WEBVTT 00:00:03.537 --> 00:00:10.497 This is the Convergent Science Network podcast. Leading researchers in the domain 00:00:10.497 --> 00:00:16.717 of neuroscience, brain theory and technology are interviewed by Paul Verschoor and Tony Prescott. 00:00:21.837 --> 00:00:28.057 This is Paul Verschoor, here together with Tony Prescott. but this is the last 00:00:28.057 --> 00:00:32.457 podcast of the 10th edition of BCBT. 00:00:35.277 --> 00:00:39.297 And essentially it's up to Tony and me now to discuss our talks. 00:00:40.557 --> 00:00:43.317 But also I think to reflect a little bit of what we tried to achieve, 00:00:44.597 --> 00:00:47.717 all the preceding BCBT's and the current BCBT. 00:00:48.817 --> 00:00:53.417 Because in some sense we also started this whole thing, the summer school and 00:00:53.417 --> 00:00:57.637 the Living Machine Conference, that's now running for six years, 00:00:59.397 --> 00:01:04.737 with the idea to really build or carve out, if you want, a multidisciplinary domain, 00:01:05.517 --> 00:01:10.497 that combines the study of mind and brain with the appropriate technologies 00:01:10.497 --> 00:01:15.757 to really build, if you want, artificial mind and brain as a model. 00:01:16.497 --> 00:01:20.817 And this is definitely reminiscent of, let's say, the cybernetics revolution 00:01:20.817 --> 00:01:22.917 of the 40s and 50s. Right. 00:01:22.957 --> 00:01:27.697 And I'm definitely greatly inspired by that period. 00:01:27.857 --> 00:01:31.137 And also, Tony, you started your talk with this famous Rosenbluth and Wiener 00:01:31.137 --> 00:01:36.157 quote, like the best model of a cat is the cat and preferably the same cat. 00:01:37.920 --> 00:01:43.620 And you use it more often in your talk, so you find it apparently an important point of departure. 00:01:44.520 --> 00:01:50.000 So why do you find that a useful observation? Well, I think it's interesting 00:01:50.000 --> 00:01:58.260 because so many people misunderstand that quote, because it's meant ironically. 00:01:59.280 --> 00:02:05.440 And people use it as if it, occasionally people use it as if they think it's 00:02:05.440 --> 00:02:06.780 just meant to take that literally. 00:02:06.780 --> 00:02:09.460 And always study the cat if you 00:02:09.460 --> 00:02:13.260 want to understand the cat whereas what uh rosenbluth and 00:02:13.260 --> 00:02:16.660 vena were saying is look uh 00:02:16.660 --> 00:02:19.880 you can't understand the cat uh 00:02:19.880 --> 00:02:26.540 in its full detail um because you will end up with a description that's as complicated 00:02:26.540 --> 00:02:30.640 as the cat indeed you won't ever get there because it's difficult very difficult 00:02:30.640 --> 00:02:35.780 to measure and the whole point of of what they were saying is that actually 00:02:35.780 --> 00:02:38.300 if we want to understand any animal, 00:02:38.360 --> 00:02:41.640 such as a cat or a human, then we have to use models. 00:02:42.100 --> 00:02:45.800 And in biology, everyone accepts this. 00:02:46.500 --> 00:02:50.600 But what they use is one animal as a model of another. 00:02:50.940 --> 00:02:56.220 So typically in neuroscience now, we use mice as a model of the human. 00:02:56.840 --> 00:03:01.460 And that involves a whole lot of assumptions, which we don't often discuss, 00:03:01.520 --> 00:03:06.360 or they're not often pushed to the fore, perhaps the way they should be um but 00:03:06.360 --> 00:03:08.880 even more than that we we use um. 00:03:09.923 --> 00:03:14.603 A a mouse that might be uh anesthetized or 00:03:14.603 --> 00:03:17.623 uh you know it might be uh that we 00:03:17.623 --> 00:03:20.423 are uh holding it down pinning it down in 00:03:20.423 --> 00:03:23.963 some sort of head restraint and making it do some behavior and in 00:03:23.963 --> 00:03:26.703 each of these situations we are stepping further and further 00:03:26.703 --> 00:03:29.503 away from the thing we want to understand which might be the 00:03:29.503 --> 00:03:33.383 freely moving freely behaving animal or person so 00:03:33.383 --> 00:03:36.343 um i like that norbert vena a quote 00:03:36.343 --> 00:03:39.103 because the rest of that article it's a lovely 00:03:39.103 --> 00:03:42.223 short article which sets out the 00:03:42.223 --> 00:03:44.943 value of physical models and of course they were 00:03:44.943 --> 00:03:50.803 writing in the 1950s 1940s at the time when they didn't really have computer 00:03:50.803 --> 00:03:55.703 models but they could build physical devices to try and understand stuff and 00:03:55.703 --> 00:04:00.303 that what that's what he was advocating he was saying let's build some simplified 00:04:00.303 --> 00:04:04.843 physical models in order to understand these complex biological systems. 00:04:05.163 --> 00:04:11.263 And of course, now we can build these very sophisticated physical models, which we call robots. 00:04:11.603 --> 00:04:15.223 Right. So what is interesting about this, I was in your answer, 00:04:15.343 --> 00:04:19.303 in some sense now we can look at different kinds of models, right? 00:04:19.363 --> 00:04:25.383 Because using one animal species to try to understand another animal species. 00:04:25.723 --> 00:04:30.643 For that You might use the word model, but it's in a rather different meaning 00:04:30.643 --> 00:04:36.243 than if you talk about constructing an artifact that is a model, right? 00:04:36.303 --> 00:04:40.943 So I think we should also keep that apart because in terms of accessibility. 00:04:41.803 --> 00:04:46.923 You basically say, look, I have more experimental control over this animal model 00:04:46.923 --> 00:04:50.503 in order to make inferences about, let's say, the human brain. 00:04:51.323 --> 00:04:56.463 But if you talk about an artifact, you really have to put that together from the bottom up. 00:04:56.463 --> 00:04:59.423 So it's it's a level of control is is 00:04:59.423 --> 00:05:02.583 of course an order of magnitude larger than 00:05:02.583 --> 00:05:05.583 what you would have with an animal model yeah i mean so with the 00:05:05.583 --> 00:05:08.763 um the physical uh model let's say 00:05:08.763 --> 00:05:12.063 it's a robot you have access to absolutely everything 00:05:12.063 --> 00:05:16.403 that you're interested in because you've built it you also pretty much know 00:05:16.403 --> 00:05:19.983 how it works although what you can never be quite sure until you've built it 00:05:19.983 --> 00:05:24.923 what it's going to do uh and uh that's the great thing about it and of course 00:05:24.923 --> 00:05:30.503 the The reason that people are sometimes skeptical about us using robots as models is, 00:05:30.563 --> 00:05:35.123 of course, all the biological substrates of the animal aren't present in the robots. 00:05:35.223 --> 00:05:39.283 We're trying to approximate them with, you know, silicon and plastic and metal. 00:05:39.603 --> 00:05:42.843 And that's a hard job to do. And so…. 00:05:43.988 --> 00:05:49.408 Whether people accept that this is a good strategy depends a little bit on whether 00:05:49.408 --> 00:05:54.808 they agree with the sort of position, I think, that you and I share that these 00:05:54.808 --> 00:05:58.048 physical artifacts are, in some interesting sense, 00:05:58.268 --> 00:06:05.388 devices that can obey the same principles as humans and animals in the way they operate. Right. 00:06:05.628 --> 00:06:10.488 But there are two things here. On the one hand, you automatically then fall 00:06:10.488 --> 00:06:14.508 into this question like, okay, what then makes a good model? 00:06:15.388 --> 00:06:18.468 So we have discussed abstraction. Abstraction is required. 00:06:18.788 --> 00:06:24.608 But then there already we see an issue that models are being – given the computational 00:06:24.608 --> 00:06:27.588 technologies we have, we can start adding a lot of detail. 00:06:27.588 --> 00:06:35.528 So we are slipping away from this original Rosenblatt and Wiener idea of abstraction 00:06:35.528 --> 00:06:38.128 and construction because now we say, no, 00:06:38.188 --> 00:06:42.248 we can throw in as much detail as we want and end up with something like blue 00:06:42.248 --> 00:06:46.408 brain that is as complicated as the original neural system we measured from 00:06:46.408 --> 00:06:48.008 and as incomprehensible, right? 00:06:48.008 --> 00:06:52.608 So this issue of abstraction, I think, is an important one to really protect 00:06:52.608 --> 00:06:55.348 when we speak about models. You have to be clear about your abstractions. 00:06:56.128 --> 00:07:00.688 And then the second thing, and I don't think it's sufficiently appreciated, 00:07:00.828 --> 00:07:08.188 is that robots give us access to behavior, and behavior is a further constraint we can impose on models. 00:07:08.188 --> 00:07:14.368 Like, with animal models, what you see happening is that to keep things controllable, 00:07:14.408 --> 00:07:17.368 because that's the key constraint of any experiment, 00:07:17.628 --> 00:07:23.288 you push animals into a corner of their behavioral space where you might never 00:07:23.288 --> 00:07:24.608 find them in the real world. 00:07:24.688 --> 00:07:28.448 And this is something Lea was pointing to, right, with her animals that live 00:07:28.448 --> 00:07:31.368 now in these pens in the real world. Right. 00:07:32.871 --> 00:07:40.411 You might have constrained or distorted their behavioral output so much that 00:07:40.411 --> 00:07:42.371 actually you don't really understand what these animals are doing. 00:07:43.191 --> 00:07:47.851 And I think behavior is one of the key constraints we have, and we lost it largely 00:07:47.851 --> 00:07:52.131 over many decades, also in the reaction against, if you want, 00:07:52.191 --> 00:07:54.251 behaviorism, which is strong adherence to behavior. 00:07:54.511 --> 00:07:58.411 But the robot allows us, again, to bring in behavior as a constraint on these 00:07:58.411 --> 00:08:01.351 abstract constructs that we call models. 00:08:02.071 --> 00:08:07.671 Yeah, I think, and this is sort of ethological realism that you can begin to 00:08:07.671 --> 00:08:13.271 bring back with robot models is going to really help. 00:08:13.431 --> 00:08:20.171 But I think there's also a little bit of a problem in this field that you also 00:08:20.171 --> 00:08:22.371 need to work with these models. 00:08:22.471 --> 00:08:25.871 You have to have an understanding of how to do controlled experiments still. 00:08:26.151 --> 00:08:29.191 And so we find a lot of people chucking together 00:08:29.191 --> 00:08:33.531 physical devices that resemble some animal 00:08:33.531 --> 00:08:36.471 and it does some behavior that resembles the animal behavior but 00:08:36.471 --> 00:08:40.091 it's not necessarily helping us to answer any question so 00:08:40.091 --> 00:08:42.991 uh the methodology for how we 00:08:42.991 --> 00:08:46.951 go about this what um breitenberg 00:08:46.951 --> 00:08:50.331 who is another great you know uh inspiration for 00:08:50.331 --> 00:08:53.311 me and i think for us both what what he called uh synthetic 00:08:53.311 --> 00:08:57.431 psychology um it's good 00:08:57.431 --> 00:09:00.451 to try and thrash out what is the best methodology or 00:09:00.451 --> 00:09:03.571 what are appropriate methodologies for doing synthetic psychology 00:09:03.571 --> 00:09:06.551 well right and i think that's a bit what we've 00:09:06.551 --> 00:09:10.771 been trying to grapple with at bcbt and 00:09:10.771 --> 00:09:15.071 living machines is to look across the range of approaches people are taking 00:09:15.071 --> 00:09:21.011 and ask questions like how do biologists and engineers need to work together 00:09:21.011 --> 00:09:26.891 in order to progress this field how what is an appropriate way to design an 00:09:26.891 --> 00:09:30.071 experiment that involves a robot as opposed to an animal? 00:09:30.451 --> 00:09:35.551 And I think that your talk today was a nice example of when you were talking 00:09:35.551 --> 00:09:41.071 about vicarious trial and error and linking the literature from Tolman through 00:09:41.071 --> 00:09:47.131 to early studies on how rats look both ways in a maze before deciding which way to turn. 00:09:47.431 --> 00:09:53.951 Recent data from David Reddish showing that there's really activity in the hippocampus 00:09:53.951 --> 00:09:57.371 that shows the rats really thinking about turning left before it goes right, 00:09:57.491 --> 00:10:02.731 and then showing in a robot model that has embedded within it a model of that 00:10:02.731 --> 00:10:05.251 neural system that we can capture those properties. 00:10:05.491 --> 00:10:10.791 And I think within the 10 years we've been doing BCBT, I think our understanding 00:10:10.791 --> 00:10:12.671 of that methodology has progressed, actually. 00:10:13.891 --> 00:10:17.831 And it's not straightforward to see how to do this when you first get going 00:10:17.831 --> 00:10:19.611 with it. Right. No, absolutely. 00:10:20.051 --> 00:10:26.591 So I think this, uh, so, so what I see that methodologically what helped me a lot, um, is that, 00:10:27.668 --> 00:10:31.948 We always have been advancing our models along two lines or two levels of abstraction. 00:10:33.588 --> 00:10:39.368 So one, the distributed adaptive control theory, which is very strongly anchored 00:10:39.368 --> 00:10:44.688 in behavior and robotics, which helps me to think about behavior and underlying, 00:10:44.868 --> 00:10:47.348 if you want, psychological processes of memory, attention, 00:10:47.728 --> 00:10:52.568 decision-making, action selection, and so on, without immediately constraining 00:10:52.568 --> 00:10:56.208 myself by substrate, like how can neurons do this? 00:10:56.208 --> 00:11:02.508 What I always did try to do at that level is to say, well, let's at least not violate the obvious. 00:11:02.588 --> 00:11:08.848 Let's not violate things we know, such as neurons do not broadcast global information 00:11:08.848 --> 00:11:10.708 to each other, for instance. 00:11:10.908 --> 00:11:15.028 So you try to impose these kinds of constraints. You try to keep any kind of 00:11:15.028 --> 00:11:18.668 rules of learning based on local information and so on. 00:11:19.168 --> 00:11:24.928 But then, so that helped us to really build these integrative behavioral models, 00:11:25.088 --> 00:11:29.948 which means you have to really think about real-time, real-world, embodied control. 00:11:31.188 --> 00:11:34.428 And then in parallel, we would run these more detailed models where we say, 00:11:34.488 --> 00:11:38.588 okay, here we make a prediction about, let's say, conjunctive representations 00:11:38.588 --> 00:11:43.548 playing a role in optimal decision-making, right? 00:11:43.548 --> 00:11:48.908 One of the early predictions we had was that sensory motor states are like the 00:11:48.908 --> 00:11:53.248 primitive representation elements for optimal decision-making in a Bayesian sense. 00:11:53.888 --> 00:11:56.928 And that then became a driving hypothesis to look at the brain. 00:11:57.008 --> 00:12:01.848 And with that, we went to the hippocampus. And using very detailed and also 00:12:01.848 --> 00:12:07.988 anatomically and physiologically sophisticated models, working with Cesar Renacosta and John Lisman, 00:12:08.728 --> 00:12:12.048 we really could interpret a lot 00:12:12.048 --> 00:12:15.088 of properties of this hippocampal system that looked initially 00:12:15.088 --> 00:12:17.788 very puzzling but in that case we didn't worry 00:12:17.788 --> 00:12:20.708 too much yet about behavioral consequence but once 00:12:20.708 --> 00:12:24.788 we nailed that model and we had the basic principles in place then we could 00:12:24.788 --> 00:12:28.648 bring it together again in a robot model because now I could replace a bunch 00:12:28.648 --> 00:12:33.288 of the let's say memory systems that I had defined more algorithmically with 00:12:33.288 --> 00:12:36.828 a much more constrained trained model of hippocampal processing that then suddenly 00:12:36.828 --> 00:12:39.488 gave me new features that I had never thought of before. 00:12:39.668 --> 00:12:43.148 Because initially in DEC, I was always thinking about, 00:12:44.250 --> 00:12:49.410 short-term memory in terms of a sequential representation of sensory-motor states 00:12:49.410 --> 00:12:52.690 that I could then link to goals for policy generation. 00:12:53.910 --> 00:12:58.450 But then by mapping that back to the hippocampus, we learned that actually hippocampus 00:12:58.450 --> 00:13:00.770 is doing mind travel, right? 00:13:00.810 --> 00:13:04.370 It's exploiting these sequences in vicarious trial and error to actually make 00:13:04.370 --> 00:13:08.350 predictions and then have an internal simulation of what the world might look 00:13:08.350 --> 00:13:11.010 like when you go in one direction as opposed to the other. 00:13:11.010 --> 00:13:15.050 And that's now a whole new feature that we unlocked in the context of this overall 00:13:15.050 --> 00:13:22.810 framework of DAC and optimal control or robust control in the context of behavior. 00:13:23.150 --> 00:13:26.810 So there's a real, very constructive synergy now between these lines of modeling 00:13:26.810 --> 00:13:31.710 that I think it took, of course, some time to build that momentum. 00:13:31.990 --> 00:13:36.450 But I think now we're really gathering the fruits of that. I mean, 00:13:36.450 --> 00:13:40.290 if we say that the core problem, and I don't think there's one problem, 00:13:40.350 --> 00:13:44.830 there's many problems we're interested in, but one core problem is the question of architecture. 00:13:45.090 --> 00:13:48.790 What is the control architecture of the brain? And obviously, 00:13:48.890 --> 00:13:52.270 that's the core question behind a lot of your work. 00:13:53.410 --> 00:13:56.430 Then there's there's a top-down way of approaching that 00:13:56.430 --> 00:14:00.130 which is you look at behavior you look at what systems 00:14:00.130 --> 00:14:02.850 in principle could generate that behavior and you 00:14:02.850 --> 00:14:08.190 try and build those systems and demonstrate that they're sufficient and then 00:14:08.190 --> 00:14:12.230 there's the more bottom-up approach you look at the circuits that you find in 00:14:12.230 --> 00:14:15.730 in the biological systems that have that behavior and you see how how those 00:14:15.730 --> 00:14:21.750 circuits could have properties that could instantiate those principles. 00:14:22.290 --> 00:14:26.270 I mean, so would you agree there's this combination of top-down, 00:14:26.330 --> 00:14:30.210 bottom-up that you're trying to do sort of in parallel lines, 00:14:30.310 --> 00:14:35.030 and they're trying to feed in to answer these central questions about the architecture. 00:14:35.250 --> 00:14:39.070 So there are the different levels of description here, 00:14:39.230 --> 00:14:43.510 but in a way where we want to push for the high-level description, 00:14:43.510 --> 00:14:46.390 which is going to be then the most powerful set of principles 00:14:46.390 --> 00:14:49.330 yeah no look absolutely you're right 00:14:49.330 --> 00:14:52.150 and what's really important to also not give 00:14:52.150 --> 00:14:57.410 any kind of exclusive status to either of the two you have to do both to to 00:14:57.410 --> 00:15:03.210 to find to identify the constraints um so yeah i see this really as as a combined 00:15:03.210 --> 00:15:08.470 top-down bottom-up approach however i think it's important to observe that But 00:15:08.470 --> 00:15:11.290 to make that work constructively, 00:15:11.430 --> 00:15:14.670 you must commit yourself to clear empirical benchmarks. 00:15:15.430 --> 00:15:19.510 Otherwise, we're in fantasy land and then we very quickly wave our hands and 00:15:19.510 --> 00:15:21.890 we speak of biological inspiration and so on. 00:15:21.950 --> 00:15:25.610 But scientifically, that's not going to help us. So it's really important to 00:15:25.610 --> 00:15:30.590 anchor both of these, the top-down and the bottom-up view and the models that 00:15:30.590 --> 00:15:34.890 come out of it, to very clear empirical benchmarks, which must be grounded in 00:15:34.890 --> 00:15:36.730 the anatomy, the physiology, and the behavior. 00:15:37.859 --> 00:15:40.939 That we know brains and minds generate. 00:15:42.159 --> 00:15:48.739 When we talk about that, though, with modeling, there's different goals that you can have. 00:15:49.059 --> 00:15:52.779 And it's not always prediction that you're going to put first. 00:15:52.959 --> 00:15:58.859 It's more things like sufficiency. Is the model sufficient to generate the behavior 00:15:58.859 --> 00:16:00.439 that we're interested in? 00:16:00.699 --> 00:16:05.899 And you might also have predictions. But my experience is that people who have 00:16:05.899 --> 00:16:10.979 built robot models have occasionally come up with some interesting predictions 00:16:10.979 --> 00:16:12.439 that biologists might not have had. 00:16:12.999 --> 00:16:17.959 But that doesn't happen as often as maybe it could or should. 00:16:18.459 --> 00:16:21.979 But actually they have another contribution to make which they can test the 00:16:21.979 --> 00:16:22.899 sufficiency of the model. 00:16:23.019 --> 00:16:27.279 They can say, I instantiate this theory. Either it does or doesn't work or it 00:16:27.279 --> 00:16:28.899 needs to be improved in this way. 00:16:30.099 --> 00:16:33.719 And I think another thing that it does that people overlook is that it also 00:16:33.719 --> 00:16:35.279 helps the biologists focus on, 00:16:35.919 --> 00:16:42.219 what are the important questions, because biologists, as any natural scientist, 00:16:42.539 --> 00:16:47.399 tend to be attracted by phenomena that they can observe, and then they think, 00:16:47.399 --> 00:16:48.459 oh, well, I can study this phenomena. 00:16:48.939 --> 00:16:55.099 And they don't always pick, and because you can't, the ones that are going to 00:16:55.099 --> 00:16:56.679 be important in terms of understanding function. 00:16:56.999 --> 00:17:01.539 Whereas there's an engineering approach which says, well, in order to solve 00:17:01.539 --> 00:17:05.439 this problem, and we probably have to have this kind of mechanism. 00:17:05.699 --> 00:17:10.399 Can you see that in the brain? So it forces a different kind of question into the biology. 00:17:11.739 --> 00:17:17.139 Well, I know you are a nice guy, but the sufficiency argument, 00:17:17.439 --> 00:17:20.559 I'm not so convinced by that, you know, because it's very weak, 00:17:20.659 --> 00:17:22.759 right? We need testability. 00:17:23.679 --> 00:17:27.399 And sufficiency means something like, well, I haven't rejected it yet, 00:17:27.439 --> 00:17:28.459 but I might in the future. 00:17:28.679 --> 00:17:32.759 And I think what we should not forget get is behind any modular assumptions. 00:17:34.564 --> 00:17:37.564 Assumptions are testable predictions or that from these 00:17:37.564 --> 00:17:40.484 assumptions you can derive testable predictions and i 00:17:40.484 --> 00:17:43.284 think people are often not paying enough attention to that 00:17:43.284 --> 00:17:46.384 they often run away from their predictions because probably 00:17:46.384 --> 00:17:49.804 they know if they get declared the model collapses immediately because 00:17:49.804 --> 00:17:53.424 the the assumptions were just too strong and so 00:17:53.424 --> 00:17:56.684 so that's one thing about sufficient behind is a sufficient model 00:17:56.684 --> 00:17:59.764 are testable assumptions that that 00:17:59.764 --> 00:18:02.764 that we should you'd look for okay well i i mean 00:18:02.764 --> 00:18:05.984 i i think sufficiency is part of it completeness is 00:18:05.984 --> 00:18:09.084 another you know sort of the more of detail of 00:18:09.084 --> 00:18:11.864 the system you can capture the better you're going i think that's 00:18:11.864 --> 00:18:16.544 where your principle of convergent validation comes in it's a it's a question 00:18:16.544 --> 00:18:20.924 of completeness can i account for the biology at multiple levels no no so i 00:18:20.924 --> 00:18:25.984 think if you can take that that that together with the sufficiency then you 00:18:25.984 --> 00:18:30.164 start to uh really be able to say well well, 00:18:30.164 --> 00:18:34.864 anything that can fit all these criteria is going to be a good model. 00:18:34.984 --> 00:18:37.384 And yeah, great if you can do prediction too. 00:18:37.724 --> 00:18:45.484 But I think to some extent, neuroscience, for instance, has been too captured 00:18:45.484 --> 00:18:48.644 by this hypothesis testing idea. 00:18:48.884 --> 00:18:54.084 So to get in a top journal, you have to say that you're proposing a new theory 00:18:54.084 --> 00:18:55.804 of X, you've got this strong hypothesis, 00:18:56.084 --> 00:18:58.884 you tested it, the data you know uh 00:18:58.884 --> 00:19:01.684 came out the right way and you published 00:19:01.684 --> 00:19:04.804 your paper and that encourages people not to 00:19:04.804 --> 00:19:07.444 build on what's gone before so much but to say that oh 00:19:07.444 --> 00:19:10.384 i've done something new uh it it encouraged an 00:19:10.384 --> 00:19:13.644 element of cherry picking which has resulted in 00:19:13.644 --> 00:19:17.024 i think in recent years people saying well we need to go back uh 00:19:17.024 --> 00:19:21.404 to the beginning and measure the brain completely without having these a priori 00:19:21.404 --> 00:19:25.044 theories about how it works so i don't know if hypothesis testing interesting 00:19:25.044 --> 00:19:30.584 the way it's been done in neuroscience can just be imported into synthetic psychology 00:19:30.584 --> 00:19:36.684 so but i'm i really disagree with you now i mean sorry now you're going too far because. 00:19:37.804 --> 00:19:44.804 First i believe that um first convergence validation is it's more about how 00:19:44.804 --> 00:19:50.844 do you deal with the intrinsic indeterminacy of a model right so so models have certain parameters, 00:19:51.364 --> 00:19:56.184 and you use these parameters to fit the curve in the end, right? 00:19:56.244 --> 00:19:58.384 It's like it's a curve-fitting exercise in some way. 00:19:59.195 --> 00:20:03.775 And now you can add as many parameters as you have data points. 00:20:04.295 --> 00:20:06.615 So now you have a problem of overfitting. 00:20:07.735 --> 00:20:12.815 And to counteract that, you want to add more constraints from different levels of description. 00:20:13.115 --> 00:20:18.015 So the conversion validation is a way to think about this indeterminacy of models. 00:20:18.535 --> 00:20:25.435 And we saw examples of that also in BCBT that people would explain a sort of response curve. 00:20:25.515 --> 00:20:28.315 And suddenly we have a new magic parameter in a model to do that. 00:20:28.315 --> 00:20:32.975 And it's nice, great, it's a way to think about that specific phenomenon, 00:20:33.255 --> 00:20:36.855 but as a model, it's not completely satisfactory, you know. 00:20:36.875 --> 00:20:41.375 And then if you now map it to neuroscience, I don't feel at all that neuroscience 00:20:41.375 --> 00:20:44.755 is sort of suffering from an excess of hypothesis. 00:20:45.115 --> 00:20:51.175 To the contrary, I think it's suffering from an excess of technologies that, 00:20:51.255 --> 00:20:55.955 you know, people get enslaved or entrained by the technologies they have available. 00:20:55.955 --> 00:21:00.775 And every technology will unlock another set of potential correlations in the 00:21:00.775 --> 00:21:03.075 universe, and we're going to chase them all down. 00:21:03.395 --> 00:21:09.395 So, actually, I feel that we're really hypothesis-starved and data-rich. 00:21:09.695 --> 00:21:13.995 Well, I think that what you mean by hypothesis or what I mean by hypothesis 00:21:13.995 --> 00:21:15.095 is something a bit different. 00:21:15.095 --> 00:21:24.455 So, what I'm thinking of in terms of a hypothesis is it can be some quite relatively straightforward, 00:21:24.695 --> 00:21:30.115 say, a way of saying, well, people in the past have said this system is wired 00:21:30.115 --> 00:21:33.275 up in this way, and here we have data that shows something different. Oh, like that. 00:21:33.415 --> 00:21:41.495 Okay. Yeah, so what we're lacking in neuroscience is truly explanatory theories 00:21:41.495 --> 00:21:42.935 of systems. Absolutely. 00:21:43.175 --> 00:21:47.315 And yes, explanatory theories should give rise to predictions. 00:21:47.775 --> 00:21:51.375 But I think they should explain first. So, I mean, I've... 00:21:53.217 --> 00:21:57.857 My reading of philosophy of science, I've been kind of, I was, 00:21:57.877 --> 00:22:02.217 you know, skilled in the 1980s when Popper was everything. 00:22:03.097 --> 00:22:06.077 I always forget how old you are. Yeah, exactly. 00:22:06.797 --> 00:22:13.517 But I think the sort of the, all the emphasis on Popperian falsification, 00:22:13.677 --> 00:22:16.777 which is good experimental design, 00:22:17.057 --> 00:22:22.117 has detracted from the need to have strong explanatory theories. 00:22:23.317 --> 00:22:26.597 And I read David Deutsch's Fabric of Reality, 00:22:26.897 --> 00:22:33.257 and he's a physicist, but the first chapter of that book really drives home 00:22:33.257 --> 00:22:38.997 the difference between science as explanation and science as a predictive tool. 00:22:39.257 --> 00:22:43.297 And these are two quite different things. Science as a predictive tool is great, 00:22:43.457 --> 00:22:45.957 but that doesn't necessarily mean you're explaining anything. 00:22:46.157 --> 00:22:49.557 And to me, the first goal of science is to explain. Sure. 00:22:49.677 --> 00:22:56.597 Look, I completely agree with you. So, actually, I did my philosophy of science in German. 00:22:57.077 --> 00:23:01.757 So, you can imagine what the content was like, rather normative. 00:23:02.497 --> 00:23:05.517 But, of course, it included Popper, but not only. 00:23:06.397 --> 00:23:10.997 But, indeed, it's limiting. Popper is definitely limiting, and Deutsch is right. 00:23:11.097 --> 00:23:14.697 And the criteria for any theory is, in my opinion, threefold. 00:23:14.797 --> 00:23:17.617 It is to explain and predict and control. 00:23:18.297 --> 00:23:22.437 Right? And these are, for me, the criteria of a theory. That's why DAC is phrased 00:23:22.437 --> 00:23:26.617 the way it is, because we try to explain adaptive behavior, different forms. 00:23:27.557 --> 00:23:30.657 We make testable predictions, and then we control real-world systems, 00:23:30.717 --> 00:23:32.417 or we control recovery in patients. 00:23:33.417 --> 00:23:38.117 For me, those are the three criteria. And what has helped me a lot to make sense 00:23:38.117 --> 00:23:42.817 of these methodological requirements are these ideas of Bas van Fraassen, 00:23:43.057 --> 00:23:47.397 a philosopher of science who wrote his book, The Scientific Image, 00:23:47.397 --> 00:23:51.297 who is basically advancing the idea that theories... 00:23:52.659 --> 00:23:55.539 Are actually empirically just need 00:23:55.539 --> 00:23:58.319 to be empirically adequate so we just have to accept that 00:23:58.319 --> 00:24:01.079 they're the best possible description we have 00:24:01.079 --> 00:24:04.759 today of a set of phenomena but they 00:24:04.759 --> 00:24:07.479 are contingent they're evolving they're dynamic and at 00:24:07.479 --> 00:24:10.399 some point they might be rejected and that helped me 00:24:10.399 --> 00:24:14.599 a lot to escape from this very stringent normative popperian 00:24:14.599 --> 00:24:17.419 framework where you always are so stuck with 00:24:17.419 --> 00:24:20.639 with falsification because it it is 00:24:20.639 --> 00:24:24.579 definitely limiting and and not helping us forward especially if you're in a 00:24:24.579 --> 00:24:28.379 domain where you have to cut across many levels of description in particular 00:24:28.379 --> 00:24:33.359 mind and brain yeah well i think yeah i think that's absolutely right and the 00:24:33.359 --> 00:24:37.379 problem with with popper and indeed when i was studying this and i think it 00:24:37.379 --> 00:24:38.819 was 1980 this was already 00:24:38.999 --> 00:24:42.479 evident was that if your theory 00:24:42.479 --> 00:24:46.439 is wrong if it's been falsified there's there's 00:24:46.439 --> 00:24:49.359 no way in that framework to say where that's a better theory than any other 00:24:49.359 --> 00:24:53.419 but in fact we we live every day with with theories that are falsified and as 00:24:53.419 --> 00:24:57.879 you i think you said today you know all models are wrong so uh we've got to 00:24:57.879 --> 00:25:01.779 live with these wrong theories wrong models and we've got to try and improve 00:25:01.779 --> 00:25:07.179 them and i think i i think you're right explanation prediction control are three very 00:25:07.359 --> 00:25:10.499 good drivers for how we can make better theories. 00:25:10.639 --> 00:25:14.839 Yeah. So, but I, I do think that we, we, we can't just look at the biological 00:25:14.839 --> 00:25:19.859 sciences as a model for how to do this, uh, new kind of synthetic science. 00:25:20.679 --> 00:25:22.219 I do think that there's a. 00:25:23.783 --> 00:25:26.923 A gap in science for these 00:25:26.923 --> 00:25:30.263 explanatory theories and uh that's 00:25:30.263 --> 00:25:34.103 what we're we're trying to fill and maybe yeah we can we can make predictions 00:25:34.103 --> 00:25:38.363 off that and you made some predictions based on which you described today in 00:25:38.363 --> 00:25:42.383 which uh the the other hard part of this is of course persuading a biologist 00:25:42.383 --> 00:25:48.263 then to test your prediction right yeah sure uh but actually we we have succeeded 00:25:48.263 --> 00:25:50.123 I've succeeded in doing that once in a while, right? 00:25:50.163 --> 00:25:53.763 Including Edward Moser has been testing some of our predictions, which is great. 00:25:54.963 --> 00:25:58.363 But also there, you should see these predictions as a form of dialogue with 00:25:58.363 --> 00:25:59.103 the empirical sciences. 00:26:00.163 --> 00:26:05.483 So it's not like, oh, now I have this normative statement and you empiricists go test it, right? 00:26:05.543 --> 00:26:07.683 It's part of the dialogue that you have to try to establish, 00:26:07.863 --> 00:26:13.663 which is not always easy because the empirical scientists or the biologists 00:26:13.663 --> 00:26:17.603 are not necessarily trained to be very receptive to that. It's not a problem we have. 00:26:18.263 --> 00:26:22.823 But the biggest problem I see is that, actually, I started also with Tolman. 00:26:25.223 --> 00:26:29.723 And actually, what's interesting is that since Tolman, and especially Hull, 00:26:29.763 --> 00:26:34.603 Clark Hull, so we talk early 50s now, there have been no more comprehensive 00:26:34.603 --> 00:26:36.483 integrative theories in psychology. 00:26:37.363 --> 00:26:42.023 And in my opinion, whether you like it or not, neuroscience is largely dealing 00:26:42.023 --> 00:26:46.563 with the mind. All these functional properties of brains, in the end, 00:26:46.603 --> 00:26:50.103 traditionally were in the area of psychology. 00:26:50.483 --> 00:26:54.023 So this is what we tried to explain, but we threw it away. 00:26:54.303 --> 00:26:58.003 And now we're very worried about what neuron X does to neuron Y. 00:26:58.243 --> 00:27:01.023 And of course, since we can measure thousands of them at the same time, 00:27:01.103 --> 00:27:02.883 now it's sort of huge population we're trying to analyze. 00:27:03.303 --> 00:27:07.623 And then we might talk about forms of signal transduction and whatever. Right. 00:27:08.396 --> 00:27:12.216 But we'll have sometimes lost sight of the actual questions we're dealing with, 00:27:12.316 --> 00:27:13.896 which is what's memory, what's attention. 00:27:14.456 --> 00:27:19.276 As you also saw yesterday in Francesca's talk on hippocampal system, I think she did great. 00:27:19.876 --> 00:27:24.636 And she also clearly showed that, right? How we have to get back to these larger 00:27:24.636 --> 00:27:27.516 questions in this case about cognitive development, for instance. 00:27:27.776 --> 00:27:31.896 And I think this is where the synthetic psychology can really help to link mechanism 00:27:31.896 --> 00:27:37.296 to again function or in other words, brain and body to mind. 00:27:37.296 --> 00:27:42.356 I think, I mean, I would agree, but I would say Piaget was also somebody who 00:27:42.356 --> 00:27:46.556 was looking for these large scale theories and, you know, and up until certainly 00:27:46.556 --> 00:27:48.896 the 70s was developing them. 00:27:48.976 --> 00:27:54.516 And I think you can look at Piaget and you can see almost where psychology has 00:27:54.516 --> 00:27:56.016 gone right, but also gone wrong. 00:27:56.016 --> 00:27:59.076 Because in you know taking apart all 00:27:59.076 --> 00:28:02.196 of Piaget's experimental work and he 00:28:02.196 --> 00:28:05.196 was a good experimentalist but he uh over 00:28:05.196 --> 00:28:10.136 interpreted his results I think is what we can say uh and uh people have thrown 00:28:10.136 --> 00:28:15.576 away his whole framework on the basis of of that you know whereas uh and I think 00:28:15.576 --> 00:28:20.916 now development psychology is is fairly theory free or or they're into different 00:28:20.916 --> 00:28:23.396 camps you know which are quite far apart. 00:28:23.576 --> 00:28:30.076 And it's difficult to see how we can pull some of these areas of experimental psychology. 00:28:30.216 --> 00:28:32.756 I know development cognitive neuroscience is kind of a field, 00:28:32.896 --> 00:28:37.516 but it's not as big as it should be, certainly in the psychology community. 00:28:37.856 --> 00:28:43.796 And there is this kind of gulf between the neuroscientists, the developmentalists, 00:28:43.916 --> 00:28:47.476 the cognitive science who are trying to bridge that gap, but not necessarily 00:28:47.476 --> 00:28:52.056 succeeding. Well, look, you're right. But in that sense, I think what happened. 00:28:52.256 --> 00:28:57.116 So, for me, it's really a transitory figure because what's interesting about 00:28:57.116 --> 00:29:00.856 Tolman and Hull, they had this physics metaphor in mind. 00:29:00.916 --> 00:29:05.736 So, they really wanted to build a still logical, positivistic model of theories 00:29:05.736 --> 00:29:09.216 of the mind, which didn't really work out for them. 00:29:09.316 --> 00:29:12.596 But they definitely had the right intuitions. They're really on the right track. 00:29:13.356 --> 00:29:19.636 Piaget already starts to get dissolved in a universe of different experimental 00:29:19.636 --> 00:29:23.256 manipulations and interpretations and resonances with interpretations. 00:29:23.656 --> 00:29:29.256 So Piaget, it's very difficult to sort of condense that into one comprehensive 00:29:29.256 --> 00:29:31.036 theory that covers the whole of psychology. 00:29:31.516 --> 00:29:39.016 Well, he has this theory of equilibration, which has adaptation and assimilation as its two sides. 00:29:39.176 --> 00:29:43.456 And you can really link that. I mean, he didn't talk about dynamical systems 00:29:43.456 --> 00:29:50.476 as such, but essentially what he was describing was how dynamical systems can self-organize. 00:29:50.756 --> 00:29:56.236 And he was writing just before the second wave of connectionism, 00:29:56.256 --> 00:30:02.456 and you can really see how his ideas have been validated to a large degree by 00:30:02.456 --> 00:30:06.756 what people have been doing with those kinds of learning models. 00:30:06.756 --> 00:30:09.696 Um uh it hasn't yet really fed back 00:30:09.696 --> 00:30:12.876 into psychology because i know when i talk to developmentalists they 00:30:12.876 --> 00:30:16.136 all sort of uh get very very upset 00:30:16.136 --> 00:30:19.176 and annoyed when you say oh piaget was was great 00:30:19.176 --> 00:30:22.456 because they think he's ancient history i mean he is a historical figure 00:30:22.456 --> 00:30:25.896 and his experimental results are 00:30:25.896 --> 00:30:28.736 now you know for the museum but i think there 00:30:28.736 --> 00:30:31.836 was a core set of ideas there he was 00:30:31.836 --> 00:30:34.776 really battling cognitivism as well i mean he was he was 00:30:34.776 --> 00:30:38.016 at a time when cognitivism was very strong he was he was 00:30:38.016 --> 00:30:40.816 one of the figures it was exactly they were trying to push aside sure he 00:30:40.816 --> 00:30:45.716 was on the lone voices yeah it's of reason yeah and so i agree with that but 00:30:45.716 --> 00:30:50.536 what i want to say i'm not disagreeing about the the key role that piaget should 00:30:50.536 --> 00:30:56.396 be playing in our current thinking about about the mind but as compared to hull 00:30:56.396 --> 00:30:59.876 and tolman piaget's outlook was not. 00:31:01.690 --> 00:31:05.730 Covering the whole of psychology, he was clearly focusing on adaptation, 00:31:06.050 --> 00:31:07.090 development, and change. 00:31:07.570 --> 00:31:11.950 But for instance, if you look at Tolman, for instance, he really wants to go 00:31:11.950 --> 00:31:14.170 the whole way from genetics, 00:31:14.590 --> 00:31:20.510 the endocrine system, motivation, to cognition, moral thinking, 00:31:20.690 --> 00:31:23.490 and consciousness, right? The whole enchilada. 00:31:24.170 --> 00:31:27.450 And Piaget already reduced it a little bit. 00:31:27.510 --> 00:31:32.810 That's not a criticism of Piaget but but that's it it is for me it was signifying 00:31:32.810 --> 00:31:36.510 the direction psychology was taking it sort of becoming more fragmented and 00:31:36.510 --> 00:31:41.330 also reduce this ambition and I feel that the synthetic psychology we're now 00:31:41.330 --> 00:31:46.450 advancing using that term from Breitenberg also is a real opportunity, 00:31:47.030 --> 00:31:52.070 to put psychology back on the map of science to say no this is what it's really 00:31:52.070 --> 00:31:55.250 about if you want to understand the brain these are the questions we got to 00:31:55.250 --> 00:31:56.630 worry about yeah I mean I mean, 00:31:56.630 --> 00:32:01.930 I think one of the reasons maybe why we're harking back to these figures from the 50s, 00:32:01.930 --> 00:32:07.930 60s, and 70s is that it has become much harder in science to be the sort of 00:32:07.930 --> 00:32:14.590 generalist that these figures were in some way and to be able to go across different 00:32:14.590 --> 00:32:18.290 disciplines of mind and understand them. 00:32:18.290 --> 00:32:23.210 And so there's a need to build sort of interdisciplinary expertise, 00:32:23.270 --> 00:32:25.830 which is going to be largely teams. 00:32:27.610 --> 00:32:32.550 And there needs to be an understanding that that should be the strategy. 00:32:32.710 --> 00:32:38.350 And I think part of the issue at the moment is it's still hard to do that. 00:32:38.350 --> 00:32:41.670 There's a lot of talk about interdisciplinarity and how to support that, 00:32:41.810 --> 00:32:50.350 but still some of the aspects of the world of science militate against really 00:32:50.350 --> 00:32:51.690 interdisciplinary work. 00:32:51.910 --> 00:32:56.650 Well, I agree. This is a very important point because I think interdisciplinarity 00:32:56.650 --> 00:32:59.670 definitely is high risk. 00:33:00.699 --> 00:33:05.339 And in the face of scientific disciplines, they get more specialized, 00:33:05.599 --> 00:33:09.539 they get more technology-driven, that are also very much influenced by their 00:33:09.539 --> 00:33:14.379 own local cliques and networks and tribes, if you want. It's very tribal. 00:33:15.039 --> 00:33:22.379 On top of that, we have incentive systems that are driving people into very ruthless competition. 00:33:24.219 --> 00:33:28.619 Of course, this leads to that all the forces are stacked against what we need. 00:33:28.619 --> 00:33:32.299 So the forces are stacked against the multidisciplinarity and also the freedom 00:33:32.299 --> 00:33:35.099 to try risky hypothesis, right? 00:33:35.159 --> 00:33:39.879 So certainly in an environment where right now much more value is placed on 00:33:39.879 --> 00:33:44.039 what's called innovation or immediate impact as opposed to, let's say, 00:33:44.079 --> 00:33:48.199 fundamental advances. So yeah, you're absolutely right. 00:33:48.379 --> 00:33:53.619 And I think that the way through that for us has been that in the synthetic 00:33:53.619 --> 00:33:59.079 psychology, we actually are building artifacts and we are also having impact in robotics. 00:33:59.079 --> 00:34:03.659 We can solve certain problems or we can contribute to solving problems, 00:34:03.879 --> 00:34:08.559 not only in the psychology or the application of robots, but also really in 00:34:08.559 --> 00:34:12.779 the control of robots that gives us a bit more traction and maybe a bit more 00:34:12.779 --> 00:34:17.199 also if you want protection from this kind of criticism. 00:34:17.199 --> 00:34:22.619 And of course, also in our case, that we were successful in mapping the theory 00:34:22.619 --> 00:34:27.859 of DEC to the most effective neurorehabilitation method for stroke recovery. 00:34:28.179 --> 00:34:30.599 Of course, has helped us a lot. 00:34:30.679 --> 00:34:35.439 Because basically, we have changed, if you want, the criteria on which we have 00:34:35.439 --> 00:34:39.759 to judge theories in neuroscience and psychology by saying, well, 00:34:39.799 --> 00:34:43.639 it's not only about convincing your peers, it's about actually having measurable 00:34:43.639 --> 00:34:44.879 impact in the real world. 00:34:44.879 --> 00:34:48.759 Well, if you claim to know how the brain works, please go fix a patient somewhere. 00:34:48.959 --> 00:34:55.599 And if we would just apply this criterion consistently, I think we would reduce 00:34:55.599 --> 00:34:56.699 a lot of noise in the field. 00:34:57.378 --> 00:35:01.898 And get things recalibrated in a maybe a bit more constructive way. 00:35:02.118 --> 00:35:07.078 Yeah, I think that's right. And it's actually, with hindsight, 00:35:07.398 --> 00:35:13.158 it's a bit surprising to me that there isn't more emphasis within the field 00:35:13.158 --> 00:35:20.078 of cognitive science on trying to discover new therapies. I think it's coming back now. 00:35:20.338 --> 00:35:25.358 I certainly hope so. But it didn't feel that way 10, 20 years ago. 00:35:25.358 --> 00:35:28.518 And to some extent i think cognitive science 00:35:28.518 --> 00:35:32.158 has has tried to keep its focus on understanding 00:35:32.158 --> 00:35:36.138 mind and understanding the uh unimpaired 00:35:36.138 --> 00:35:39.358 mind as a first step is a good way to go uh 00:35:39.358 --> 00:35:45.758 it's also seeded some of the potential uh to do uh applications and impact to 00:35:45.758 --> 00:35:50.718 disciplines like ai unnecessarily i think because actually a lot of the best 00:35:50.718 --> 00:35:57.598 ideas in ai come through psychology and cognitive science and then get sort of rediscovered as AI. 00:35:58.998 --> 00:36:03.818 But then I think the other way around, the cognitive science hasn't done itself any favors, 00:36:04.518 --> 00:36:09.798 by disconnecting a little bit from psychiatry and all these other disciplines 00:36:09.798 --> 00:36:14.018 that are concerned with brain disease and impairments. 00:36:14.218 --> 00:36:17.738 Yeah, but you know what plays a role there? I think that given the incentive 00:36:17.738 --> 00:36:22.158 structures we try to survive in, people have become very risk-averse. 00:36:22.898 --> 00:36:26.298 And as we earlier agreed, all models are wrong. 00:36:27.310 --> 00:36:32.790 But as long as you can resonate with your peers and support each other's models, 00:36:32.990 --> 00:36:37.430 everybody can feel comforted and supported. 00:36:38.210 --> 00:36:41.730 But as soon as you map to the real world and you say, okay, here's my model 00:36:41.730 --> 00:36:45.290 of schizophrenia and now I'm going to bring it to the clinic, that's high risk. 00:36:45.750 --> 00:36:48.630 Because then if it doesn't work, it's really very obvious. 00:36:49.130 --> 00:36:52.050 And I think, so that's not a problem. People are very risk averse. 00:36:52.130 --> 00:36:55.190 That's why they don't want to make that step. So we have to really rethink the 00:36:55.190 --> 00:36:58.090 incentive structures in order to open that up. 00:36:58.470 --> 00:37:04.510 But in a way, I think what we're also saying is that it's not helpful to have 00:37:04.510 --> 00:37:10.390 this gap between so-called pure science on one hand and so-called applied science on the other. 00:37:10.430 --> 00:37:16.850 The best basic or pure science also has implications in the real world that 00:37:16.850 --> 00:37:19.190 people have overlooked sometimes. 00:37:19.190 --> 00:37:25.110 And I think there's also, again, within sort of academic publishing, 00:37:25.370 --> 00:37:31.170 there's sort of a preference for these sort of big discoveries in pure science 00:37:31.170 --> 00:37:35.810 so that you can make your career by just advancing knowledge without having 00:37:35.810 --> 00:37:37.430 to have innovation impact. 00:37:37.690 --> 00:37:44.270 So I'm kind of in favor of having more emphasis on the innovation side as a 00:37:44.270 --> 00:37:48.510 way of driving thinking within science, provided it's done sensitively. Right. 00:37:49.010 --> 00:37:53.290 This is really, these are good points, and I agree with you. 00:37:53.350 --> 00:37:56.990 And sometimes we can go back to the cyberneticians again, right? 00:37:57.090 --> 00:38:00.990 Because many of them had actually real-world concerns. Yeah. 00:38:00.990 --> 00:38:03.770 Like Wiener working on control systems, also... 00:38:06.210 --> 00:38:11.010 Ashby was a psychiatrist, right? So they were worrying about real problems. 00:38:11.530 --> 00:38:16.730 And for them, as far as I understand reading about them and reading their work, 00:38:16.910 --> 00:38:22.350 there was no divide between these applied concerns and the principles they took from it. 00:38:22.370 --> 00:38:28.470 They never complained about it or saw any thresholds or obstacles between that. 00:38:28.830 --> 00:38:32.970 And now this has been all reconceptualized in a way. 00:38:33.870 --> 00:38:37.090 Actually, also Vannevar Bush, right? Shortly after the Second World War, 00:38:37.190 --> 00:38:41.110 science, the endless frontiers, setting up the National Science Foundation of 00:38:41.110 --> 00:38:44.130 the States, was very much prosperity depends on science. 00:38:44.290 --> 00:38:47.870 There's a natural linkage between the human condition and science. 00:38:48.750 --> 00:38:53.090 And that was never really questioned. And now it seems we have to have this 00:38:53.090 --> 00:38:56.790 divergence like, oh, we have all these people concerned about the so-called 00:38:56.790 --> 00:39:00.450 basic questions. And then we have other people who do the sort of application 00:39:00.450 --> 00:39:02.810 of these and the two shall never meet. 00:39:03.110 --> 00:39:06.310 And I think that's a massive mistake. take. I don't think that that needs to 00:39:06.310 --> 00:39:08.270 be the case. And I also see it in our own work. 00:39:09.610 --> 00:39:14.090 Let's take again the example of neurorehabilitation. Every intervention is like 00:39:14.090 --> 00:39:20.430 an experiment and every patient provides new and valuable information on the basic theories we have. 00:39:20.510 --> 00:39:26.570 Like we have all sorts of ideas about how error might modulate learning in stroke patients. 00:39:27.150 --> 00:39:30.930 These are hypotheses and they're being tested every day in the clinic with real 00:39:30.930 --> 00:39:34.830 patients and we get the information back in real time, and it's advancing our theories. 00:39:35.170 --> 00:39:39.450 And I think this tight coupling between basic and applied science, 00:39:39.670 --> 00:39:43.110 which I then call Vico's loop, after one of my heroes, John Battista Vico, 00:39:43.610 --> 00:39:46.050 of the fact and the truth are reversible. 00:39:47.730 --> 00:39:50.890 We can find these loops, we can find these synergies between application and 00:39:50.890 --> 00:39:52.390 basic science if you look for it. 00:39:53.508 --> 00:39:58.428 And it's not happening enough. And another thing which is happening now, 00:39:58.528 --> 00:40:06.028 which certainly wasn't as evident a decade ago, is people actually looking at 00:40:06.028 --> 00:40:09.608 the technologies that are coming out of robotics, 00:40:09.828 --> 00:40:12.748 AI, cognitive science, and now saying, well, do we want these technologies? 00:40:13.928 --> 00:40:19.728 And so there's much more of a focus on, okay, where is this going to go? 00:40:19.868 --> 00:40:25.568 You know, who's going to benefit from this? and there's a growing realization 00:40:25.568 --> 00:40:29.448 that some of the technologies that have developed in the past 20 years have 00:40:29.448 --> 00:40:32.528 increased prosperity, but for the few rather than for the many. 00:40:33.948 --> 00:40:38.368 And we want to do something to counter against that. 00:40:38.568 --> 00:40:43.648 So, I mean, if we want, so I think it, you know, a lot of our science is publicly 00:40:43.648 --> 00:40:47.108 funded, but also I think as scientists, we want to feel that the contribution 00:40:47.108 --> 00:40:49.468 we're going to make is going to have a positive impact. 00:40:49.648 --> 00:40:52.868 So how do we build that into, uh 00:40:52.868 --> 00:40:56.608 what we're trying to do here how do we ensure that our 00:40:56.608 --> 00:40:59.508 goals and well not just our goals 00:40:59.508 --> 00:41:02.328 of the work that we're doing we're going to focus it 00:41:02.328 --> 00:41:07.148 towards positive impacts beneficial impacts right well that's really i think 00:41:07.148 --> 00:41:12.948 the key question we got to answer collectively and i i think that um we have 00:41:12.948 --> 00:41:17.728 to pick our problems carefully but maybe also rethink a little bit how How have 00:41:17.728 --> 00:41:20.248 we really structured the scientific enterprise? 00:41:20.528 --> 00:41:23.388 I think the main criterion right 00:41:23.388 --> 00:41:28.228 now for our science is how much you get appreciated by your colleagues. 00:41:29.769 --> 00:41:34.709 And that's very strange, right? Because these tribes also developed their own 00:41:34.709 --> 00:41:38.469 biases and expectations that might be completely besides reality. 00:41:38.889 --> 00:41:44.409 And if you also look at impact, for instance, you can look at many domains, 00:41:44.609 --> 00:41:50.529 cardiology, cancer, brain disease, education. 00:41:50.529 --> 00:41:57.609 Education, I don't think that after many decades of following this sort of peer-based, 00:41:58.469 --> 00:42:03.189 validation of a field has led to now a massive progress. 00:42:03.609 --> 00:42:09.049 I mean, of course, we cannot say that's been zero progress, but it has not been fantastic. 00:42:09.289 --> 00:42:13.929 Like in the case of neurorehabilitation, we looked at this over the period 1975 00:42:13.929 --> 00:42:18.629 till now. Now, we analyzed dozens of meta-analysis 00:42:18.629 --> 00:42:22.889 of the impact of stroke rehabilitation, and it stayed the same. 00:42:23.009 --> 00:42:28.689 So your chances of recovery 40 years ago are the same as to now. 00:42:29.909 --> 00:42:32.929 Well, that should give us some pause because in the same period of time, 00:42:32.989 --> 00:42:40.489 billions of euros have been spent on brain research and associated fields. Zero impact. 00:42:41.449 --> 00:42:45.649 So what's wrong here? And I think what's wrong here is really the model in which 00:42:45.649 --> 00:42:49.349 we have been pursuing this science, which was sort of divide and conquer, 00:42:49.569 --> 00:42:54.309 run after the technologies, go for greater levels of detail, 00:42:54.449 --> 00:42:58.689 or essentially just measure whatever the tools allow us to measure without even 00:42:58.689 --> 00:43:00.549 posing questions or even driving hypotheses. 00:43:01.049 --> 00:43:06.469 And I think by sacrificing psychology as our anchor point of questions, 00:43:06.889 --> 00:43:08.409 this has been the consequence. 00:43:08.709 --> 00:43:11.689 And I think the way back is still long and we're not doing very well. 00:43:12.971 --> 00:43:17.411 Uh, so that sounds a bit pessimistic. No, because we're doing something about 00:43:17.411 --> 00:43:20.511 it. We're in the clinic. We already treated 800 people and that's the beginning. 00:43:20.751 --> 00:43:22.931 There's still about 59 million to go. 00:43:23.311 --> 00:43:28.891 But in terms of, you know, getting, uh, science back on track towards solving 00:43:28.891 --> 00:43:32.111 some of these problems facing, uh, humanity, 00:43:32.571 --> 00:43:38.031 uh, what is, what are the quick steps that we could make, you know, 00:43:38.031 --> 00:43:41.911 sort of collectively, or at least we could agitate as a community to do more of. 00:43:41.911 --> 00:43:44.231 Okay, yeah, sure. Now, look, you're right. 00:43:44.391 --> 00:43:49.051 And that's also interesting. You see, both with Tolman and Hull, 00:43:49.151 --> 00:43:52.971 I talked about earlier, they really had this clear outlook. 00:43:53.271 --> 00:43:56.311 They also, of course, lived through the Second World War and they also saw it 00:43:56.311 --> 00:44:01.471 as their responsibility as psychologists to make recommendations and have impact at that level. 00:44:01.551 --> 00:44:04.071 Like, how do we advance the human condition? 00:44:04.551 --> 00:44:08.651 And I really think that should be our concern today. So that also means we have 00:44:08.651 --> 00:44:12.051 to rethink our science. Like, how are we going to really deal with the human condition? 00:44:12.231 --> 00:44:17.671 Is the way in which we have organized the scientific enterprise actually helping 00:44:17.671 --> 00:44:20.571 us to advance human condition? 00:44:20.691 --> 00:44:25.751 Take as an example, economy. Economy, in the end, is strongly dependent on human behavior. 00:44:26.071 --> 00:44:30.731 So how can we advance economy without linking it very, very closely to psychology 00:44:30.731 --> 00:44:37.551 and psychology to neuroscience, right? So I think there are bridges we have 00:44:37.551 --> 00:44:40.071 to build to make progress there. 00:44:40.251 --> 00:44:47.331 And we also have to, I think, be willing to pose a larger question. 00:44:47.371 --> 00:44:49.891 Because if we want to change the human condition positively, 00:44:50.271 --> 00:44:54.271 which we better do sooner than later because we're facing some serious challenges, 00:44:54.591 --> 00:44:56.991 and global warming is only one of them. 00:44:56.991 --> 00:45:02.571 But let's say income inequality, which is a very much social psychological phenomenon, 00:45:02.971 --> 00:45:06.691 I think is a massive stress on our society we have to deal with. 00:45:06.851 --> 00:45:11.131 But that means we have to understand things like greed and hoarding and tribal 00:45:11.131 --> 00:45:15.111 behaviors and forces that drive inequality. 00:45:15.951 --> 00:45:20.331 Do we have anything in our hands today to do that? No. Then you can say, 00:45:20.511 --> 00:45:22.451 well, look, you know, we have to improve education. 00:45:24.151 --> 00:45:30.811 But the most recent meta-analysis on the main factors that drive education in 00:45:30.811 --> 00:45:36.011 terms of its impact in terms of learning outcomes are not very conclusive. 00:45:36.171 --> 00:45:38.451 So there we still don't really know what we're doing and I think, 00:45:39.240 --> 00:45:44.560 It is really urgent that we rethink very carefully how we organize the scientific 00:45:44.560 --> 00:45:49.920 enterprise, how we link it to the real world, and also how we advance more integrative 00:45:49.920 --> 00:45:53.060 paradigms to bring these disciplines together. 00:45:53.280 --> 00:45:56.380 Because I think to understand the human condition in the end means we have to 00:45:56.380 --> 00:46:02.920 go from almost, let's say, the physics of bodies and brains to the sociology 00:46:02.920 --> 00:46:05.100 and culture that they can give rise to. 00:46:05.140 --> 00:46:08.400 And right now, we don't have these integrative paradigms. I think also, 00:46:08.420 --> 00:46:14.860 I mean, economics is vitally important for, you know, sort of fulfilling basic human needs, 00:46:15.020 --> 00:46:21.480 but also there are other needs and there are, you know, sort of epidemics of 00:46:21.480 --> 00:46:25.880 social diseases or, I mean, loneliness is not a disease, 00:46:26.020 --> 00:46:31.440 but it's academic portions in a world that has never had higher population. 00:46:32.580 --> 00:46:36.980 And higher levels of social networking, I can add to that. Well, that's true. 00:46:37.120 --> 00:46:41.920 Yeah. So if we can advance theories that can help address that. 00:46:42.000 --> 00:46:48.460 So I think, and partly this is about what collectively we are trying to optimize. 00:46:48.640 --> 00:46:56.320 And so I think the beginning of this has to be some theory about the human life and what it's for. 00:46:56.320 --> 00:46:59.600 You know i mean it's not for anything i don't think as materialistically 00:46:59.600 --> 00:47:02.820 would say but at the same time you know we have 00:47:02.820 --> 00:47:05.960 uh this ability to be aware of ourselves and 00:47:05.960 --> 00:47:08.820 to decide you know uh what we want our life 00:47:08.820 --> 00:47:12.480 to be for uh to set our own goals um and 00:47:12.480 --> 00:47:15.800 so there's uh one 00:47:15.800 --> 00:47:18.880 of the goals i think of our science is to try and help answer 00:47:18.880 --> 00:47:21.680 that question because what personally i wouldn't 00:47:21.680 --> 00:47:27.300 want to do is to spend my life basing my my goals and and my daily efforts on 00:47:27.300 --> 00:47:32.300 uh things which turn out to be illusory and i think that is a risk you know 00:47:32.300 --> 00:47:36.780 that there are lots of false prophets out there giving you reasons to live which 00:47:36.780 --> 00:47:40.340 aren't aren't good ones well this is this is a. 00:47:41.382 --> 00:47:44.842 This is the crux of the whole story, right? Because basically what you're saying 00:47:44.842 --> 00:47:50.042 is that we must go back to the basic question of eudaimonia. 00:47:51.042 --> 00:47:57.522 And that was the reason also why the Greeks never built a rocket to fly to colonize 00:47:57.522 --> 00:48:01.642 Mars, because their main question was, what is the virtuous life? 00:48:02.302 --> 00:48:05.702 And that really means like, what's the good life, right? What is virtue? 00:48:05.962 --> 00:48:11.762 What are the norms we should apply to that? And we have drifted away from those questions a lot. 00:48:11.862 --> 00:48:17.002 We always have felt like, well, there are forces shaping our society that we 00:48:17.002 --> 00:48:20.082 don't need to touch from that scientific perspective, but maybe we should. 00:48:20.162 --> 00:48:23.542 Because in our also definitely more and more secular society, 00:48:23.822 --> 00:48:28.062 where else can we get our norms from, if not from a deep understanding of who 00:48:28.062 --> 00:48:29.682 we are and what our limitations are? 00:48:29.762 --> 00:48:33.402 And I think this is an important responsibility now for psychology and neuroscience 00:48:33.402 --> 00:48:39.082 to pursue, but there are very few people who do that. And if they do it, 00:48:39.082 --> 00:48:43.102 they do it after retirement because it is just a very risky proposition right now. 00:48:43.202 --> 00:48:46.422 Yeah. I mean, I think it was not this Pope, but the last one that said that 00:48:46.422 --> 00:48:49.002 some of these questions are for religion, not for science. 00:48:50.822 --> 00:48:52.862 But you can sort of see what he's 00:48:52.862 --> 00:48:59.302 getting at in that maybe the scientific outlook as it is, is too narrow. 00:48:59.302 --> 00:49:02.782 Arrow um on the other hand uh 00:49:02.782 --> 00:49:05.742 i don't think in in principle that we 00:49:05.742 --> 00:49:10.482 should be excluded from taking a more scientific approach to these questions 00:49:10.482 --> 00:49:16.322 of of how we uh what we should go for in life but perhaps we also have to stretch 00:49:16.322 --> 00:49:20.742 what we're doing as science as scientists to encompass more of the humanities 00:49:20.742 --> 00:49:25.402 as an outlook absolutely what i find interesting there is that. 00:49:26.403 --> 00:49:33.443 And it's very funny that we live in this sort of bubble of happy beliefs about who we are as humans. 00:49:33.923 --> 00:49:39.543 For some reason, we have great difficulties with recognizing also the kind of 00:49:39.543 --> 00:49:45.643 destructive animals we are and definitely can be. 00:49:45.923 --> 00:49:49.843 And we often see that now in our society as sort of anomalies that we can sort 00:49:49.843 --> 00:49:52.043 of, we don't need to worry about too much. 00:49:52.403 --> 00:49:56.083 Criminals, we can lock up their way and for the rest, everything is fine. 00:49:56.403 --> 00:50:00.423 But understanding human condition to me also means especially the destructive 00:50:00.423 --> 00:50:05.383 forces that humans can mobilize because only then can we, if you want, 00:50:05.523 --> 00:50:09.763 defend the future of humanity against those destructive forces. 00:50:09.763 --> 00:50:13.723 And I think I would see that as an important objective of our research. 00:50:13.783 --> 00:50:19.283 Also, for that reason, this is one of the motivations why we are very much involved 00:50:19.283 --> 00:50:24.743 in the study of the Holocaust and the commemoration of the Holocaust and Nazi crimes, 00:50:24.823 --> 00:50:32.543 because I feel this is a source of highly relevant information about the limitations of humanity. 00:50:32.543 --> 00:50:38.283 I mean, however ugly and terrible this is, and it really is terrible, 00:50:38.483 --> 00:50:45.383 we have to understand these boundaries on what humans can accomplish for good 00:50:45.383 --> 00:50:52.183 and bad, because only then can we find countermeasures to protect ourselves from ourselves. Yeah. 00:50:52.443 --> 00:50:56.063 I mean, this is one of the things that has stood out for me in this BCBT is 00:50:56.063 --> 00:50:59.863 actually how many times, sometimes in the talks, 00:51:00.083 --> 00:51:05.723 sometimes in interviews and very often over dinner that the conversations have 00:51:05.723 --> 00:51:10.103 come around to these bigger, if you like, questions about, 00:51:10.303 --> 00:51:16.683 you know, what are we going to do about human society and where it's going? 00:51:17.843 --> 00:51:25.703 How we can take account of the fact that we are, I think, as John Doyle put 00:51:25.703 --> 00:51:29.123 it, the sort of apes with guns, and that's a very dangerous situation. 00:51:29.883 --> 00:51:33.523 And as you say, we have to understand our own limitations better. 00:51:34.361 --> 00:51:40.981 We also are now not just apes with guns, but apes with internets and AIs, 00:51:41.021 --> 00:51:44.401 and there's all the potential that that can have, and robots, 00:51:44.621 --> 00:51:51.021 of course, to both improve our condition, but fundamentally change it in a way. 00:51:51.021 --> 00:51:56.881 Because some of these devices we can use to almost transform what we are. 00:51:56.981 --> 00:52:00.141 Sure, absolutely. And, you know, we're getting to the point, 00:52:00.181 --> 00:52:04.261 and I think this is one of the exciting but also frightening things about this 00:52:04.261 --> 00:52:06.921 area of living machines where we can interface our brains. 00:52:07.301 --> 00:52:10.621 We already do, of course. I mean, a screen is a kind of interface, 00:52:10.781 --> 00:52:16.681 but we are able more and more to connect to these technologies in a very direct 00:52:16.681 --> 00:52:20.501 and intuitive way, which is very exciting. 00:52:20.501 --> 00:52:25.001 But that the implications of that i think for the human mind i think we also 00:52:25.001 --> 00:52:29.881 touched on this you know how technology changes the minds of children yes and 00:52:29.881 --> 00:52:33.401 the minds of our children are going to be different from ours when we were growing 00:52:33.401 --> 00:52:36.961 up so uh and then i think so what we want to do probably. 00:52:37.761 --> 00:52:41.521 More in the future with our school and our uh our 00:52:41.521 --> 00:52:44.821 conference is is to see how we can 00:52:44.821 --> 00:52:48.101 bring all these things together because they do they do seem to be strands that converge 00:52:48.101 --> 00:52:51.681 in a way as you which suggests why we call it convergent science absolutely now 00:52:51.681 --> 00:52:54.701 look i fully agree and to also 00:52:54.701 --> 00:52:58.221 finish up this this last point we were discussing one one 00:52:58.221 --> 00:53:04.121 massive weakness that that the human mind has is that we always adjust our set 00:53:04.121 --> 00:53:10.681 points very quickly you know so um we see disasters happening around us it might 00:53:10.681 --> 00:53:15.281 be mad dictators with nuclear arms or or mad dictators dictators, 00:53:15.481 --> 00:53:19.721 controlling the most powerful nation in the world, it might be hurricanes, 00:53:20.261 --> 00:53:26.081 it might be global warming, the Anthropocene is upon us, it might be self-driving cars, right? 00:53:26.101 --> 00:53:30.581 There's a long list of things that not that long ago we saw as real threats 00:53:30.581 --> 00:53:32.361 and we were really worried about it. 00:53:32.881 --> 00:53:36.741 And very quickly we changed our set point. We are like... 00:53:38.844 --> 00:53:43.964 Cognitive homeostats, you know, so we very quickly zoom into this new set point 00:53:43.964 --> 00:53:52.364 that sits at this new average now of disastrous challenges and it doesn't worry us anymore. 00:53:52.604 --> 00:53:58.784 So I think this is a weakness of the human mind that we are often driven to 00:53:58.784 --> 00:54:06.224 action by homeostatic and also emotional systems that very quickly readjust to these new norms. 00:54:06.224 --> 00:54:11.304 And I think we have to really develop a metacognition here, a critical self-reflection 00:54:11.304 --> 00:54:17.684 that tells us, no, we cannot give in to certain outrages of our human norms. 00:54:17.764 --> 00:54:22.624 And we must insist that we're going to act against them with the best of our science. 00:54:23.444 --> 00:54:27.324 And there, the program, I think, is also partially spelled out already at the 00:54:27.324 --> 00:54:32.184 level of the United Nations where the sustainable development goals have been defined. 00:54:33.784 --> 00:54:38.304 If you look at them in detail, then I think here they're a bit redundant, 00:54:38.464 --> 00:54:42.704 but there's some obvious things like equal opportunity, food security, 00:54:43.304 --> 00:54:45.044 right for education, and so on. 00:54:46.024 --> 00:54:50.764 So that agenda is defined, but what we're completely lacking today is a comprehensive 00:54:50.764 --> 00:54:58.404 science, technology, and also socioeconomic agenda to bring these things about. 00:54:58.404 --> 00:55:04.384 And I really feel that the living machines community and the BCBT community 00:55:04.384 --> 00:55:09.284 that we're trying to grow should start to focus itself more, 00:55:09.364 --> 00:55:13.564 occupy themselves more with those challenges because this is our future, 00:55:13.684 --> 00:55:14.784 this is the future of our children. 00:55:17.257 --> 00:55:19.557 And, of course, we don't want to compromise their future too much. 00:55:19.617 --> 00:55:26.117 So, yes, we have to really rethink to place the school and living machines and 00:55:26.117 --> 00:55:32.617 also the conversion science network on a plateau where we can really start to 00:55:32.617 --> 00:55:35.037 be also agents of change, 00:55:35.237 --> 00:55:39.997 if you want, or positive change by helping us to build new frameworks, look upon ourselves, 00:55:40.337 --> 00:55:43.457 and how we can change the reality for the better. 00:55:43.937 --> 00:55:49.377 I mean, I think you're right that we need to be very much aware of what we might 00:55:49.377 --> 00:55:52.137 be losing as we are gaining new things. 00:55:52.577 --> 00:55:59.337 We also, I think, have to be, and it's the flexibility that you're talking about 00:55:59.337 --> 00:56:03.017 to sort of adapt to culture is almost what's got us here as a species. 00:56:03.017 --> 00:56:06.137 And you know and maybe we are a bit complacent 00:56:06.137 --> 00:56:09.237 but we have in the last hundred years done amazing things 00:56:09.237 --> 00:56:12.037 i think we've reduced absolute poverty in the world from 00:56:12.037 --> 00:56:15.137 80 percent of the population down to below 20 and 00:56:15.137 --> 00:56:19.497 it's forecasted to go down to 10 so there's there are achievements that we can 00:56:19.497 --> 00:56:23.917 point to that things are going in the right direction so i'm i'm not absolutely 00:56:23.917 --> 00:56:27.737 pessimistic about you know the fact that we can't solve these challenges and 00:56:27.737 --> 00:56:31.737 in the past we've always sort of have relied on technology and science as one 00:56:31.737 --> 00:56:34.537 of the paths, not the only one, to resolve challenges. 00:56:34.717 --> 00:56:37.497 And I think that's part of the agenda is to say that, you know, 00:56:37.537 --> 00:56:42.257 we can push science and technology in the right way and we can deal with these problems. 00:56:44.037 --> 00:56:47.417 I don't want to sound too pessimistic because actually intrinsically I'm not. 00:56:47.597 --> 00:56:51.517 I'm an optimist. That's why we engage with many challenging questions. 00:56:51.737 --> 00:56:55.837 But we should be careful not to fool ourselves, right? To claim like, 00:56:55.957 --> 00:56:58.597 okay, humans have done all these great things. 00:56:59.297 --> 00:57:03.997 And I think that's also very doubtful. It's also a story we like to tell ourselves. 00:57:04.137 --> 00:57:06.677 Like, in the end, we were great. We put a man on the moon. 00:57:06.897 --> 00:57:11.037 But if you look, for instance, at the domain of health where we are active, 00:57:11.277 --> 00:57:13.277 the amazing thing is that the 00:57:13.277 --> 00:57:18.217 life expectancy has improved over the last decades dramatically. Right. 00:57:18.707 --> 00:57:23.067 So right now it would be people who are born now will be like over 80. 00:57:23.647 --> 00:57:27.767 And it used to be the 60s, right? So for us, it will be the 70s somewhere. 00:57:27.927 --> 00:57:29.047 So yes, dramatic improvements. 00:57:29.467 --> 00:57:36.647 But if you look at the change in healthy life expectancy, it has stagnated in 00:57:36.647 --> 00:57:39.287 the same period, right? That's very interesting. 00:57:40.047 --> 00:57:44.347 So we can keep people alive longer, but that means it will be miserable longer. 00:57:44.887 --> 00:57:48.287 All right? So we should also… Well, they'd be paying for healthcare for longer. 00:57:48.287 --> 00:57:51.527 Good there might be something there yeah no but but you see that that's 00:57:51.527 --> 00:57:55.027 very tricky so when you say okay poverty has been reduced but what 00:57:55.027 --> 00:57:57.847 does it really mean so have we have built 00:57:57.847 --> 00:58:02.127 maybe a poor middle class that is living every day under all sorts of stress 00:58:02.127 --> 00:58:05.867 because they have to pay off their debts of their credit cards and their their 00:58:05.867 --> 00:58:11.387 banks and their as we have seen also in the the crash that happened in our economies 00:58:11.387 --> 00:58:16.247 about 10 years ago right so so i think we should become more critical about 00:58:16.247 --> 00:58:17.567 these stories we tell ourselves. 00:58:17.747 --> 00:58:21.747 Like also this popular story you hear about, oh, humans have become less violent 00:58:21.747 --> 00:58:23.967 because less people are being killed. 00:58:24.287 --> 00:58:27.887 But maybe that's exactly the same point as with improved life expectancy. 00:58:28.127 --> 00:58:29.687 Yeah, we live longer, but we're miserable. 00:58:30.227 --> 00:58:35.547 So, okay, we're not being killed anymore, but maybe now we're just slowly, 00:58:35.887 --> 00:58:40.007 we're made to die slowly because of cardiovascular disease that's induced by 00:58:40.007 --> 00:58:42.847 stress, but doesn't count anymore as aggression, right? 00:58:42.847 --> 00:58:47.107 So I think we should be more subtle in how we interpret these kinds of so-called 00:58:47.107 --> 00:58:48.287 accomplishments of humanity. 00:58:48.767 --> 00:58:55.267 Yeah, I think we maybe reduce one problem and we create another one somewhere else by doing that. 00:58:55.367 --> 00:58:57.627 And I think that's sort of the way it goes. 00:58:57.807 --> 00:59:03.567 But to some extent, these problems we're talking about are a product of our 00:59:03.567 --> 00:59:05.627 own misconception about ourselves. 00:59:06.007 --> 00:59:11.007 So for me, the fundamental goal or a fundamental goal for our field is. 00:59:11.640 --> 00:59:15.720 To understand what we are and perhaps correct some of those misconceptions. 00:59:16.460 --> 00:59:22.080 So some of those misconceptions, I think you can trace their roots in European 00:59:22.080 --> 00:59:27.140 philosophy and thought and to people that have. 00:59:27.220 --> 00:59:32.380 You know, taken the idea of the soul and then translated it into the modern 00:59:32.380 --> 00:59:35.900 notion of consciousness and this idea that you are a consciousness, 00:59:35.900 --> 00:59:38.940 which is somehow in your body but 00:59:38.940 --> 00:59:41.720 not necessarily of it uh and you know 00:59:41.720 --> 00:59:44.500 that the most recent version of this is the notion that you 00:59:44.500 --> 00:59:47.700 could somehow take that and upload it into a machine 00:59:47.700 --> 00:59:51.320 or a robot and have eternal life and i think this this um 00:59:51.320 --> 00:59:54.920 sort of very western idea of the self is 00:59:54.920 --> 00:59:58.600 at the root of some of these own problems uh our 00:59:58.600 --> 01:00:01.460 own unhappiness for instance uh because i think 01:00:01.460 --> 01:00:04.420 our happiness is linked to all these other physical things not just 01:00:04.420 --> 01:00:07.200 in a sort of the consciousness is floating inside our. 01:00:07.200 --> 01:00:10.100 Heads and then uh a lot of 01:00:10.100 --> 01:00:12.860 these other problems that we have in the world are are down 01:00:12.860 --> 01:00:15.740 to this kind of individualism that we have really sort 01:00:15.740 --> 01:00:19.100 of put on a pedestal in the west exactly so that that's 01:00:19.100 --> 01:00:22.300 a really interesting point you know because in the western 01:00:22.300 --> 01:00:25.580 cultural tradition there is this this myth 01:00:25.580 --> 01:00:29.300 of the the the lone or the individual genius who 01:00:29.300 --> 01:00:32.000 sort of gives rise to change right so we 01:00:32.000 --> 01:00:34.840 think about also in art for instance you would think about the 01:00:34.840 --> 01:00:37.960 great composers mozart beethoven and so on who now 01:00:37.960 --> 01:00:40.980 are sort of dictating their will on all future 01:00:40.980 --> 01:00:45.420 generations of musicians because there's a fixed score from which that piece 01:00:45.420 --> 01:00:51.360 is now forever produced and we seem to apply that model to or to ourselves also 01:00:51.360 --> 01:00:58.640 when we are We have outstanding individuals like the captains of the technology industry. 01:00:59.620 --> 01:01:03.680 They have suddenly forgotten that they might have gotten in that position more 01:01:03.680 --> 01:01:08.700 by sheer luck and contingency and not necessarily by individual genius. 01:01:09.260 --> 01:01:13.280 And so it's really interesting to see that these are the characters that are 01:01:13.280 --> 01:01:18.180 now suddenly pontifying about how they want to upload their mind or how they want to live forever. 01:01:18.180 --> 01:01:24.780 But think about, imagine we would have the mind of Nero on a hard disk somewhere. 01:01:25.220 --> 01:01:28.320 Who would care about turning it on, right? 01:01:28.560 --> 01:01:34.720 Maybe we should have some norms or some ethical rules that would prevent us 01:01:34.720 --> 01:01:38.800 from poisoning the cultural and psychological environment of future generations 01:01:38.800 --> 01:01:44.280 with the complete nonsense that the current self-declared prophets of the techno 01:01:44.280 --> 01:01:47.700 religion seem to have of themselves. 01:01:47.700 --> 01:01:51.320 Yeah, I think that that is a risk and we should, you know, sort of one of the 01:01:51.320 --> 01:01:55.540 things that we should be talking about, actually one of the more near-term risks. 01:01:56.220 --> 01:02:01.220 But I think what we can do in a positive way, I think, is to, 01:02:02.386 --> 01:02:05.366 help people including those people understand themselves 01:02:05.366 --> 01:02:08.846 better because they're suffering from a delusion uh and 01:02:08.846 --> 01:02:14.066 that delusion is probably making them unhappy uh and also you know wishing for 01:02:14.066 --> 01:02:20.246 this silicon afterlife or whatever um and money bro yeah and i think the so 01:02:20.246 --> 01:02:28.026 the the interest that we have in human subjectivity which i know you're doing a lot of work on. 01:02:29.346 --> 01:02:32.186 Should allow us the goal should be to have a 01:02:32.186 --> 01:02:36.786 new understanding of the human self or that helps 01:02:36.786 --> 01:02:40.626 us be more content with what we are because my own 01:02:40.626 --> 01:02:43.726 sort of ideas in this i guess i look to 01:02:43.726 --> 01:02:47.006 sort of more to the western religion eastern religion such as uh 01:02:47.006 --> 01:02:50.246 zen buddhism which i think have had this idea for a 01:02:50.246 --> 01:02:53.566 longer time but they've never you know they got 01:02:53.566 --> 01:02:56.666 us to a certain point with it and i think we can take this 01:02:56.666 --> 01:02:59.706 idea and actually well say why these things that they discovered 01:02:59.706 --> 01:03:02.646 uh might be true um through our understanding of 01:03:02.646 --> 01:03:05.766 the sure but buddha buddha is a bit in some 01:03:05.766 --> 01:03:11.086 sense also a bit frustrating right because hey look unhappiness results from 01:03:11.086 --> 01:03:18.086 wanting things so that means i can be content and avoid unhappiness by not wanting 01:03:18.086 --> 01:03:23.906 things so So now the whole religious exercise is focusing on not wanting things. 01:03:24.726 --> 01:03:27.726 Okay, it's a method. You can try to do that. 01:03:27.786 --> 01:03:33.426 But it also means you're denying part of our humanity, which is we intrinsically want things. 01:03:33.706 --> 01:03:36.926 Yeah. So there might be other ways to deal with that that might be, 01:03:36.946 --> 01:03:41.186 let's say, more liberating and creative than saying, oh, let's not want. Yeah. Right? 01:03:41.846 --> 01:03:45.446 In some sense, the Judeo-Christian tradition is something similar, 01:03:45.566 --> 01:03:51.306 right? We should also not want outside of a very well-defined framework and 01:03:51.306 --> 01:03:55.206 then in some sense use religious meditation to stay within that want. 01:03:55.366 --> 01:04:00.366 And if we exceed the framework, then there are all sorts of rituals to get back into it. 01:04:00.706 --> 01:04:03.846 But maybe we also have to accept the fact that we are religious. 01:04:04.952 --> 01:04:08.532 Wanting things that we do pursue these 01:04:08.532 --> 01:04:11.912 wants often in irrational ways maybe it's 01:04:11.912 --> 01:04:14.752 also through that acceptance that we can find a better 01:04:14.752 --> 01:04:18.172 way to to deal with it because if you look now if 01:04:18.172 --> 01:04:21.152 you look at at the way humans now go off 01:04:21.152 --> 01:04:24.352 the rails in our society for instance through 01:04:24.352 --> 01:04:27.792 addiction right which is a massive problem especially in the 01:04:27.792 --> 01:04:30.532 states um i think this 01:04:30.532 --> 01:04:33.652 tells us something think very deeply about our society not 01:04:33.652 --> 01:04:37.132 being able to deal with people's wants in in 01:04:37.132 --> 01:04:41.032 a sort of a well-managed way and there 01:04:41.032 --> 01:04:43.812 you see for instance here in Europe or also in Holland where I'm coming 01:04:43.812 --> 01:04:46.692 from the attitude towards let's say drugs of 01:04:46.692 --> 01:04:50.892 abuse has been very different and less suppressive and 01:04:50.892 --> 01:04:53.732 the problems of want in that domain of 01:04:53.732 --> 01:04:56.592 addiction are there but it's often more towards alcohol and. 01:04:56.592 --> 01:05:00.372 Less towards the kind of opiates that you see creating havoc 01:05:00.372 --> 01:05:03.192 in the states right so so yes we can then say oh 01:05:03.192 --> 01:05:06.172 let's all become buddhists but okay the west 01:05:06.172 --> 01:05:09.092 is also very expensive to go to these sort of buddhist monasteries to 01:05:09.092 --> 01:05:12.432 meditate and look at the wall and so on but maybe there 01:05:12.432 --> 01:05:15.332 are also more scientific informed methods we can 01:05:15.332 --> 01:05:18.412 apply here to manage our wants through 01:05:18.412 --> 01:05:21.132 let's say other forms of metacognition i mean i think you're 01:05:21.132 --> 01:05:25.132 right and i think uh you know i was um a 01:05:25.132 --> 01:05:28.632 big reader of herman hess when i was young and 01:05:28.632 --> 01:05:31.812 uh his book the glass speed game is uh i 01:05:31.812 --> 01:05:34.912 always read that as a metaphor for cognitive science it 01:05:34.912 --> 01:05:38.072 was these groups of group of people that studied music 01:05:38.072 --> 01:05:40.992 and maths and all these esoteric things 01:05:40.992 --> 01:05:43.672 as a kind of path to enlightenment uh and it 01:05:43.672 --> 01:05:46.692 was one of the things that inspired me to be a cognitive scientist and i think you 01:05:46.692 --> 01:05:49.532 know buddhism is a bit like that there are the monks in in their 01:05:49.532 --> 01:05:52.392 monastery of which there are very few and they meditate all day and then 01:05:52.392 --> 01:05:55.332 there's everybody else uh and not everybody can 01:05:55.332 --> 01:05:58.192 be a monk clearly and not everybody can or wants 01:05:58.192 --> 01:06:01.092 to be a cognitive scientist so what we can do is 01:06:01.092 --> 01:06:03.872 uh you know sort of with the insights that we 01:06:03.872 --> 01:06:06.972 gain we can hope to reshape society and one 01:06:06.972 --> 01:06:13.352 of the things that stood out for me in this school was uh the talk that we had 01:06:13.352 --> 01:06:20.432 on decision making and agency from patrick haggard and you know patrick was 01:06:20.432 --> 01:06:24.272 you know a little bit hazy about uh you know. 01:06:25.054 --> 01:06:27.794 Getting rid of the homunculus, I mean, he wanted to do that, 01:06:27.834 --> 01:06:29.894 but he didn't want to entirely vanish it. 01:06:29.974 --> 01:06:33.594 He still wanted something which was influencing our decision process. 01:06:34.314 --> 01:06:39.054 But I mean, I think for me, perhaps for you, whatever's influencing that decision 01:06:39.054 --> 01:06:41.554 process, there's not going to be a subject in there. 01:06:41.614 --> 01:06:43.634 Ultimately, there's going to be other brain processes. processes, 01:06:43.634 --> 01:06:48.374 maybe some of those processes we are going to label as self-processes, 01:06:48.394 --> 01:06:51.474 and that's getting close to our theory of self. 01:06:51.614 --> 01:06:54.074 But I think what Patrick was saying, and I would agree as well, 01:06:54.214 --> 01:06:57.614 is this is going to change our whole way of thinking. 01:06:57.734 --> 01:07:01.414 If this scientific view percolates into society, 01:07:01.734 --> 01:07:06.254 it's going to change our way of thinking, for example, about how we deal with 01:07:06.254 --> 01:07:07.714 people that abuse drugs, 01:07:07.914 --> 01:07:12.534 the whole notion that we We punish people because they do crimes, 01:07:12.614 --> 01:07:18.254 may have to go out the window, and we may have to globally adopt a more Scandinavian 01:07:18.254 --> 01:07:22.054 view of prisoners' rehabilitation. 01:07:22.894 --> 01:07:28.214 And there might be a path to do that by bringing some of these scientific ideas out to the public. 01:07:28.274 --> 01:07:30.914 Sure. No, I completely agree with that. You're right. Right. 01:07:31.274 --> 01:07:36.274 And what's interesting is I showed it in our own models, in this case of foraging, 01:07:36.434 --> 01:07:44.254 that we have this hidden assumption about that the mind is operating as a single integrated entity. 01:07:45.194 --> 01:07:48.854 But that's an assumption, right? And that's also what Patrick, 01:07:48.934 --> 01:07:50.174 in some sense, was alluding to. 01:07:50.574 --> 01:07:54.254 And also others like Mike Kazanica has been talking about this, 01:07:54.314 --> 01:08:00.114 that you have these sort of dual process ideas about the mind, where, 01:08:00.294 --> 01:08:05.434 or as Mike Kazanica indeed also showed in the split brain patients he worked 01:08:05.434 --> 01:08:07.054 on from Sperry when he was a student, 01:08:07.834 --> 01:08:14.534 but we don't, our metacognitive system that leads to our ability to declare 01:08:14.534 --> 01:08:16.934 and experience and have theories about ourselves, 01:08:17.534 --> 01:08:24.414 might actually be rather disconnected from the subconscious processes that drive 01:08:24.414 --> 01:08:28.974 our behaviors, as for instance, core behavior system that Bjorn Merker would 01:08:28.974 --> 01:08:30.834 talk about, that I mentioned also this morning. 01:08:30.994 --> 01:08:35.474 These are very primitive systems that drive our wants, right? 01:08:35.674 --> 01:08:40.034 And they're automatic, they're strongly genetically defined, 01:08:40.334 --> 01:08:43.194 they're operating really outside of the window of consciousness. 01:08:43.894 --> 01:08:46.454 We have no idea what they do, but we see ourselves doing things, 01:08:46.514 --> 01:08:51.614 right? So I think this hidden assumption of this continuity and transparency 01:08:51.614 --> 01:08:55.194 of operation in the mind is breaking us up here. 01:08:55.314 --> 01:09:00.934 And if we are able to see more also the internal contradictions that exist in 01:09:00.934 --> 01:09:05.254 minds, I think it might help us a lot by devising better interventions. 01:09:06.771 --> 01:09:11.811 So there is a path to increasing contentment through this, but what does it 01:09:11.811 --> 01:09:18.151 involve then if my midbrain is leading me down paths which may be immediately 01:09:18.151 --> 01:09:21.211 rewarding but ultimately bad for me? What's the strategy? 01:09:21.591 --> 01:09:24.491 Well, at least have the metacognition to recognize that. 01:09:24.731 --> 01:09:28.431 I think very often there is no metacognition about it. 01:09:28.431 --> 01:09:31.531 If people don't even know that they 01:09:31.531 --> 01:09:35.051 can be driven by these forces that they 01:09:35.051 --> 01:09:37.991 don't have direct cognitive access to then stuff happens 01:09:37.991 --> 01:09:42.731 to them and they would say like well yeah i just it happened to me i did it 01:09:42.731 --> 01:09:47.171 right but maybe what we have to understand is that through our consciousness 01:09:47.171 --> 01:09:53.331 we can will ourselves into the future we can will ourselves towards being one 01:09:53.331 --> 01:09:56.711 person or another other by biasing our actions against. 01:09:56.811 --> 01:10:00.191 Well, now you're getting a bit mystical because this is, where's this homunculus 01:10:00.191 --> 01:10:03.691 that can will come back from? No, it's a dual process theory, right? 01:10:03.851 --> 01:10:09.451 So I'm saying that, that your executive systems combined with medical systems 01:10:09.451 --> 01:10:11.771 can allow you to build a theory about yourself. 01:10:12.011 --> 01:10:15.831 And we all have this. And these theories can be more or less elaborate because 01:10:15.831 --> 01:10:20.311 they say, well, I like running and cycling and I'm not much into chess nowadays. 01:10:20.751 --> 01:10:24.451 Right? So I have a theory about myself. But these theories can be expanded to 01:10:24.451 --> 01:10:30.391 how it would behave under certain conditions where people might show destructive behavior, right? 01:10:30.571 --> 01:10:36.991 And if we just understand that we are able to have an insight in these factors, 01:10:37.111 --> 01:10:40.391 it's in some sense a very psychoanalytic view on how we operate. 01:10:40.551 --> 01:10:43.231 But yes, I think this can at least be beneficial. 01:10:43.691 --> 01:10:48.371 Yeah. And so by acknowledging its dual process or multiprocess, 01:10:48.391 --> 01:10:51.831 and one process might not have direct access to the other. 01:10:53.052 --> 01:10:56.972 We have to learn to develop theories of ourselves. Right now, 01:10:56.972 --> 01:10:58.112 we don't train people to do that. 01:10:59.312 --> 01:11:03.732 But it sounds a little bit like monks learning Zen Buddhism, 01:11:03.832 --> 01:11:07.072 but we're now training them to think metacognitively. 01:11:07.252 --> 01:11:11.992 Exactly. To say, oh, I have wands, and I have wands that might damage others 01:11:11.992 --> 01:11:14.232 or that might damage myself on the long run. 01:11:14.372 --> 01:11:20.372 So now I can rescale these wands because I have a theory. In the case of the 01:11:20.372 --> 01:11:24.552 Buddhist monks, you have to say, I'm going to meditate the hell out of my wands 01:11:24.552 --> 01:11:26.852 by staring at this wall for a long time. 01:11:27.012 --> 01:11:29.132 I have to reduce the wands that have to disappear. 01:11:29.792 --> 01:11:33.112 While an alternative might be that you say, look, I have to understand myself 01:11:33.112 --> 01:11:38.052 why I have these wands and how I can regulate them without damaging others and myself. 01:11:38.492 --> 01:11:42.912 I'm not necessarily claiming this is a route to success, but I'm just giving 01:11:42.912 --> 01:11:47.132 a hypothetical scenario of how a more rational approach could be deployed to this. 01:11:47.132 --> 01:11:49.752 Yeah yeah but now the one thing i was 01:11:49.752 --> 01:11:52.472 wondering maybe one thing that that you could 01:11:52.472 --> 01:11:55.232 do to this i mean we discuss often how how are 01:11:55.232 --> 01:11:58.132 we going to coexist with technology right right would 01:11:58.132 --> 01:12:00.952 you see technology as as assisting you 01:12:00.952 --> 01:12:05.632 in developing these kinds of math and cognitive theories of of your own behavior 01:12:05.632 --> 01:12:13.652 um you mean as a scientist or or personally well both um well yeah i think we 01:12:13.652 --> 01:12:19.472 we export a lot of our cognition into these devices that we have and they're hugely powerful. 01:12:19.732 --> 01:12:24.392 I would, I mean, I would personally hope that with the advances in AI we are 01:12:24.392 --> 01:12:30.572 now making that machines will take on some of the load of trying to develop theories, 01:12:30.872 --> 01:12:35.532 you know, and I think we can, there perhaps hasn't been enough focus on. 01:12:37.272 --> 01:12:42.272 Sort of automating scientific theory development. I know it's a, there was a, 01:12:43.330 --> 01:12:47.930 a machine a few years ago that rediscovered Newton's laws. So it's maybe several 01:12:47.930 --> 01:12:50.910 hundred years behind where it needs to be to help us now. 01:12:51.130 --> 01:12:57.850 But I do think that we already gain a lot from having all these tools and machines 01:12:57.850 --> 01:13:02.570 and also databases that support our science. 01:13:02.890 --> 01:13:06.090 And I think the role of the scientist is changing as well, 01:13:06.170 --> 01:13:10.050 because we no longer need to remember everything 01:13:10.050 --> 01:13:12.810 thing we've read and have it in our heads we just can access it so 01:13:12.810 --> 01:13:15.950 for somebody like me who's kind of big picture it it's 01:13:15.950 --> 01:13:18.690 great you know i can i i don't have 01:13:18.690 --> 01:13:22.110 to remember the details but i can go and find them when i need them so 01:13:22.110 --> 01:13:27.190 uh i do think that our cognitive capacity to do science is expanded by these 01:13:27.190 --> 01:13:31.830 tools and maybe there is some exciting develops in the future where ais will 01:13:31.830 --> 01:13:39.070 be helping us identify new paths to take but but are you then do you see that develop 01:13:39.390 --> 01:13:43.770 along these kinds of paranoid boss rom scenarios like ai's 01:13:43.770 --> 01:13:47.730 will have a zillion different ways to cheat us out of reality and take over 01:13:47.730 --> 01:13:54.050 or or do you think that that our coexistence these ai's will take a different 01:13:54.050 --> 01:13:59.810 form um i mean i'm kind of quite optimistic about that i mean i uh. 01:14:01.541 --> 01:14:03.801 I think superintelligence will happen. 01:14:04.001 --> 01:14:08.061 I mean, I think there already is AI superintelligence in lots of domains. 01:14:08.321 --> 01:14:12.241 You know, sort of chess playing is one where AIs are much better than us. 01:14:12.781 --> 01:14:18.121 Already, stock trading is another where they do 90% of the stock trades. 01:14:18.121 --> 01:14:20.801 Um so uh there are 01:14:20.801 --> 01:14:24.261 a few domains of human behavior where we're still ahead 01:14:24.261 --> 01:14:27.441 of ais and will be for some time a scientific 01:14:27.441 --> 01:14:30.681 discovery i think is is one that probably be 01:14:30.681 --> 01:14:33.521 okay for a while but uh i do think that in 01:14:33.521 --> 01:14:37.001 these different areas ais are going to advance uh 01:14:37.001 --> 01:14:39.921 i the notion super intelligence is 01:14:39.921 --> 01:14:43.021 i think you you also agree this is is is it's 01:14:43.021 --> 01:14:46.381 not a good idea because i think there are multiple uh kinds of 01:14:46.381 --> 01:14:49.521 intelligence and of course they all correlate so uh 01:14:49.521 --> 01:14:52.361 iq is a good measure across the board 01:14:52.361 --> 01:14:54.941 but it doesn't mean that all of the different parts of your 01:14:54.941 --> 01:14:57.821 intelligence are the same and operate in the same way uh so 01:14:57.821 --> 01:15:00.521 and one of the goals i guess of what 01:15:00.521 --> 01:15:03.701 we're trying to do is to work out what those systems and 01:15:03.701 --> 01:15:06.661 components are um and then build 01:15:06.661 --> 01:15:09.901 a theory of mind based on those i think eventually we will 01:15:09.901 --> 01:15:12.961 be able to build machines that can reason about 01:15:12.961 --> 01:15:16.441 themselves which is one of the defining aspects of 01:15:16.441 --> 01:15:19.181 ourselves um my own view and i think 01:15:19.181 --> 01:15:21.961 if you watch a lot of science fiction movies as i do. 01:15:21.961 --> 01:15:25.181 Uh the the bad ais and the bad robots are 01:15:25.181 --> 01:15:28.201 always the ones that don't understand themselves and 01:15:28.201 --> 01:15:31.121 they have some some mission uh perhaps one 01:15:31.121 --> 01:15:35.021 that's been programmed in by some misguided human uh and 01:15:35.021 --> 01:15:38.561 i think humanity will still be its own biggest uh 01:15:38.561 --> 01:15:41.441 risk even in these coming 01:15:41.441 --> 01:15:45.481 days of ai that people will misuse ai but 01:15:45.481 --> 01:15:48.201 i'm more optimistic that if we build ai's that can 01:15:48.201 --> 01:15:51.701 reason about themselves as ai's they will actually be 01:15:51.701 --> 01:15:55.241 useful and helpful and not be motivated to 01:15:55.241 --> 01:15:57.921 take over and destroy everything else but i do think it's 01:15:57.921 --> 01:16:00.741 right that we that we're thinking and talking about this so of course 01:16:00.741 --> 01:16:04.541 but do you feel that these kinds of alarmist messages 01:16:04.541 --> 01:16:07.521 going around the internet right now 01:16:07.521 --> 01:16:10.541 the social networks and all these declarations being signed 01:16:10.541 --> 01:16:17.281 by scientists like oh stop ai we cannot weaponize ai um we have to this will 01:16:17.281 --> 01:16:21.621 be the most dangerous things we invent right so do you think we should we should 01:16:21.621 --> 01:16:26.921 be jumping on that train as well and sign all these declarations well i did 01:16:26.921 --> 01:16:29.881 sign uh one or two of them so um. 01:16:31.618 --> 01:16:36.638 I think we do have a responsibility as scientists to think about even sort of 01:16:36.638 --> 01:16:40.558 these low likelihood, but very high risk scenarios. 01:16:42.018 --> 01:16:48.778 So, you know, sort of a super intelligence, which would be very anti-human is, 01:16:48.878 --> 01:16:50.338 I think, theoretically possible. 01:16:50.638 --> 01:16:55.678 And we have to think about what are the paths that could lead to that and try 01:16:55.678 --> 01:16:57.998 and make sure that those can't happen. 01:16:58.058 --> 01:17:01.238 So I think it's right that we're already thinking about that now. 01:17:01.618 --> 01:17:08.738 Um, and, but there is the, there are quite a few people who are stoking paranoia, 01:17:08.798 --> 01:17:10.498 which I don't think is necessarily helping. 01:17:11.398 --> 01:17:16.418 Um, and there's also a risk that we, you know, throw out the good with the bad, 01:17:16.558 --> 01:17:20.998 you know, that we, that we decide not to have, uh, AIs and robots that are really 01:17:20.998 --> 01:17:22.458 going to improve the human condition. 01:17:22.458 --> 01:17:26.458 Yeah, but don't you think that it's a bit of a propaganda exercise? 01:17:26.678 --> 01:17:31.298 Because first, we know what goes on in the field, and we know that there's still 01:17:31.298 --> 01:17:33.378 many obstacles before we reach general AI. 01:17:34.738 --> 01:17:41.238 So for scientists to sign these declarations, to me, is also testifying to an 01:17:41.238 --> 01:17:43.718 overstatement of the capabilities of the field. 01:17:43.718 --> 01:17:46.878 Like it's also a little bit hyping a 01:17:46.878 --> 01:17:50.458 field that actually doesn't have the capabilities that it claims to have. 01:17:50.558 --> 01:17:55.098 For instance, take the DARPA robot challenge from 2015, 01:17:55.478 --> 01:18:01.918 where the benchmark was essentially Fukushima nuclear power plant disaster and 01:18:01.918 --> 01:18:05.938 have robots, humanoid robots that can operate in these environments. It was one big disaster. 01:18:06.238 --> 01:18:09.198 And these robots are still remote controlled. In the meantime, 01:18:09.318 --> 01:18:14.238 we have these huge propagandistic statements about, oh, AI is going to take over. 01:18:14.398 --> 01:18:18.838 Well, it's still far removed from that. And then secondly, what worries me is 01:18:18.838 --> 01:18:21.898 that this is driven very strongly by people with business interests. 01:18:22.338 --> 01:18:28.998 And if, for instance, if I'm making money by selling cars that have autopilots…, 01:18:29.998 --> 01:18:34.718 Then indeed, just for my marketing purposes, I would like to give my customers 01:18:34.718 --> 01:18:38.998 the impression or the illusion that I care about their well-being and I care 01:18:38.998 --> 01:18:40.078 about ethical standards. 01:18:40.538 --> 01:18:45.098 So by sort of fanning the flames of, oh, let's be concerned about AI. 01:18:45.518 --> 01:18:50.638 I'm creating trust with my customer so they buy my cars in which they can kill 01:18:50.638 --> 01:18:54.678 themselves if their autopilot is facing an unpredictable situation, 01:18:54.998 --> 01:18:55.838 which also has happened. 01:18:55.838 --> 01:19:00.938 So there's also a hoax behind it, which we should be very careful with. 01:19:01.058 --> 01:19:07.398 And don't forget, as you also said yourself, the biggest threat to humans are humans. 01:19:07.738 --> 01:19:11.918 And before the AI will take over, we might already have killed ourselves. 01:19:12.138 --> 01:19:16.538 So I think AI should be much more worried about the real problems we're facing. 01:19:16.538 --> 01:19:21.538 And I can tell you, if you go to applied domains where we need more science, 01:19:21.678 --> 01:19:27.418 education, healthcare, and so on, we are still far removed from having effective systems. 01:19:27.578 --> 01:19:32.458 And I wish that we would put more energy in that direction as opposed to running 01:19:32.458 --> 01:19:37.718 around like headless chickens who lost their intelligence to be worried about 01:19:37.718 --> 01:19:38.658 artificial intelligence. 01:19:38.658 --> 01:19:43.258 Because on top of that, there's this very strange and naive belief that there 01:19:43.258 --> 01:19:46.478 will be this discrete moment in time where it's all going to happen. 01:19:46.558 --> 01:19:55.118 And again, we always tune our standards, our baseline to the reality we find ourselves in. 01:19:55.198 --> 01:19:59.418 So we seem to forget that we are, as you said, already co-evolving with our technology. 01:19:59.678 --> 01:20:04.058 So it is a very gradual process where we change as the technology changes. 01:20:04.058 --> 01:20:08.638 So this discrete point of transition, like Skynet woke up and now it's conscious, 01:20:08.818 --> 01:20:13.098 because that's essentially the scenario they have in mind. It's just Terminator, you know. 01:20:13.838 --> 01:20:15.838 So these guys don't even have any imagination. 01:20:17.078 --> 01:20:21.418 It's just not what we see happening around us. Yeah, I mean, I think you're right. 01:20:21.578 --> 01:20:25.158 I think there are people who are using this as a smokescreen, 01:20:25.258 --> 01:20:29.378 you know, look over there while I do this, and hopefully you won't notice. 01:20:29.378 --> 01:20:36.178 I think there's another bunch of people riding this anti-AI bandwagon, and they are, 01:20:36.318 --> 01:20:40.798 I think, people who are developing this older version or defending this older 01:20:40.798 --> 01:20:46.258 version of humanism that see actually AI robotics as a threat to their whole 01:20:46.258 --> 01:20:48.558 conception of the human condition. 01:20:48.558 --> 01:20:52.818 Because if we were to build a robot that could walk and talk and describe itself 01:20:52.818 --> 01:20:58.198 as having internal states, then that would really challenge a lot of preconceptions 01:20:58.198 --> 01:21:00.938 that, uh, people have about themselves. 01:21:01.098 --> 01:21:05.558 Sure. Um, so I think, I think people find that threatening and I think, 01:21:05.558 --> 01:21:08.838 uh, you know, robotics in particular has, uh. 01:21:09.833 --> 01:21:16.633 You know, it's almost taken over from the zombie or the psychopath as, 01:21:16.833 --> 01:21:19.113 you know, the sort of the bad guy in the movie. 01:21:19.313 --> 01:21:23.973 And there's a reason for that, actually, because robots are much better psychopaths 01:21:23.973 --> 01:21:24.733 than a human psychopath. 01:21:24.953 --> 01:21:29.613 Robots are even more devoid of emotion or compassion, potentially. 01:21:30.053 --> 01:21:35.393 So it fits with our cultural imagination to see these things as dangerous. 01:21:35.453 --> 01:21:37.993 So it's an easy story for the media to sell. 01:21:38.133 --> 01:21:41.653 Of course. I mean, for all of these reasons, yes, it's in the headlines a lot, 01:21:41.733 --> 01:21:43.973 perhaps far more than it needs to be. 01:21:44.073 --> 01:21:49.213 And what would be better, I guess, is if we can have, and I think it's partly 01:21:49.213 --> 01:21:55.113 the fault of the community not to do more to communicate about the science and the research. 01:21:55.113 --> 01:22:00.113 I mean, so like you say, sometimes people look behind the curtain and, 01:22:00.213 --> 01:22:05.633 you know, they see the Wizard of Oz there controlling his DARPA robot and the 01:22:05.633 --> 01:22:11.133 DARPA robot falls over and it becomes evident that maybe things aren't as advanced as they might be. 01:22:11.133 --> 01:22:14.553 On the other hand, if you look at the previous DARPA road challenge, 01:22:14.953 --> 01:22:19.453 the initial one, the cars got a couple of miles and then crashed. 01:22:19.613 --> 01:22:21.333 But then a few years later, they 01:22:21.333 --> 01:22:25.713 were driving hundreds of miles across the Arizona desert by themselves. 01:22:25.933 --> 01:22:31.153 So you can make rapid progress. So it's not an exponential by any means, 01:22:31.233 --> 01:22:36.013 but there are some sort of sharp lifts which take you up to another level. 01:22:36.173 --> 01:22:40.133 And perhaps we have just had one. I mean, I think people are extrapolating from 01:22:40.133 --> 01:22:45.573 the recent advances in machine learning to say that we're up on the exponential. 01:22:45.793 --> 01:22:48.693 I think we're not. We've jumped up a level. 01:22:49.193 --> 01:22:54.093 And we're now maybe flatline again for a while. And then there might be another sharp jump. Okay. 01:22:54.153 --> 01:22:59.533 But don't forget with the self-driving challenge where the Stanford robot Stanley 01:22:59.533 --> 01:23:07.553 won, the team that won was the team with the most planners of GPS waypoints for the car to reach. 01:23:07.733 --> 01:23:13.133 And the only sort of adaptive element that made the difference between Stanley 01:23:13.133 --> 01:23:18.253 and Carnegie Mellon was that Stanley had a little reinforcement learning system 01:23:18.253 --> 01:23:20.733 that would accelerate when there were no obstacles on the road. 01:23:20.733 --> 01:23:24.533 But the whole planning was done by humans prior to the start. 01:23:25.133 --> 01:23:29.593 So again, much less impressive than you think it is, right? So we should be 01:23:29.593 --> 01:23:31.253 very careful how we interpret those results. 01:23:31.453 --> 01:23:36.173 But look, so you're in this human brain project as well, which, 01:23:36.293 --> 01:23:40.573 okay, you are an exception, but it's largely a waste of money, in my opinion. 01:23:40.893 --> 01:23:45.293 And we should do something about that because we cannot afford in the face of 01:23:45.293 --> 01:23:48.233 limited resources to throw it away like that. 01:23:48.233 --> 01:23:56.453 But imagine I give you 10 million, let's give you euros because the pound has no value anymore. So... 01:23:57.761 --> 01:24:03.501 If I give you 10 million euros for a 10-year project, what would you dedicate that money to? 01:24:04.661 --> 01:24:09.781 In one word, what's the key concept or question you would pursue with that? 01:24:10.321 --> 01:24:13.601 So this is just a normal-sized project. What about a flagship? 01:24:14.041 --> 01:24:17.061 A billion flagships. I'll give you a billion. No problem. Okay. 01:24:17.161 --> 01:24:21.521 I'll give you free coffee as well. I think if I had a billion euros, 01:24:21.681 --> 01:24:30.361 I would be setting out to build human AI systems that can really think about 01:24:30.361 --> 01:24:32.941 the challenges that the world faces. 01:24:32.941 --> 01:24:35.841 Um i think if you look at the 01:24:35.841 --> 01:24:39.621 ipcc and the way 01:24:39.621 --> 01:24:42.981 that that the international panel for climate 01:24:42.981 --> 01:24:46.281 change and the way that groups of scientists have 01:24:46.281 --> 01:24:49.361 and people also looking at the economic and 01:24:49.361 --> 01:24:52.141 social impacts of climate change these people have come along 01:24:52.141 --> 01:24:55.761 and they've worked with very sophisticated computer models and 01:24:55.761 --> 01:24:58.701 over a period of time they've refined these models and 01:24:58.701 --> 01:25:02.401 they've refined their thinking to the point that we can now with pretty 01:25:02.401 --> 01:25:05.301 high confidence say that the result of human 01:25:05.301 --> 01:25:08.281 activity generating co2 is going to cause 01:25:08.281 --> 01:25:13.961 levels of temperature rise which are going to make life bad for us and as a 01:25:13.961 --> 01:25:18.601 result of that i mean you say you mentioned the power to explain predict and 01:25:18.601 --> 01:25:23.461 control i think that's a really nice example of it because the first of all 01:25:23.461 --> 01:25:26.561 they they had to build systems that could explain data. 01:25:26.961 --> 01:25:30.401 And then they had to build systems that could predict what happened in the future. 01:25:30.721 --> 01:25:34.201 They are now being able to show that they can predict, you know, 01:25:34.241 --> 01:25:37.101 from predictions they made 10 years ago, they can point back and say, 01:25:37.201 --> 01:25:42.281 this is, uh, no, it's at least as bad as we said it was going to be possibly worse. 01:25:42.781 --> 01:25:47.381 Uh, and then they've also got a program of control, which they are proposing. 01:25:47.561 --> 01:25:48.901 It's, it's not fully fledged. 01:25:49.321 --> 01:25:52.841 Uh, and if you look at the paris agreement it looks 01:25:52.841 --> 01:25:56.761 like the whole world with one or two notable exceptions are 01:25:56.761 --> 01:26:00.241 starting to get behind this and so that was uh 01:26:00.241 --> 01:26:02.921 i wouldn't say it's a science driven thing but it 01:26:02.921 --> 01:26:06.021 was driven by a community that really cared about 01:26:06.021 --> 01:26:10.621 the science of the climate and really thought that if we did this in a global 01:26:10.621 --> 01:26:16.401 way we could make it in a difference and i think uh with a billion euros we 01:26:16.401 --> 01:26:20.881 could do the same for some of the other big problems that we face And I think 01:26:20.881 --> 01:26:23.621 the first one we might want to look at is wealth inequality. 01:26:24.261 --> 01:26:29.401 So you would want to build these models to try and understand human social behavior, 01:26:29.741 --> 01:26:35.201 human economy, you know, as we said, bring the psychology and the economics together. Right. 01:26:35.681 --> 01:26:39.481 And bring forth some proposals about how we can reorganize. 01:26:40.595 --> 01:26:43.815 World markets you know sort of maybe rethink the structure 01:26:43.815 --> 01:26:46.875 of capitalism in a way that won't favor us 01:26:46.875 --> 01:26:49.775 going to this extreme because the other alternative is that 01:26:49.775 --> 01:26:52.855 that we're going to end up as a society collapsing in 01:26:52.855 --> 01:26:55.835 some way uh because you know the 01:26:55.835 --> 01:26:59.335 extremes of wealth inequality you can only see it some cities uh 01:26:59.335 --> 01:27:02.775 some people living in gated communities uh with 01:27:02.775 --> 01:27:05.875 more wealth than they can imagine what to do with you know booking trips to 01:27:05.875 --> 01:27:08.695 the moon and things and then uh you know over 01:27:08.695 --> 01:27:11.615 the world people living in extreme poverty so that ought to 01:27:11.615 --> 01:27:14.955 be a priority and i think uh what the 01:27:14.955 --> 01:27:18.055 climate change model has showed is that 01:27:18.055 --> 01:27:20.955 you can really use computers for good to predict 01:27:20.955 --> 01:27:24.495 and then you know with using these 01:27:24.495 --> 01:27:27.375 computers under control this is this is so i 01:27:27.375 --> 01:27:30.435 completely agree with you but and what you also see here is that 01:27:30.435 --> 01:27:34.295 why does the climate change project actually 01:27:34.295 --> 01:27:37.095 work because it gave the tools to 01:27:37.095 --> 01:27:39.995 humanity to change its metacognition about its 01:27:39.995 --> 01:27:42.675 own state yes right so this is 01:27:42.675 --> 01:27:46.995 very powerful and therefore indeed i agree with you the computer models are 01:27:46.995 --> 01:27:50.895 not necessarily immediately there to interfere with that reality but they should 01:27:50.895 --> 01:27:54.775 help us as humans to develop a metacognitive state to understand our condition 01:27:54.775 --> 01:27:59.635 so that we can change it for the better indeed that's one of the things i think 01:27:59.635 --> 01:28:01.735 that i got from nick bostrom's book about. 01:28:02.175 --> 01:28:05.175 Super intelligence because he talks about different kinds 01:28:05.175 --> 01:28:08.315 of super intelligence that you can build and one 01:28:08.315 --> 01:28:11.115 of them is the oracle and the oracle you just go 01:28:11.115 --> 01:28:13.915 along to and like the oracle of ancient greece you can 01:28:13.915 --> 01:28:17.215 ask any question gives you the answer uh and so 01:28:17.215 --> 01:28:20.255 uh essentially the climate change scientists have 01:28:20.255 --> 01:28:23.035 built an oracle they say well what's the temperature going to be like in 10 01:28:23.035 --> 01:28:29.475 years it gives you an answer um you know within some bounds uh and then uh he 01:28:29.475 --> 01:28:33.655 also talks about well we could build genies and genies would be oracle but they'd 01:28:33.655 --> 01:28:37.355 also have power to change stuff and you know genies are potentially much more 01:28:37.355 --> 01:28:40.815 dangerous so if we build these ai oracles. 01:28:41.515 --> 01:28:47.175 Then we still have the control as humanity decide what's the best advice you 01:28:47.175 --> 01:28:51.355 can give us in this situation the oracle was said if you did this then there 01:28:51.355 --> 01:28:55.395 might be less wealth inequality and then we can then choose or not choose to 01:28:55.395 --> 01:29:00.255 do that right now i think over time perhaps we probably will do what the AI says, 01:29:00.555 --> 01:29:03.995 because it will turn out that those predictions are good. Well, 01:29:03.995 --> 01:29:05.515 we will shape it. We will also shape it. Exactly. 01:29:05.715 --> 01:29:09.615 And it will be a human machine system. So there will never, I don't think ever 01:29:09.615 --> 01:29:12.955 in these climate change systems are they just running an algorithm. 01:29:13.055 --> 01:29:17.095 It's people interpreting data, putting it into the algorithm, 01:29:17.315 --> 01:29:20.415 tweaking the algorithm, working out, improving it all the time. 01:29:20.515 --> 01:29:26.755 So it's a human machine system which understands an important aspect of our world. 01:29:26.875 --> 01:29:31.195 So we should build more of those. Sure. Actually, in 2005, as one of our art 01:29:31.195 --> 01:29:32.855 projects, we presented the Synthetic Oracle. 01:29:33.726 --> 01:29:36.726 And the whole idea was indeed that you engage with, let's say, 01:29:36.746 --> 01:29:42.746 an interactive sound and light composition that depends on your actions to create, 01:29:42.906 --> 01:29:49.346 let's say, a very implicit kind of immersive experience that helps you to meditate on your state. 01:29:49.586 --> 01:29:51.106 Okay. And actually, it was very 01:29:51.106 --> 01:29:53.206 effective. Of course, it didn't tell you anything about climate change. 01:29:53.406 --> 01:29:57.406 But I think this – so I think we have a principle here, right? 01:29:57.406 --> 01:30:00.726 It's all about also changing our metacognition so we can understand our own 01:30:00.726 --> 01:30:07.046 situation, that we can step in and make a difference, as was done with the Paris Climate Accord. 01:30:07.126 --> 01:30:11.046 But now we have to scale it up because our problems are much broader than that. 01:30:11.506 --> 01:30:15.706 So, okay, Tony, look, I'll get back to you once I have the billion, 01:30:15.886 --> 01:30:18.946 but I think we have a good plan for the future. 01:30:20.046 --> 01:30:26.106 I forgot to sort of try to rip some holes in your theories about the brain, 01:30:26.186 --> 01:30:27.186 but I'll do that next time. 01:30:27.406 --> 01:30:29.446 Well, thank you very much for this conversation. Thank you. 01:30:33.486 --> 01:30:39.506 The CSN Podcast was produced by the Convergent Science Network of Biometrics 01:30:39.506 --> 01:30:45.946 and Biohybrid Systems, a project funded by the European Sevens Research Framework Program. 01:30:47.586 --> 01:30:52.786 For more interviews, recorded lectures, or upcoming conferences in the field 01:30:52.786 --> 01:30:59.026 of biometrics and biohybrid systems, go to csnnetwork.eu. 01:31:00.080 --> 01:31:07.600 Music.