amount 0:14:57.640 --> 0:15:00.680 of reasoning. Can we measure that? Can we can we 0:15:01.240 --> 0:15:03.720 what's the metric? We don't have that yet, but I 0:15:03.720 --> 0:15:07.160 think we will come to understand that there's a difference 0:15:07.200 --> 0:15:12.360 between say, learning and reasoning. And again, this is our idea, 0:15:12.440 --> 0:15:15.120 like we're at the beginning of understanding the periodic table 0:15:15.880 --> 0:15:20.520 of cognition, and so another one would be you could 0:15:20.560 --> 0:15:27.560 have minds that forget things. Forgetting is actually very very 0:15:27.600 --> 0:15:30.040 important in many kinds of cognition. 0:15:30.400 --> 0:15:34.240 As the writer ball Zach said, memories beautify life, but 0:15:34.360 --> 0:15:36.200 only forgetting makes it bearable. 0:15:36.320 --> 0:15:39.720 Right, you could have ais that never forget anything, and 0:15:39.800 --> 0:15:42.600 you could have ais that have learned how to forget 0:15:42.600 --> 0:15:47.600 certain things. For again, we're going to again these are 0:15:47.600 --> 0:15:52.360 being engineered to do different tasks, and so there will 0:15:52.360 --> 0:15:54.960 be certain kind of tests where we want them to forget. 0:15:55.280 --> 0:15:59.240 So there'll be some you know, Nobel prize in the 0:15:59.240 --> 0:16:02.840 future for some AI researchers who figures out how to 0:16:02.920 --> 0:16:05.240 do intentional forgetting. 0:16:05.840 --> 0:16:08.760 Yeah, it appears, by the way, that is an aspect 0:16:08.760 --> 0:16:12.960 of our intelligence. We forget most of the things happening 0:16:12.960 --> 0:16:17.240 in our lives. And one hypothesis that Francis Kriik suggested 0:16:17.320 --> 0:16:20.520 is that dreams are our way of taking out the 0:16:20.520 --> 0:16:22.680 garbage at nighttime. Yeah. 0:16:22.760 --> 0:16:28.760 Yeah, so another aspect of different kinds of possible minds 0:16:28.800 --> 0:16:33.320 or minds that are would be easy to migrate versus 0:16:33.400 --> 0:16:36.359 ones that were difficult to migrate off of the substrate. 0:16:36.520 --> 0:16:41.440 So I am a big believer that the Church Turing 0:16:41.520 --> 0:16:44.120 hypothesis about computation. 0:16:45.200 --> 0:16:46.840 Can you explain that this misunderstood? 0:16:46.920 --> 0:16:50.920 So the Church Turing, how about this is about computation says, 0:16:52.720 --> 0:16:59.800 given enough storage and memory, all forms of computation are equivalent. 0:17:00.840 --> 0:17:03.960 That you can emulate what one computer can do another 0:17:04.000 --> 0:17:09.040 computer can do if you have enough storage and capacity. 0:17:10.040 --> 0:17:13.199 And I think that's the key thing is that in 0:17:13.280 --> 0:17:19.400 real time there isn't equivalency that that it makes a difference. 0:17:19.480 --> 0:17:21.879 It actually makes a difference. What's the substrate you're running on. 0:17:21.960 --> 0:17:25.160 They're not equivalent because there's a matter of time. Yeah, 0:17:25.440 --> 0:17:27.919 if you have an infinite space. But no computer has 0:17:27.960 --> 0:17:33.240 infinite tape. You have you're always finite. That's reality, your finite. 0:17:33.280 --> 0:17:35.480 You have finite time, you have to make a decision. 0:17:35.920 --> 0:17:38.160 If you emilate it, you're going slower. And so those 0:17:38.200 --> 0:17:41.040 make the difference when we come to intelligence, and so 0:17:41.160 --> 0:17:46.200 that the kinds of ais that we make on silicon 0:17:46.880 --> 0:17:54.000 will always behave differently than one running on wetwear on tissue. Okay, 0:17:54.160 --> 0:17:58.160 So one of the other hypotheses about the possible mind 0:17:58.240 --> 0:18:04.280 landscape is that a lot of those varieties come from 0:18:04.359 --> 0:18:09.359 operating on different substrates on different brains, that the brains. 0:18:09.600 --> 0:18:13.320 It's not like the materialists where it doesn't matter what 0:18:13.359 --> 0:18:15.720 the brain is made out of. I'm saying it absolutely 0:18:15.800 --> 0:18:19.080 does matter what the brain is made out of, because 0:18:19.240 --> 0:18:25.880 you will get a different compound from that, because there 0:18:25.960 --> 0:18:30.120 is a spatial element that makes a difference in the 0:18:30.160 --> 0:18:35.480 actual output of this. And so one of the things 0:18:35.800 --> 0:18:38.640 that the possible mind landscape would say is that all 0:18:38.680 --> 0:18:41.640 the minds that we're making are alien. 0:18:42.560 --> 0:18:45.320 They're not human. They're not human like. 0:18:46.240 --> 0:18:49.280 The only way we can make human like intelligence is 0:18:49.320 --> 0:18:53.080 to have a human like brain. We could make artificial 0:18:53.600 --> 0:18:58.960 minds that are based on cells, and the more cellular 0:18:59.280 --> 0:19:01.840 and gray matter or like they would be, the more 0:19:01.960 --> 0:19:06.520 that compound intelligence would resemble ours. But when we're making 0:19:06.520 --> 0:19:11.120 them on silicon, there are going to be alien intelligences, 0:19:11.640 --> 0:19:14.359 which doesn't mean that they're stupid. They could be very, 0:19:14.440 --> 0:19:17.800 very smart. It's just that they're different. They have a 0:19:17.800 --> 0:19:22.080 different character, they have a different personality, they're different. They're 0:19:22.119 --> 0:19:26.359 different in the way that Spack a Star Trek is 0:19:26.440 --> 0:19:31.680 different than Kirk, and that difference. 0:19:32.680 --> 0:19:36.320 Is actually their benefit. It's the fact that they're not 0:19:36.359 --> 0:19:37.440 thinking like we think. 0:19:39.280 --> 0:19:43.520 Right in the sense that a show like Westworld, there's 0:19:44.080 --> 0:19:47.040 an amusement park that's made, and we build robots that 0:19:47.080 --> 0:19:49.440 look just like humans so we can interact with them. 0:19:49.480 --> 0:19:51.760 But I've often thought that we probably are never going 0:19:51.840 --> 0:19:54.600 to build robots that look just like humans, because there's 0:19:54.600 --> 0:19:56.840 no point. We already have humans. Humans are easy enough 0:19:56.880 --> 0:20:00.240 to make, and what we do with our machine means 0:20:00.320 --> 0:20:03.439 in general is build things that do things that we 0:20:03.680 --> 0:20:05.240 are different. Yeah, exactly right, right. 0:20:06.080 --> 0:20:10.119 I think we will make humanoid robots because that's the interface, 0:20:10.520 --> 0:20:13.600 because we want them operating in our world, and we're 0:20:13.680 --> 0:20:19.120 comfortable with that scale. We're comfortable with the emotional connection, 0:20:19.600 --> 0:20:20.880 and so there are a lot of reasons to make 0:20:20.880 --> 0:20:23.960 them humanoid, but there's no reason to make them look 0:20:24.040 --> 0:20:27.080 like exactly like us. Okay, that they will be alien, 0:20:27.640 --> 0:20:32.639 and that's good because we need the alien intelligences to 0:20:32.720 --> 0:20:35.439 do the things that we can't do by ourselves. We 0:20:35.480 --> 0:20:40.399 can make another human in nine months, so we want 0:20:40.600 --> 0:20:45.200 to make other kinds of minds, and all those minds 0:20:45.240 --> 0:20:48.160 are going to be aliens to us in the sense 0:20:48.200 --> 0:20:51.280 that because they're running on a different substrate, they can 0:20:51.440 --> 0:20:55.600 fake a lot of human behavior, and they will because 0:20:55.680 --> 0:20:58.399 we're going to be comfortable with it, but they won't 0:20:58.520 --> 0:21:02.440 be exactly like this because they're running on a different substrate. 0:21:03.440 --> 0:21:06.240 Now, can you imagine a scenario where a system is 0:21:06.359 --> 0:21:09.240 built such that it's just call it a bigger a 0:21:09.280 --> 0:21:11.679 much bigger brain than we are, and it can emulate 0:21:11.760 --> 0:21:14.240 our intelligence on part of its hardware. 0:21:14.640 --> 0:21:19.480 It will imitate a lot of it, but because it's 0:21:19.600 --> 0:21:25.639 running on a different substrate, it cannot do exactly what 0:21:25.680 --> 0:21:31.040 we're doing. So my hypothesis is that this brain makes 0:21:31.040 --> 0:21:36.560 a difference to the mind. Okay, that they're not equivalent 0:21:37.200 --> 0:21:40.119 if you take into account the fact that they're that 0:21:40.160 --> 0:21:44.320 they're limited by time and space. Doing the conversation that 0:21:44.359 --> 0:21:48.600 we do in our brains is will give a different 0:21:49.640 --> 0:21:56.080 quality and maybe different answers to something that's done on silicon. 0:22:10.680 --> 0:22:13.439 Okay, And so so that we can keep stretching our 0:22:13.520 --> 0:22:15.919 minds here give us some other possibilities. 0:22:16.240 --> 0:22:20.280 So one is I think we could imagine minds that 0:22:20.480 --> 0:22:27.520 are very very specialized in working with individual peoples. It's like, 0:22:27.960 --> 0:22:30.960 not quite a clone to me, but a dedicated, very 0:22:31.080 --> 0:22:36.200 personalized AIS, very very personalized to my specific need and 0:22:37.080 --> 0:22:40.920 personality in my own brain that would be different from 0:22:41.000 --> 0:22:41.720 other humans. 0:22:41.800 --> 0:22:42.919 And they're pairing. 0:22:43.040 --> 0:22:47.119 So it's sort of like a paired mind, a mind 0:22:47.160 --> 0:22:50.600 that is built to be paired with my mind to 0:22:50.720 --> 0:22:53.760 work with my mind, and that's its only job, and 0:22:53.800 --> 0:22:56.240 it's really good at that. It's not good at anything else, 0:22:56.400 --> 0:22:59.240 and it's not good with working with you. So this 0:22:59.440 --> 0:23:01.840 idea of a paired intelligence. 0:23:01.720 --> 0:23:03.960 Oh, that's cool. So right now people are doing that, 0:23:04.000 --> 0:23:07.240 of course, with AI that learned your stuff, maybe reads 0:23:07.280 --> 0:23:09.080 all your emails and song. But you're saying, instead of 0:23:09.080 --> 0:23:13.000 a general AI that learns learns me, one that actually is. 0:23:13.080 --> 0:23:18.520 Built, right, It's constructed, its programmed, it's the weights, it's 0:23:18.560 --> 0:23:24.080 trained on me, so that it is unusable by somebody else. 0:23:24.600 --> 0:23:28.199 That's how dedicated it is. It's it's it's paired in 0:23:28.240 --> 0:23:32.439 that sense. Another one would be so so we can imagine, 0:23:32.680 --> 0:23:35.080 of course I had this, I had this kind of 0:23:35.119 --> 0:23:39.800 a grid of like the four possible directions of humanity 0:23:40.920 --> 0:23:46.000 with just two axises. One is, we can imagine a 0:23:46.040 --> 0:23:51.280 future of humanity that has many species of humans and 0:23:51.400 --> 0:23:59.399 many minds. Okay, so we speciate, yes, over time, we 0:23:59.440 --> 0:24:04.280 could imagine. Another one is we have many species of 0:24:04.359 --> 0:24:10.400 humans and one mind. Well, like, okay, where we are 0:24:11.040 --> 0:24:15.600 telepathy just does we're connected so much, even though we're 0:24:15.600 --> 0:24:18.880 different species, we're all in tune, and we create kind 0:24:18.880 --> 0:24:21.880 of a single borg intelligence on the planet. 0:24:22.560 --> 0:24:25.480 Then you can imagine before we move on to what 0:24:25.600 --> 0:24:28.199 degree do we have something moving in that direction with us, 0:24:28.200 --> 0:24:30.520 say the Internet? Yeah, yeah, right. 0:24:31.119 --> 0:24:33.320 So that's just to kind of indicate that there's a 0:24:33.520 --> 0:24:36.560 there's a there's a cold class of minds, that these 0:24:36.560 --> 0:24:40.960 are kind of superminds, these these aggregated minds made up 0:24:41.000 --> 0:24:46.040 of many many humans, and so we have we have 0:24:46.160 --> 0:24:48.000 the mind of all humans working together. 0:24:48.040 --> 0:24:49.000 If we had some kind of. 0:24:48.920 --> 0:24:53.880 Technologies that would allow us to telepathically connect to each other. 0:24:54.840 --> 0:24:56.840 It was like a neuralink, right, So you had a 0:24:56.840 --> 0:24:58.920 neuralink that kind of works at the large scale, and 0:24:58.960 --> 0:25:02.240 we're connected to others and we are we're creating some 0:25:02.359 --> 0:25:07.120 kind of superhuman mind at the skill of literally having humans, 0:25:07.720 --> 0:25:10.159 billions of them connected together. That would be a mind. 0:25:10.320 --> 0:25:13.200 That's the possible mind that we don't know very much about. 0:25:13.960 --> 0:25:17.640 So that's a possible mind. Then there's the possible mind 0:25:17.800 --> 0:25:21.320 of all the little ais linked up together. You have 0:25:21.520 --> 0:25:25.960 a thousand different species of ais and they are also connected, 0:25:26.320 --> 0:25:31.840 forming another level of AI that's operating much different dimension. 0:25:33.080 --> 0:25:35.199 And so, in a sense, is what everyone's going for 0:25:35.359 --> 0:25:38.320 with agentic AI. Where you have agents, you have lots 0:25:38.320 --> 0:25:40.240