WEBVTT

00:00:03.617 --> 00:00:06.477
This is the Convergent Science Network podcast.

00:00:08.517 --> 00:00:13.017
Leading researchers in the domain of neuroscience, brain theory and technology

00:00:13.017 --> 00:00:16.717
are interviewed by Paul Verschure and Tony Prescott.

00:00:20.417 --> 00:00:24.837
This is Paul Verschure with the Convergent Science Network podcast together

00:00:24.837 --> 00:00:26.457
with my colleague Tony Prescott.

00:00:27.557 --> 00:00:30.837
And we have here as our guest José Aloy. Hello, José.

00:00:32.537 --> 00:00:37.917
José was a speaker as well at our BCBT Summer School 2016.

00:00:40.237 --> 00:00:45.377
So, José, you came from physics and now actually what you're studying is collective

00:00:45.377 --> 00:00:50.557
behavior of animals and also in hybrid sort of animal-robot societies.

00:00:51.377 --> 00:00:56.897
So, how did you go from physics into this domain of the ethology or neuroethology

00:00:56.897 --> 00:01:04.377
of behavior? Well, it's because in physics, people are interested in collective systems anyway.

00:01:04.557 --> 00:01:09.177
Statistical physics deals with collective systems. Then you have dynamical systems

00:01:09.177 --> 00:01:11.057
that can deal with collective systems.

00:01:11.317 --> 00:01:17.697
And so it came to the mind of people in that field that you can use the tools

00:01:17.697 --> 00:01:24.877
of physics to build up mathematical models to describe experiments done with animals.

00:01:24.877 --> 00:01:29.377
And for instance, I started to work on that with Jean-Louis de Nouveau in Brussels,

00:01:29.617 --> 00:01:33.417
and he was very successful in showing

00:01:33.417 --> 00:01:36.657
that he was capable of building

00:01:36.657 --> 00:01:41.657
experiments that led to mathematical models based on differential equation or

00:01:41.657 --> 00:01:47.737
stochastic equation that gave extremely good results to describe what uns societies

00:01:47.737 --> 00:01:53.737
and other collective behavior in animals were happening.

00:01:53.737 --> 00:02:02.037
So from there, Jean-Louis had the idea that because all those results had a

00:02:02.037 --> 00:02:05.797
high impact in the new field of collective robotics.

00:02:07.214 --> 00:02:11.934
People in collective robotics were trying to mimic also, building also a collection

00:02:11.934 --> 00:02:14.734
of agents that together can perform a task.

00:02:15.474 --> 00:02:20.754
They were getting inspiration from biology and it came, a very simple idea.

00:02:20.994 --> 00:02:26.574
If we have animals on one side and you have a robot, biomimetic robot on the

00:02:26.574 --> 00:02:31.114
other side mimicking their behavior, if you succeed in mixing them,

00:02:31.274 --> 00:02:35.314
you get a new group, a bio-hybrid group of robots and animals.

00:02:35.314 --> 00:02:40.534
And because you're able to tune the behavior of the robot, you're capable of

00:02:40.534 --> 00:02:42.734
tuning the behavior of the whole group.

00:02:42.974 --> 00:02:47.734
Because essentially what the mathematical models were showing is that those

00:02:47.734 --> 00:02:51.854
collective behaviors can be described by emergent properties of the system.

00:02:52.314 --> 00:02:56.554
Which means that there is no clear hierarchy. Even if the hierarchy is completely

00:02:56.554 --> 00:03:01.754
flat, every individual is equal, you still get emergent behavior.

00:03:01.754 --> 00:03:07.914
And then if you inject a few individuals in the system, you can tune the whole system.

00:03:08.214 --> 00:03:12.774
And it's a field that is now quite common in physics.

00:03:13.094 --> 00:03:18.114
More generally, it's the field of active matters, where now physics is trying

00:03:18.114 --> 00:03:24.294
to build from matter systems that reproduce some of the collective behavior

00:03:24.294 --> 00:03:26.954
observed in higher animals, like ants, bees,

00:03:27.074 --> 00:03:30.234
or birds, or fish, at the level of matter.

00:03:30.234 --> 00:03:35.994
So there has always been an interplay between statistical physics and animal collective behavior.

00:03:36.494 --> 00:03:39.394
Right. But in some sense, you're

00:03:39.394 --> 00:03:43.154
also then proposing that you can build robots or use robots as a probe.

00:03:44.123 --> 00:03:49.443
Into an animal population to understand what the principles are by which such a population operates.

00:03:49.903 --> 00:03:56.123
Yeah, it's a tool also. So you can test if your model works when it is embodied.

00:03:56.443 --> 00:04:03.023
Because of course you can show that simulations reproduce results that are similar to the animals.

00:04:03.223 --> 00:04:07.923
But then it's even better if you can prove that when you embody that model in

00:04:07.923 --> 00:04:14.823
a completely different agent, which is a robot, that has nothing to do with the animal, Still,

00:04:15.123 --> 00:04:20.883
it works, which proves that your model is capturing something true from collective behavior.

00:04:21.283 --> 00:04:25.943
And then, of course, if you try to mix those robots inside the animal group,

00:04:26.023 --> 00:04:27.003
you're testing hypotheses.

00:04:28.403 --> 00:04:33.843
And you're testing modulating parameters. It corresponds to building an experiment

00:04:33.843 --> 00:04:37.423
where you're able to modulate a parameter of the collective,

00:04:37.683 --> 00:04:40.303
which is not very easy to do experimentally.

00:04:40.343 --> 00:04:44.643
You can imagine other systems. For instance, people have been trying to teach

00:04:44.643 --> 00:04:46.923
animals something, to inform them.

00:04:47.063 --> 00:04:52.463
Then you re-inject them in a group of naive animals.

00:04:52.803 --> 00:04:57.623
And because they already know, have learned the solution, they are influencing the whole group.

00:04:57.843 --> 00:05:01.863
Or you can even imagine doing genetic mutations and some of the individuals

00:05:01.863 --> 00:05:04.223
have different capabilities, etc.

00:05:04.783 --> 00:05:08.683
But learning on some animals is very difficult or impossible.

00:05:08.683 --> 00:05:15.123
And genetic mutation, the link between genetic mutation and higher level behavior is very weak.

00:05:15.303 --> 00:05:20.103
I mean, it's not very easy to do, especially with animals like ants, bees or fish.

00:05:20.323 --> 00:05:23.803
It's complicated. So the robot might be an interesting tool,

00:05:23.923 --> 00:05:28.163
a kind of microscope to probe collective behavior at the animal level.

00:05:30.080 --> 00:05:35.480
So you're proposing that we can model the behavior of a group of animals using

00:05:35.480 --> 00:05:37.960
a small set of differential equations,

00:05:38.280 --> 00:05:46.480
and that's quite a strong claim because we're obviously going to ignore most

00:05:46.480 --> 00:05:51.100
of the richness of those animals and what they're like and how they're made up,

00:05:51.140 --> 00:05:55.520
and we're going to look at some very high-level property which we can describe mathematically.

00:05:55.800 --> 00:05:58.520
Why do physicists think that this is going to succeed?

00:05:59.080 --> 00:06:04.140
Okay, so it has been proved that it works on some experimental physical cases,

00:06:04.140 --> 00:06:08.300
like for ants, for bees, for fish, for birds.

00:06:08.500 --> 00:06:15.560
Those models have been capturing the essential mechanism of the group,

00:06:15.660 --> 00:06:21.120
how it works. But of course, and it's the philosophy of a physical approach

00:06:21.120 --> 00:06:23.080
of systems, you want a minimal model.

00:06:23.820 --> 00:06:30.560
You want the most simple model capable of capturing what you're observing in an experiment, right?

00:06:30.700 --> 00:06:36.860
But of course, it doesn't mean that these mathematical models are a model of the animal.

00:06:37.760 --> 00:06:44.020
They are not. They're just capturing a specific mechanism in a specific experiment.

00:06:44.020 --> 00:06:46.840
And so there are strong limitations to that.

00:06:47.000 --> 00:06:52.960
I mean, for the last 25 or 30 years, there have been a lot of success in animal

00:06:52.960 --> 00:06:57.100
collective behavior, but still, when you look at the global result,

00:06:57.400 --> 00:07:01.060
there are not so many models that give interesting results.

00:07:01.060 --> 00:07:08.120
But still, the idea that you can have simple mechanism, simple mechanism that

00:07:08.120 --> 00:07:13.160
produce interesting collective behavior without taking into account the whole

00:07:13.160 --> 00:07:14.600
complexity of the animal.

00:07:15.703 --> 00:07:21.343
Is a message. You see what I mean? So you described an experiment with cockroaches

00:07:21.343 --> 00:07:27.583
where you have two containers and you put the cockroaches, I think,

00:07:27.603 --> 00:07:29.743
randomly into the containers and then you see how they,

00:07:30.503 --> 00:07:35.603
choose to go into one bin or another and you use differential equations to show

00:07:35.603 --> 00:07:38.163
how that system might evolve over time.

00:07:39.283 --> 00:07:42.943
And I guess one of the things that can be confusing about this.

00:07:43.063 --> 00:07:47.203
Is your model a description of the system,

00:07:47.343 --> 00:07:51.623
or is it not capturing the causality in the system in the way that,

00:07:51.643 --> 00:07:56.323
say, a mechanistic model of individual cockroaches and how they make decisions

00:07:56.323 --> 00:07:58.643
would capture causality?

00:07:58.963 --> 00:08:03.483
I think it's both. It gives you a description of the system,

00:08:03.563 --> 00:08:08.783
and it gives you also the cause of the solution that emerged in the system.

00:08:08.903 --> 00:08:13.643
It gives you the mechanism that produce this emergent behavior at the population level.

00:08:13.923 --> 00:08:21.503
But still, you can go down at the individual level and find yet other causes

00:08:21.503 --> 00:08:25.903
that lead to this solution that emerge in the system.

00:08:26.523 --> 00:08:31.283
And what those models have shown, if you think in terms of dynamical system.

00:08:32.223 --> 00:08:39.463
You have a network of feedback regulation in the system based on positive feedback and negative feedback.

00:08:39.863 --> 00:08:45.923
And nonlinear effects, right? And if you mix that, you get emergence of interesting properties.

00:08:46.323 --> 00:08:51.223
But that thing, you find it at all levels of living systems.

00:08:51.443 --> 00:08:55.103
You find it at the genetic level, you find it at the metabolic level,

00:08:55.223 --> 00:09:00.303
you find it at the physiological level, and you find it at the population level.

00:09:00.563 --> 00:09:06.663
Those abstract concepts of network of regulatory feedback and nonlinear effect

00:09:06.803 --> 00:09:12.563
leading to the emergence of interesting patterns has been a lesson of the domain.

00:09:13.263 --> 00:09:17.143
But it seems to me that there are different aspects to this, right?

00:09:17.243 --> 00:09:22.063
I think on the one hand, you're saying, I don't need to model the animal as such.

00:09:22.303 --> 00:09:27.043
I just want to have a model that allows me to do a meaningful perturbation of

00:09:27.043 --> 00:09:29.643
that animal group or individual.

00:09:30.723 --> 00:09:35.523
And that then also brings this whole issue, okay, but what would be then a sufficient approximation.

00:09:37.363 --> 00:09:41.143
Of a meaningful intervention in such a group. And also in that context,

00:09:41.243 --> 00:09:44.763
you actually spent quite some time discussing the Turing test, right?

00:09:44.863 --> 00:09:48.123
Because in some sense, in the cockroach experiment, as Tony mentioned,

00:09:48.483 --> 00:09:50.783
you have this problem of imitation.

00:09:50.863 --> 00:09:55.743
You want to sufficiently imitate an individual cockroach to have a meaningful

00:09:55.743 --> 00:09:58.263
perturbation of the cockroach group, right?

00:09:58.323 --> 00:10:01.863
So would you, if you take the cockroach experiment as an example,

00:10:01.923 --> 00:10:07.043
would you really see that as an insect Turing test you're performing or it's different?

00:10:07.363 --> 00:10:10.583
Well, it's somehow, I think if

00:10:10.583 --> 00:10:17.123
you reduce the Turing test to a social election, I mean, it's like that.

00:10:17.163 --> 00:10:22.663
You're interacting with an artificial system and at a certain point you agree

00:10:22.663 --> 00:10:27.743
to interact with that system because it's becoming interesting enough for you to interact with.

00:10:28.123 --> 00:10:31.663
So animals are not at the symbolic level, they don't exchange,

00:10:31.883 --> 00:10:37.303
they don't, cockroaches or many, nearly all animals, maybe primate do that,

00:10:37.423 --> 00:10:41.583
but all animals don't really communicate symbolically, like they don't have

00:10:41.583 --> 00:10:44.383
the language, the same kind of language humans have.

00:10:45.083 --> 00:10:50.363
So their interaction are based on other modalities.

00:10:51.503 --> 00:10:56.883
And then, yes, if you prove that your robot is accepted by the group,

00:10:57.743 --> 00:11:03.243
And if the animals take into account that robot as a member of the group,

00:11:03.283 --> 00:11:07.903
whatever that means, I'm not saying that the cockroaches believe or think or

00:11:07.903 --> 00:11:11.963
whatever they do that the robot is a cockroach.

00:11:12.303 --> 00:11:17.583
We cannot answer that question, but they do take into account the presence of

00:11:17.583 --> 00:11:21.443
the robot as another member of the group. And because they do that.

00:11:22.933 --> 00:11:27.193
The robot is capable of influencing the whole group decision-making process.

00:11:27.933 --> 00:11:31.753
But what is interesting in the cockroach experiment, it's because we designed

00:11:31.753 --> 00:11:37.053
a biomimetic, a very simple, but yet biomimetic at the behavioral level.

00:11:37.113 --> 00:11:41.693
It was biomimetic at the behavioral level, not the shape, not the material and stuff like that.

00:11:42.173 --> 00:11:47.853
A social robot that was programmed to behave like a cockroach in the same experiment,

00:11:48.053 --> 00:11:51.993
the robot were taking also into account what the cockroaches were doing.

00:11:51.993 --> 00:11:55.313
And so they were also influenced by the cockroaches.

00:11:55.413 --> 00:12:00.793
And sometimes the cockroaches were driving the robots as a group to a place

00:12:00.793 --> 00:12:04.673
they wouldn't have chosen alone because they were programmed to prefer,

00:12:04.893 --> 00:12:06.493
to have another preference, for instance.

00:12:07.673 --> 00:12:13.753
You say that the robot is biomimetic in terms of its behavior. Yeah.

00:12:14.413 --> 00:12:19.733
But the control system that drives the robot, I think, is a finite state machine.

00:12:19.953 --> 00:12:26.333
Yeah. And that's not meant to emulate a cockroach nervous system or anything like this.

00:12:26.473 --> 00:12:33.173
So how do you get a cockroach-like robot that has behavior like a cockroach?

00:12:33.473 --> 00:12:35.093
Yeah, that's exactly the point.

00:12:35.153 --> 00:12:40.053
In fact, you don't have to copy all the levels of the living system.

00:12:40.964 --> 00:12:46.984
In fact, it's based on the idea that also has been discussed in physics.

00:12:47.064 --> 00:12:52.544
In the famous paper by Paul Anderson, Moore is different, that he published.

00:12:52.824 --> 00:12:56.524
He was a physicist of solid-state system.

00:12:56.724 --> 00:13:00.564
And he knows that you have emergence of properties at a certain level.

00:13:00.764 --> 00:13:07.884
Of course, they depend on the lower level. The properties of a solid depends on the atoms.

00:13:08.124 --> 00:13:12.704
It's obvious. But you can have a model, a level of description of the solid

00:13:12.704 --> 00:13:16.784
that is good enough to explain what the solid is about.

00:13:17.084 --> 00:13:22.204
So it means you don't need the nervous system of the cockroach necessarily.

00:13:22.424 --> 00:13:27.024
You don't necessarily need the nervous system of the cockroach to get the behavior.

00:13:27.304 --> 00:13:32.244
It can be based on another lower mechanism, like the finite state machine.

00:13:32.244 --> 00:13:37.764
Machine, but that doesn't exclude the fact that you would like to have also

00:13:37.764 --> 00:13:41.804
a biomimetic mechanism at the neural level. That's another question also.

00:13:42.044 --> 00:13:47.684
It's allowed, I mean. You could do that. You could add an extra layer of mimetism

00:13:47.684 --> 00:13:51.584
that you want to understand if you want to pile.

00:13:53.644 --> 00:13:59.244
Neural models, neural net models, biological relevant neural net models with

00:13:59.244 --> 00:14:02.324
individual behavior, with social behavior. That's another question.

00:14:02.564 --> 00:14:07.604
But you don't necessarily need to include all levels to get the result,

00:14:07.684 --> 00:14:12.704
because each level has its own properties somehow. You see what I mean? Yeah.

00:14:13.544 --> 00:14:18.864
So you're tweaking the robot control system to give it cockroach-like behavior,

00:14:18.984 --> 00:14:25.004
which will then generate results as though there was another cockroach in the group.

00:14:25.744 --> 00:14:31.344
I mean, aren't you worried there are a lot of degrees of freedom for you to tweak? And so...

00:14:32.264 --> 00:14:37.224
Yes, there were many, in fact. And what was interesting is that the model.

00:14:38.144 --> 00:14:42.024
This little set, very simple set of differential equations, were predicting

00:14:42.024 --> 00:14:44.664
many different possibilities to control the group.

00:14:45.502 --> 00:14:50.542
And we only tested experiment in one way of controlling the group.

00:14:50.742 --> 00:14:54.822
But we know from simulation and from solving the equation there are many other ways.

00:14:55.122 --> 00:14:59.542
For instance, you can build social robots that are social among themselves,

00:14:59.842 --> 00:15:02.682
but they do not take care of what the cockroaches are doing.

00:15:02.862 --> 00:15:06.342
You get a different set of solutions, so a different set of modulation.

00:15:06.722 --> 00:15:10.602
You can build robots that are completely non-social. They

00:15:10.602 --> 00:15:13.402
do not take care either of the other robots or the

00:15:13.402 --> 00:15:17.082
cockroaches you get different results and so

00:15:17.082 --> 00:15:19.962
on so you can or we chose to to

00:15:19.962 --> 00:15:23.922
to show the the case where the robot were the

00:15:23.922 --> 00:15:28.702
same behaving in the same way as the cockroaches but the model shows that you

00:15:28.702 --> 00:15:33.382
have many other different uh social capability social capability social interaction

00:15:33.382 --> 00:15:38.582
leading to different solution and of course the intensity of the relay of the

00:15:38.582 --> 00:15:41.642
interaction is also you can also modulate it.

00:15:41.722 --> 00:15:46.822
So you can add a parameter in the model that modulates the probability to respond

00:15:46.822 --> 00:15:48.122
to the presence of the other.

00:15:48.302 --> 00:15:51.082
In the case of cockroach, let's say this probability is one,

00:15:51.182 --> 00:15:54.722
because a cockroach considers another cockroach as a cockroach, so it's one.

00:15:55.042 --> 00:15:59.802
But you can, for the robot, you can reduce it. We have put one also in the experiment.

00:16:00.022 --> 00:16:03.102
We did. But in the simulation, you can put less than one.

00:16:03.302 --> 00:16:07.942
And then you can see what level of interaction, what intensity of interaction

00:16:07.942 --> 00:16:09.622
you need to still get an effect.

00:16:10.422 --> 00:16:15.842
It doesn't need necessarily to be one. Then you can also modulate the number

00:16:15.842 --> 00:16:17.622
of robots you're adding to the system.

00:16:18.162 --> 00:16:25.862
Our idea was to show that you need only a minimal number of robots to influence

00:16:25.862 --> 00:16:29.962
the system because it's a non-linear effect, which is interesting.

00:16:29.962 --> 00:16:36.202
But, of course, you can increase the number of robots in the system and you

00:16:36.202 --> 00:16:39.442
will get different kind of solution or modulation.

00:16:41.055 --> 00:16:46.675
But now, it was the case that you had to make the robot smell like a cockroach

00:16:46.675 --> 00:16:52.175
to be accepted, while in some sense, the shape didn't really matter, right?

00:16:52.495 --> 00:16:58.655
So, what then would be the minimal robot that you think would be plausible for

00:16:58.655 --> 00:17:03.935
a cockroach colony? So that's exactly the point.

00:17:04.035 --> 00:17:12.735
In fact, it's the crucial question, is how to make the animal respond and accept the robot.

00:17:12.935 --> 00:17:17.155
And it's very difficult because animals are multimodal. They use all their senses

00:17:17.155 --> 00:17:19.155
to perceive their environment and the others.

00:17:19.155 --> 00:17:22.775
And we were lucky enough, and it's well known, that insects,

00:17:22.975 --> 00:17:25.595
and in particular cockroaches, they

00:17:25.595 --> 00:17:31.515
base their recognition on tactile and olfactory cues more than vision.

00:17:31.715 --> 00:17:37.495
So the shape does not matter. What matters for them is that you smell like a

00:17:37.495 --> 00:17:39.115
member of the group, and that's enough.

00:17:39.835 --> 00:17:45.735
And of course, any mobile robot that is capable of moving around in the system

00:17:45.735 --> 00:17:51.075
and detecting the presence of shelter in terms of a different intensity of light

00:17:51.075 --> 00:17:55.035
below or outside of the shelter would be enough.

00:17:55.195 --> 00:18:00.055
Any shape. And of course, you need a size that is compatible with the physics.

00:18:00.155 --> 00:18:04.575
I mean, you're not going to put a robot 10 times larger than the cockroaches

00:18:04.575 --> 00:18:06.815
because it's a completely different physical world.

00:18:06.935 --> 00:18:10.295
But the shape in that case doesn't matter. But then we've tried with chicken,

00:18:10.375 --> 00:18:13.635
we're trying with fish, and it's a completely different story.

00:18:14.633 --> 00:18:18.473
Because those vertebrates are a bit more tricky. They are much more multimodal.

00:18:18.513 --> 00:18:22.613
So they do care about vision. They do care about shape.

00:18:22.933 --> 00:18:26.613
They also care about smell. They also care about behavior.

00:18:26.833 --> 00:18:32.253
So that's why I think, for the moment, trying to interact with fish is pretty difficult.

00:18:32.473 --> 00:18:35.733
There are a few groups that have been trying to do that.

00:18:35.853 --> 00:18:40.313
And it's much more difficult to get results with the cockroaches.

00:18:40.313 --> 00:18:47.133
Because I think fish are a much more multimodal animal, and they do take care

00:18:47.133 --> 00:18:50.573
of many different inputs from the sensory system.

00:18:51.313 --> 00:18:55.893
That's why shape matters, colors matters, speed matters, movements matters,

00:18:56.053 --> 00:18:57.833
much more than with the cockroaches.

00:18:58.053 --> 00:19:04.213
So basically, I think you may say we were lucky, but I don't think.

00:19:04.213 --> 00:19:11.093
I think Jean-Louis was clever enough to choose those insects because he has

00:19:11.093 --> 00:19:18.553
the intuition that it's going to work because it's an olfactory recognition essentially.

00:19:18.993 --> 00:19:23.493
But then in the cockroach experiment where you looked basically at the ability

00:19:23.493 --> 00:19:27.413
of groups of cockroaches or a hybrid group cockroach with robots,

00:19:27.833 --> 00:19:32.993
how they would disperse in an environment or aggregate in an environment depending on conditions.

00:19:33.173 --> 00:19:35.733
So they would aggregate under these shelters, right?

00:19:35.973 --> 00:19:43.113
And the notion that was tested there was called collegial decision making, right?

00:19:43.253 --> 00:19:46.713
So what does it really mean, collegial decision making in this context?

00:19:46.933 --> 00:19:51.753
There is a consensus emerging in the system. And this consensus emerges from

00:19:51.753 --> 00:19:55.233
individuals that are considered as perfect clone of each other.

00:19:55.613 --> 00:20:02.213
What the model is saying that even if you have a population of exactly identical

00:20:02.213 --> 00:20:09.993
individuals, you get the collective decision mechanism and the clever way of doing groups.

00:20:10.233 --> 00:20:15.613
I'm not saying that the cockroach population we're using were clonal individuals.

00:20:15.973 --> 00:20:20.053
No, there is a lot of inter-individual differences.

00:20:20.513 --> 00:20:26.313
But what the model shows that in completely total absence sense of any hierarchy,

00:20:26.633 --> 00:20:32.153
any difference from the individuals, yet you get a collective system producing

00:20:32.153 --> 00:20:37.633
this consensus on where to gather, which is not such a trivial task.

00:20:38.173 --> 00:20:43.373
Imagine you take a population of about 100 humans without specific structure,

00:20:43.573 --> 00:20:47.373
hierarchy structure in the group, naive, let's say, individuals,

00:20:48.293 --> 00:20:51.233
100 of them, and then you ask them, okay, your job is to split.

00:20:52.299 --> 00:20:59.899
50-50% between two rooms. Then you look at the process of how that decision is going to emerge.

00:21:00.299 --> 00:21:03.859
And it will take some time. We take people arguing, people discussing,

00:21:04.119 --> 00:21:09.619
people starting to count, going from one room to the other to check how many

00:21:09.619 --> 00:21:12.759
are you, how many are there on the other side. You can imagine that experiment.

00:21:13.059 --> 00:21:22.159
And it's going to be a complicated mechanism to produce this 50-50% land spread.

00:21:22.819 --> 00:21:26.199
Cockroaches don't do that at all. It's a very simplistic mechanism.

00:21:26.639 --> 00:21:30.519
They just move around. They know they are in a room. Interesting.

00:21:30.859 --> 00:21:34.559
They look around if they have enough bodies and the body, sorry.

00:21:34.739 --> 00:21:40.879
And then if you have this threshold function on the probability to get to go out, that makes a trick.

00:21:41.139 --> 00:21:45.839
So it means in terms of cognitive capabilities, it's very simple, fairly simple.

00:21:46.139 --> 00:21:51.739
What those models are showing, it goes back to your previous question,

00:21:51.839 --> 00:21:58.259
Tony, what those mathematical models are showing is even if you have high cognitive capabilities,

00:21:58.719 --> 00:22:02.599
you don't necessarily need to use them all the time, and in particular,

00:22:02.739 --> 00:22:05.179
to produce that kind of collective behavior.

00:22:06.136 --> 00:22:11.376
For instance, that's the same kind of approach that people who are modeling crowd movement.

00:22:11.636 --> 00:22:16.796
Of course, crowd movement with human beings. Human beings keep their high-level

00:22:16.796 --> 00:22:19.256
capabilities, cognitive capabilities, all the time.

00:22:19.376 --> 00:22:24.036
But when they are moving in a crowd, they're not necessarily using them.

00:22:24.276 --> 00:22:30.556
They are using much simpler mechanisms, and yet you get a structure in the crowd

00:22:30.556 --> 00:22:34.116
that emerges from the people moving.

00:22:34.816 --> 00:22:39.636
So in your collegial decision-making, you have sort of two opposing forces that

00:22:39.636 --> 00:22:40.716
guide the behavior, right?

00:22:40.776 --> 00:22:45.496
On the one hand, they want to seek shelter, and on the other hand, they want to aggregate.

00:22:45.976 --> 00:22:50.076
But that means if aggregation leads to a lack of shelter, then they will switch

00:22:50.076 --> 00:22:51.636
and look for another shelter.

00:22:51.896 --> 00:22:54.436
Yeah. This is essentially what happens, right? Yeah, of course.

00:22:54.496 --> 00:22:59.096
If the shelter is saturated, they cannot enter and they have to look for another place.

00:23:00.076 --> 00:23:03.916
But that means they would always aggregate. gate. This is a key driver.

00:23:04.216 --> 00:23:09.856
Those species of cockroaches are really a gregarious animal.

00:23:10.116 --> 00:23:13.276
They really want to be as a group, to shelter as a group.

00:23:13.716 --> 00:23:18.456
And it's been proved that they feel better and they grow better if they live

00:23:18.456 --> 00:23:23.316
in a social environment as a group than when they are isolated.

00:23:23.716 --> 00:23:28.236
And it's been shown even by biologists that if you tickle them with a plume,

00:23:29.236 --> 00:23:30.296
their physiology is better.

00:23:30.516 --> 00:23:34.876
So there's really animals that feel better as a group. So for them,

00:23:34.916 --> 00:23:37.216
it's important to aggregate.

00:23:38.456 --> 00:23:43.036
But then to describe those experiments, so we have this bifurcation model, if you want, right?

00:23:43.116 --> 00:23:50.376
There's sort of a moment of sort of scurrying around. There's exploration, if you want.

00:23:50.536 --> 00:23:55.496
And then very quickly, the population falls into this distribution of seeking

00:23:55.496 --> 00:23:58.136
shelter below the different shelters that they're offered.

00:23:58.336 --> 00:24:02.436
And then you interpreted that by saying that there's an interaction with positive

00:24:02.436 --> 00:24:04.716
and negative feedback. That there's positive feedback,

00:24:05.893 --> 00:24:08.673
generated by the animal, the internal control of the animal,

00:24:08.773 --> 00:24:11.393
and there's negative feedback coming from the environment.

00:24:11.673 --> 00:24:14.813
So how should I interpret that relative to this experiment exactly?

00:24:15.293 --> 00:24:21.453
Yeah, that's what we find. That's what also has been shown by those mathematical models.

00:24:21.873 --> 00:24:26.293
And it boils down to dynamical system theory that shows that,

00:24:26.373 --> 00:24:32.473
for instance, if a choice can be described by multiple stable steady state existing

00:24:32.473 --> 00:24:34.773
as a solution of the system,

00:24:34.933 --> 00:24:41.813
and those multiple steady states require the presence of a positive feedback in the system.

00:24:42.353 --> 00:24:46.113
But of course, a positive feedback is like a snowball effect.

00:24:46.473 --> 00:24:52.313
It's an amplification of what's going on. Then you need a limiting mechanism,

00:24:52.633 --> 00:24:57.273
otherwise the system is going to explode, to blow up, because these amplification

00:24:57.273 --> 00:24:59.093
mechanisms keep going on, going on.

00:24:59.093 --> 00:25:04.673
So you always find in physical system, in biological system, a limitation.

00:25:06.073 --> 00:25:10.373
But then what is interesting in the models, all the models I know,

00:25:10.453 --> 00:25:16.993
is that usually the positive feedback is implemented somehow at the individual level.

00:25:17.673 --> 00:25:24.293
It's very often a mimetic effect that drives the positive feedback.

00:25:24.293 --> 00:25:30.593
And the negative feedback is given by an environment constraint.

00:25:30.953 --> 00:25:33.153
Not enough room anymore.

00:25:34.293 --> 00:25:38.073
Not enough room in the shelter, then it kills the positive feedback.

00:25:38.193 --> 00:25:39.593
Nobody can enter anymore.

00:25:40.373 --> 00:25:46.153
So that is interesting. And you find that in all, again, at all level of living systems.

00:25:46.993 --> 00:25:51.213
You can describe a metabolic system as a network of positive and negative regulation.

00:25:51.213 --> 00:25:55.153
A genetic system as a network of positive and negative regulation.

00:25:55.873 --> 00:26:02.213
Neural nets, there are those with a positive and a negative feedback, inhibitor, activator.

00:26:02.393 --> 00:26:09.253
So you find that this theory of dynamical system gives you that you can build

00:26:09.253 --> 00:26:13.513
artificial system and that you can see that natural system behave,

00:26:13.793 --> 00:26:19.653
the mechanism under explaining their behavior is this network of positive and negative feedback.

00:26:21.512 --> 00:26:27.892
This dynamical systems approach that you have, so you took it from statistical

00:26:27.892 --> 00:26:31.712
physics, you brought it into animal behavior, specifically in crop roaches.

00:26:32.192 --> 00:26:35.712
And then from there, you want to try and generalize it to other species.

00:26:36.292 --> 00:26:39.712
Is that going to be an easy step? Or as you were saying before,

00:26:39.912 --> 00:26:41.692
other animals are more complicated.

00:26:42.212 --> 00:26:45.052
And does that mean that the approach is not going to scale up?

00:26:45.052 --> 00:26:51.732
So, in fact, there are a few examples, after all, that have been described like that.

00:26:53.432 --> 00:26:59.072
The classical examples are ants, or the cockroaches we've been discussing,

00:26:59.352 --> 00:27:05.632
some models with the bees, many models of schooling and schooling in fish.

00:27:05.872 --> 00:27:14.492
And then you have bird flocks, because those are very large groups of very large number of individuals.

00:27:15.052 --> 00:27:21.212
Where a physical model, a simple model inspired by the physics method,

00:27:21.432 --> 00:27:23.392
does something interesting.

00:27:23.812 --> 00:27:31.692
Then, de Nebourg again has shown that for some specific behavior in primates,

00:27:31.772 --> 00:27:37.332
you can still model that, that way, from some specific behavior.

00:27:37.692 --> 00:27:41.712
That doesn't mean you're describing the whole animal. It's just you're describing

00:27:41.712 --> 00:27:47.312
a mechanism that they use in some specific decision-making process,

00:27:47.432 --> 00:27:50.072
like who chooses the direction to go.

00:27:50.192 --> 00:27:55.812
So you have a group of primates sitting commonly there in their park,

00:27:55.932 --> 00:28:01.392
and then suddenly you trigger the group to start a movement towards some place.

00:28:01.492 --> 00:28:03.532
But then who is deciding where to go?

00:28:04.417 --> 00:28:08.757
And again, if you have a mechanism that is not based on a strict hierarchy,

00:28:09.077 --> 00:28:11.737
the boss is saying, everybody to do right.

00:28:13.577 --> 00:28:17.157
In cabbage and monkeys, it's not what is happening. So it's,

00:28:17.157 --> 00:28:23.737
again, some kind of self-organizing system that is driving the system.

00:28:23.897 --> 00:28:30.137
But it doesn't mean that all collective behavior or any other individual behavior

00:28:30.137 --> 00:28:32.237
could be described that way.

00:28:33.437 --> 00:28:37.957
So I'm not pretending that you can describe any kind of behavior or any kind

00:28:37.957 --> 00:28:39.277
of colleague behavior that way.

00:28:39.437 --> 00:28:47.097
In fact, it's very tedious to build those models because since I'm a student and getting older,

00:28:47.317 --> 00:28:54.737
I've seen people doing that and still there are not that many examples because

00:28:54.737 --> 00:29:00.117
of course animal behavior and collective behavior is much more complicated.

00:29:01.817 --> 00:29:06.657
Than, let's say, gas temperature emergence or something like that.

00:29:06.717 --> 00:29:12.097
People have taken these kinds of models and they've applied them also to phenomena of human perception.

00:29:12.417 --> 00:29:16.997
So, for example, the NECA cube, where you see this three-dimensional or this

00:29:16.997 --> 00:29:21.097
two-dimensional shape, you interpret it in one three-dimensional way and then flip into another.

00:29:22.537 --> 00:29:28.497
And that, I think, is a similar approach. So, in a way, is this technique sort of.

00:29:29.884 --> 00:29:34.224
Too sort of general, that it fits all these different kinds of systems,

00:29:34.384 --> 00:29:36.324
and so it tells us a little bit about them,

00:29:36.544 --> 00:29:43.444
but can it really give us useful insights into these systems and how we might,

00:29:43.484 --> 00:29:49.404
say, sort of interface with animal colonies more effectively or understand their

00:29:49.404 --> 00:29:50.384
behavior more effectively?

00:29:50.724 --> 00:29:53.724
Yeah, that's the research question, the ongoing research question,

00:29:53.824 --> 00:29:56.264
is what can you describe by this approach?

00:29:56.604 --> 00:30:01.424
And again, this approach works at very different level, but let's stay at the collective behavior.

00:30:01.704 --> 00:30:08.624
And then if you understand that, and if the model works for specific cases,

00:30:08.884 --> 00:30:12.984
then what we've learned also is that if you build artificial systems,

00:30:13.364 --> 00:30:15.464
they reproduce the same kind of behavior.

00:30:15.924 --> 00:30:20.504
For instance, what was surprising is that when Jean-Ryder Nabour did his AMT

00:30:20.504 --> 00:30:22.524
model for path selection,

00:30:22.864 --> 00:30:26.224
collective path selection by ants, people in

00:30:26.224 --> 00:30:29.044
computer science and robotics were inspired by that

00:30:29.044 --> 00:30:31.884
and then certainly we saw the the emergence of

00:30:31.884 --> 00:30:35.104
this un-colony optimization field people would

00:30:35.104 --> 00:30:37.924
that took up that model and then they started to use

00:30:37.924 --> 00:30:44.744
it as a heuristic for either network handling or optimization which was completely

00:30:44.744 --> 00:30:48.484
unexpected and they got apparently they use a whole field of that and they get

00:30:48.484 --> 00:30:53.524
interesting result then of course we've shown repetitively that you if you build

00:30:53.524 --> 00:30:56.824
robots that they can and reproduce those kind of behavior.

00:30:57.104 --> 00:31:02.484
But we are yet at a quite limited thing because we don't have a full description

00:31:02.484 --> 00:31:05.004
of the individual, if that's your question.

00:31:05.204 --> 00:31:08.104
It's never a full description of the individual.

00:31:08.344 --> 00:31:12.964
It's a description of a mechanism that takes place in a very specific case.

00:31:13.941 --> 00:31:21.241
So you may say, okay, it's not very useful because it's very narrow as a way

00:31:21.241 --> 00:31:23.241
to work. That's true. It's quite narrow.

00:31:23.441 --> 00:31:25.641
Not so many things explained.

00:31:25.921 --> 00:31:31.841
But that doesn't mean that more broadly, a dynamical system is not useful.

00:31:32.121 --> 00:31:37.781
Because again, if you think about the dynamical system as this network of positive

00:31:37.781 --> 00:31:44.881
feedback and negative feedback and non-linear effects, you find them to describe metabolic,

00:31:45.241 --> 00:31:47.701
genetic, neural nets.

00:31:47.921 --> 00:31:50.941
So the methodology of modeling is interesting.

00:31:51.121 --> 00:31:58.261
But I'm not claiming that it's a way to fully describe an individual in all

00:31:58.261 --> 00:32:00.121
occasions during all his life.

00:32:00.501 --> 00:32:04.761
But there's a problem here, right? Because I could also argue natural language

00:32:04.761 --> 00:32:09.281
works really well to describe cockroach behavior, so it's a valid method. it.

00:32:09.721 --> 00:32:14.261
But I think the question is, how does it help us to gain additional insight

00:32:14.261 --> 00:32:16.061
in the generation of behavior?

00:32:16.341 --> 00:32:19.481
So it's also a little bit, what's the leverage that it gives us, right?

00:32:19.561 --> 00:32:23.001
And there, I think there's an interesting issue because in some sense,

00:32:23.041 --> 00:32:27.421
the way you conceptualize behavior is in complete operational terms,

00:32:27.581 --> 00:32:30.661
also as the agent being under full control of its environment.

00:32:31.041 --> 00:32:34.381
Everything is in a direct control of external stimuli.

00:32:34.901 --> 00:32:38.921
And this, of course, was the same premise of which behaviorism built

00:32:38.921 --> 00:32:41.661
its science and also created its

00:32:41.661 --> 00:32:45.201
own doom because internal factors play

00:32:45.201 --> 00:32:50.321
a decisive role and you have to account for those as well so of course raises

00:32:50.321 --> 00:32:53.701
on the question for the for instance your case with the cockroach to what extent

00:32:53.701 --> 00:32:59.181
those internal states of that cockroach matter let's say hunger reproduction

00:32:59.181 --> 00:33:03.521
fear right so to what extent do these

00:33:03.621 --> 00:33:06.101
internal factors play a role, like motivational states,

00:33:06.301 --> 00:33:10.721
and then if so, how well can you capture those in that framework?

00:33:11.862 --> 00:33:17.782
First, they do matter, of course. And second, they are not captured by the model

00:33:17.782 --> 00:33:23.602
because the model is, again, the experiments were done in very specific conditions.

00:33:25.762 --> 00:33:29.322
So that we don't have to take care of those internal states.

00:33:29.482 --> 00:33:33.102
For instance, the cockroaches were well-fed or they were starving.

00:33:33.522 --> 00:33:36.522
So we are sure they start to explore, to look for food.

00:33:36.542 --> 00:33:40.142
There is no food in the system, so they're going to explore the system and stuff

00:33:40.142 --> 00:33:42.602
like that. They were kept in the dark

00:33:42.602 --> 00:33:48.622
so we're sure that all in the same physiologically sensitivity to light.

00:33:48.802 --> 00:33:51.802
And then you put them in an environment that has light.

00:33:51.922 --> 00:33:58.562
So the model only capture, and that's the limitation, with this specific experiment,

00:33:58.862 --> 00:34:04.102
given rather specific initial condition in terms of internal state of the animal,

00:34:04.242 --> 00:34:06.522
during a short lapse of time.

00:34:06.782 --> 00:34:11.902
So you may say, okay, but that's pretty narrow. And it is, but still we are

00:34:11.902 --> 00:34:13.882
learning, you know, it's a scientific process.

00:34:14.182 --> 00:34:17.342
It's very narrow, but still we are learning things.

00:34:17.542 --> 00:34:22.442
But then I could still argue, I could have then confirmed that model by running

00:34:22.442 --> 00:34:27.762
a pure cockroach-based experiment without inserting robots, right?

00:34:27.802 --> 00:34:30.042
Without inserting the robot probes. Yeah.

00:34:30.442 --> 00:34:36.162
And it has been done. Yeah. Yeah, so what's then the added advantage of also

00:34:36.162 --> 00:34:40.642
using the robot to test that very operational theory of the behavior?

00:34:41.042 --> 00:34:44.642
So there were multiple goals in that experiment.

00:34:45.162 --> 00:34:50.782
One of the goals was for roboticists, because roboticists from the EPFL were involved.

00:34:51.122 --> 00:34:55.342
We didn't build the robots, it's their job. So for them, for roboticists,

00:34:55.362 --> 00:34:56.602
it's interesting to build,

00:34:56.662 --> 00:35:01.722
in collective robotics, It's interesting for them to build biomimetic robots

00:35:01.722 --> 00:35:08.042
that are capable of performing a task and having a solution that looks clever, looks intelligent.

00:35:08.242 --> 00:35:14.062
So the motivation was, how do I build robots that, as a collective,

00:35:14.342 --> 00:35:15.962
can do interesting things?

00:35:16.142 --> 00:35:19.122
Okay, let's go biomimetic. That's their motivation.

00:35:19.582 --> 00:35:24.642
Our motivation was, okay, we can do experiments with cockroaches,

00:35:24.662 --> 00:35:28.822
but we can never really tune certain parameters,

00:35:29.082 --> 00:35:34.762
the number of informed individuals or the number of the type of interaction.

00:35:35.042 --> 00:35:39.822
If we do that, like a probe exactly that you inject in the system,

00:35:39.982 --> 00:35:43.322
you have another degree of freedom to perform experiments.

00:35:43.962 --> 00:35:50.022
And that's also a motivation. Then there is also another motivation that is

00:35:50.022 --> 00:35:57.102
hanging there, but we don't have any results result yet, but all we keep thinking about that is, okay,

00:35:57.282 --> 00:36:02.742
if you have this bio-hybrid system made of artificial agents and of natural

00:36:02.742 --> 00:36:05.202
agents, can it do something more?

00:36:06.317 --> 00:36:09.597
Uh than the the putting them together

00:36:09.597 --> 00:36:13.417
can you get something extra there that but

00:36:13.417 --> 00:36:18.517
due to the difficulties the experimental difficulties uh we don't have any results

00:36:18.517 --> 00:36:22.857
that show that you can do something more but at least you can show that you

00:36:22.857 --> 00:36:28.217
can connect uh living an artificial system that artificial system is interesting

00:36:28.217 --> 00:36:31.297
and that you can use them to tune some of the parameters.

00:36:31.677 --> 00:36:34.157
That's already an interesting result.

00:36:34.797 --> 00:36:39.137
So you, in this experiment, I think you already had an idea of what the dynamical

00:36:39.137 --> 00:36:43.877
system looked like, and you could sort of say what the equations might be that

00:36:43.877 --> 00:36:44.777
would govern their behavior.

00:36:45.197 --> 00:36:49.917
But if somebody is out there collecting a data set about animals cooperating

00:36:49.917 --> 00:36:55.397
or animals and robots interacting, how do they go about building a dynamical

00:36:55.397 --> 00:36:59.597
system description of that? Yeah, that's a completely different question.

00:36:59.817 --> 00:37:02.697
In fact, it's a completely, the methodology is completely different.

00:37:02.877 --> 00:37:05.757
So for instance, in the case of the cockroach again,

00:37:06.097 --> 00:37:12.177
and in the case of the ants, you have prior knowledge about those species,

00:37:12.457 --> 00:37:16.477
they were chosen because you have prior knowledge, prior biological knowledge,

00:37:16.717 --> 00:37:20.597
and, and, and Donnebourg invented an experimental methodology,

00:37:21.037 --> 00:37:24.257
more than a modeling method methodology, because the animal system,

00:37:24.477 --> 00:37:25.877
he hasn't invented them.

00:37:25.877 --> 00:37:31.577
I mean, even, so it's an old stuff, but the experimental methodology of binary chunks.

00:37:33.351 --> 00:37:37.731
With prior knowledge of specific species may lead to the fact that,

00:37:37.811 --> 00:37:41.011
okay, you understand if there is a social mechanism that produces,

00:37:41.651 --> 00:37:45.311
that makes this choice emerge. So it's a combination.

00:37:45.671 --> 00:37:49.791
No, for the fish experiments, we don't really have a prior model.

00:37:50.011 --> 00:37:55.271
So we have to build the robot, the fish experiment, all at the same time and

00:37:55.271 --> 00:37:59.291
to try to model the whole system at the same time.

00:37:59.371 --> 00:38:04.211
And it's much more difficult, in fact, because one of the drawbacks of this

00:38:04.211 --> 00:38:11.871
experimental field is you're combining two difficult tasks to perform collective behavior with animals,

00:38:12.851 --> 00:38:17.631
tedious, hard, difficult tasks, with the task of building robots from scratch.

00:38:18.551 --> 00:38:24.011
Tedious, difficult, and to program them and to make them biomimetic and stuff like that.

00:38:24.131 --> 00:38:28.211
So in a project, when you start to do that, it's hard work to get.

00:38:28.371 --> 00:38:34.091
So now you cannot just Science is not about looking around what's going on,

00:38:34.191 --> 00:38:40.231
collecting a lot of data, and magically out of the data you will have the insight that it's working.

00:38:40.351 --> 00:38:44.571
No, that's not the scientific method. At least that's not the scientific method I use.

00:38:44.831 --> 00:38:48.031
You have a predefined question.

00:38:49.011 --> 00:38:54.091
You gather data that you think is going to answer that predefined question.

00:38:54.651 --> 00:38:59.491
Or you build an experiment to gather that data, and out of that you prove that

00:38:59.491 --> 00:39:00.651
the model is working or not.

00:39:00.931 --> 00:39:06.731
So you have to design your experiment or the data acquisition in a way that

00:39:06.731 --> 00:39:13.151
captures, at least you guess, an educated guess that you capture something interesting out of there.

00:39:13.571 --> 00:39:17.911
So if you go to the classical example that Jean-Louis is always laughing about,

00:39:18.031 --> 00:39:21.831
okay, you want data, you go there, you have grass,

00:39:22.191 --> 00:39:25.891
you count the number of grass, little pieces of grass, and you have a lot of

00:39:25.891 --> 00:39:28.651
bunch of data. Are you going to learn something out of that?

00:39:29.570 --> 00:39:32.470
Not sure because it's not the amount of data

00:39:32.470 --> 00:39:35.750
gathering amount of data without having a hypothesis before

00:39:35.750 --> 00:39:38.550
so again it's a it's more an

00:39:38.550 --> 00:39:41.430
experiment experimental method than a

00:39:41.430 --> 00:39:45.990
mathematical model um methodology but there's a problem but there's a challenge

00:39:45.990 --> 00:39:51.050
here right because already the cockroach is a pretty complex organism okay and

00:39:51.050 --> 00:39:55.830
it can engage in many kinds of of complex behaviors it can navigate it has pretty

00:39:55.830 --> 00:39:59.910
good sensory capabilities um So in some sense,

00:39:59.970 --> 00:40:03.870
I could argue, but in your experiment, you push this high degree of freedom

00:40:03.870 --> 00:40:06.090
system in a very low degree of freedom task.

00:40:06.670 --> 00:40:09.430
And now you can show that with a low degree of freedom model,

00:40:09.530 --> 00:40:10.670
you can account for the behavior.

00:40:10.850 --> 00:40:17.150
But maybe you are actually, if you want, indeed counting sort of virtual grassroots

00:40:17.150 --> 00:40:23.090
that are fairly irrelevant towards really understanding what the cockroach is capable of.

00:40:23.090 --> 00:40:27.050
For instance, in the cockroach case, they're also outstanding in terms of their

00:40:27.050 --> 00:40:28.450
chemical sensing capabilities.

00:40:29.530 --> 00:40:31.530
And that is fully ignored in

00:40:31.530 --> 00:40:35.590
the current model and also in the robots that you have to work with them.

00:40:35.870 --> 00:40:42.530
So aren't you worried that there's a risk that your model is actually giving

00:40:42.530 --> 00:40:46.030
you such a low-dimensional description of this animal that it is almost meaningless

00:40:46.030 --> 00:40:50.390
given the complexity of its behavior in its actual ecological niche?

00:40:50.750 --> 00:40:55.490
Yeah, there are two answers to that. First of all, the model has implicit assumption.

00:40:55.990 --> 00:41:02.370
The implicit assumption do take into account the fabulous capabilities of those animals.

00:41:02.810 --> 00:41:06.730
What the model is taking into account is that they are very good at olfaction

00:41:06.730 --> 00:41:10.350
because that's the way they recognize each other.

00:41:10.470 --> 00:41:15.130
But you don't have to make a model of that olfaction necessarily because at

00:41:15.130 --> 00:41:17.910
that level, you can take it for granted. it, okay?

00:41:18.810 --> 00:41:24.470
So, but again, there is the trap of reductionism because it's a reductionist

00:41:24.470 --> 00:41:25.870
method, scientific method.

00:41:26.050 --> 00:41:31.190
And if you reduce too much your system, you may not get interesting results.

00:41:31.430 --> 00:41:33.370
And it's also the danger as.

00:41:34.162 --> 00:41:39.042
Any experiment to build an artifact, an artifact in the experiment,

00:41:39.222 --> 00:41:46.382
a bias experiment that is not relevant for the real stuff, the real system.

00:41:46.542 --> 00:41:50.662
It's always a danger, but it's the danger in every experiment with an experimental

00:41:50.662 --> 00:41:51.622
reductionist approach.

00:41:51.982 --> 00:41:55.762
You can either reduce too much and get not interesting result,

00:41:55.962 --> 00:41:58.282
or you can completely bias your system.

00:41:58.462 --> 00:42:02.042
But that you have all the time I mean, with that scientific method,

00:42:02.142 --> 00:42:06.082
and you have to be careful about that, of course, not to induce some.

00:42:06.302 --> 00:42:11.842
Then we have yet another interesting question, is that we don't have exactly

00:42:11.842 --> 00:42:15.042
the same mindset as the biologists.

00:42:15.442 --> 00:42:19.762
For instance, many biologists in behavioral science would say,

00:42:19.922 --> 00:42:27.742
okay, you have to really take care of what the animals are doing in their wild environment, right?

00:42:27.782 --> 00:42:33.182
In the natural environment. So for them, it's very important to study the animals

00:42:33.182 --> 00:42:36.622
in their real natural conditions, out there in the wild.

00:42:36.722 --> 00:42:39.302
And they are right, of course. We must do that.

00:42:39.482 --> 00:42:45.762
And then each time you bring an animal inside the lab, you're somehow starting

00:42:45.762 --> 00:42:48.422
to build an artifact in terms of experiment.

00:42:48.902 --> 00:42:55.882
Because they are not anymore in their natural environment. And so you may be

00:42:55.882 --> 00:42:59.602
studying something that is irrelevant for their natural environment.

00:43:00.102 --> 00:43:03.462
But I have the mindset of doing more than that.

00:43:04.782 --> 00:43:08.482
Yet, let's assume that it's really an artifact.

00:43:08.702 --> 00:43:12.442
They will never do that in their natural environment because such condition

00:43:12.442 --> 00:43:15.682
doesn't exist. What we are doing is testing the hardware.

00:43:16.742 --> 00:43:23.702
So if I can prove that the cockroach in a completely non-natural setup are capable

00:43:23.702 --> 00:43:25.982
of performing a clever choice,

00:43:26.222 --> 00:43:32.582
that means that this hardware is capable of doing it. And for me, it's interesting.

00:43:32.742 --> 00:43:37.062
For some biologists, they will argue that's irrelevant because they don't need

00:43:37.062 --> 00:43:39.322
to do that in their natural environment.

00:43:39.542 --> 00:43:44.422
It's always this discussion and it's going to be always there.

00:43:44.422 --> 00:43:48.242
But from my point of view, even if you're in a non-natural condition,

00:43:48.502 --> 00:43:52.602
if you prove that the hardware, which means the living system you're studying,

00:43:52.722 --> 00:43:55.522
is capable of doing that, you've learned something out of that.

00:43:56.812 --> 00:44:02.692
From my point of view as a psychologist, I guess what I'm hoping for out of

00:44:02.692 --> 00:44:06.692
this kind of research is some sort of general laws of social behavior that are

00:44:06.692 --> 00:44:11.652
going to apply across species and are going to generalize maybe to some interesting

00:44:11.652 --> 00:44:16.592
situations like people at football matches, things like this.

00:44:16.592 --> 00:44:20.952
And there does seem to be the potential for that with this kind of work.

00:44:21.312 --> 00:44:28.732
But I guess what you're also saying is that to actually demonstrate this concretely

00:44:28.732 --> 00:44:34.212
in terms of describing a set of equations that captures what's going on is very

00:44:34.212 --> 00:44:36.212
difficult and requires a huge amount of data.

00:44:36.732 --> 00:44:40.732
Well, Tony, what's also interesting there, in some sense, with your chicken

00:44:40.732 --> 00:44:42.872
experiment, or the experiment with the chicks,

00:44:43.112 --> 00:44:46.992
in some ways you showed the opposite Because you, in some sense,

00:44:47.012 --> 00:44:51.932
then invalidated the principles that were out there in literature for many decades,

00:44:52.252 --> 00:44:56.012
which is that imprinting works invariably for all chicks, right?

00:44:56.072 --> 00:45:00.952
They hatch, they see a mother-like figure, and they get imprinted,

00:45:00.952 --> 00:45:04.432
all of them invariably, on that object.

00:45:05.272 --> 00:45:08.852
And actually, your experiments that you were describing where you had chicks

00:45:08.852 --> 00:45:11.372
following robots, you saw a much higher variability.

00:45:12.592 --> 00:45:17.092
So, does it make you then more pessimistic about this ambition that Tony is

00:45:17.092 --> 00:45:20.912
sketching about identifying these sort of species independent principles?

00:45:22.272 --> 00:45:25.992
Well, it's a complicated question in fact.

00:45:27.632 --> 00:45:35.072
What you're looking for is somehow the ground of the and stuff like that.

00:45:35.272 --> 00:45:39.592
And it comes down, we may some do a cheap philosophy of science here,

00:45:39.712 --> 00:45:43.932
it comes down to this universal laws that we find in physics.

00:45:44.272 --> 00:45:48.812
Every theoretical physics at least is going to say, okay, Schrodinger equations

00:45:48.812 --> 00:45:54.712
and gravitation, so it's law of the universe. It's a bold statement.

00:45:55.032 --> 00:46:00.152
And we always dream of finding such simple laws because at the end of the day,

00:46:00.192 --> 00:46:03.672
you can write all of them on one single sheet of paper.

00:46:03.812 --> 00:46:07.972
Of course, to use them, it's more complicated, but at least you can write them

00:46:07.972 --> 00:46:13.092
on one single, a whole physics can be written in one single sheet of A4 paper, right?

00:46:13.212 --> 00:46:17.392
And then you are, wow, and that describes so many things and potentially including

00:46:17.392 --> 00:46:18.812
the structure of the universe.

00:46:19.212 --> 00:46:23.612
But if you work in the field of complex systems, and I'm from the statistical

00:46:23.612 --> 00:46:30.852
physics complex system field of studies, then you know that's not true.

00:46:30.852 --> 00:46:35.452
In fact, when you have a complex systems, you don't have a single model that

00:46:35.452 --> 00:46:36.772
captures everything of it.

00:46:36.952 --> 00:46:43.152
You need several models, a kind of kaleidoscope of models to answer different questions.

00:46:45.298 --> 00:46:50.378
Just to understand how this natural complex system works.

00:46:50.678 --> 00:46:54.418
And each time you ask a different question, maybe you need different kind of models.

00:46:54.618 --> 00:46:59.478
And then you say, okay, well, but still you're learning a lot of that complex system.

00:46:59.658 --> 00:47:03.978
But then it becomes a problem if you have a synthetic approach system.

00:47:04.918 --> 00:47:07.718
Because that's what we're really discussing, in fact.

00:47:07.798 --> 00:47:14.398
Let's imagine I want to synthesize something that is similar as a living system.

00:47:14.398 --> 00:47:18.638
I want an artificial cell, or I want an artificial cockroach,

00:47:18.658 --> 00:47:20.318
which is a completely different question.

00:47:20.538 --> 00:47:25.838
Then with this kaleidoscope, you don't know exactly how to use it.

00:47:26.158 --> 00:47:30.378
Because you're saying you will not have a single model of your system,

00:47:30.558 --> 00:47:33.858
then what do you do? Because you don't have the recipe to build it.

00:47:34.178 --> 00:47:39.018
And you are still, I think, I personally still don't know how to synthesize

00:47:39.018 --> 00:47:42.358
from scratch a complete system.

00:47:42.358 --> 00:47:45.438
Them so when we when you do the the cockroach

00:47:45.438 --> 00:47:48.518
robot at the end of the day that robot is not really interesting

00:47:48.518 --> 00:47:51.318
it's doing nothing it's behaving like a cockroach

00:47:51.318 --> 00:47:54.198
in a specific environment if you do if you

00:47:54.198 --> 00:47:57.718
you you look at in collective robotics the ant

00:47:57.718 --> 00:48:02.318
uh robots that they all follow each other and they do a trail a kind of pheromone

00:48:02.318 --> 00:48:07.538
trail whatever mechanism they use to do that great they do it but so what what's

00:48:07.538 --> 00:48:11.798
the use of those robots well they're a minimal description of what the cockroach

00:48:11.798 --> 00:48:14.418
must be doing in that task. Exactly. So that's useful.

00:48:14.698 --> 00:48:19.778
Yeah, but you're not synthesizing the robot that has been built is not a cockroach.

00:48:20.238 --> 00:48:25.298
It's a tiny subset of the cockroach. Now, if you want to build a robot that

00:48:25.298 --> 00:48:30.958
is much closer to what cockroaches are capable of doing, like you were mentioning,

00:48:31.158 --> 00:48:34.018
Paul, they have a lot of capabilities, internal state and so on.

00:48:34.098 --> 00:48:36.718
It's a completely different question and you are nowhere.

00:48:36.998 --> 00:48:41.898
We still have a lot of research to do. Well, just to kind of defend your research

00:48:41.898 --> 00:48:46.018
a minute, I think that what we're finding out with these kinds of experiments

00:48:46.018 --> 00:48:51.258
is how little complexity needs to be in the animal because the environment,

00:48:51.478 --> 00:48:53.398
and that includes the other individuals in it,

00:48:53.678 --> 00:48:58.058
will bring out all of these interesting emergent effects.

00:48:59.918 --> 00:49:03.378
And then that can apply also to us.

00:49:03.418 --> 00:49:06.958
We can say, well, this behavior that I thought I was doing was very sophisticated

00:49:06.958 --> 00:49:11.558
turns out to be some fairly simple control mechanism in my brain that's responding

00:49:11.558 --> 00:49:14.678
to the environment in this way.

00:49:14.918 --> 00:49:19.258
So it's a sort of classic Simon argument that the complexity is not in the ant.

00:49:19.358 --> 00:49:21.278
It's in the ant and in the environment interaction.

00:49:21.798 --> 00:49:25.598
Yeah, that I agree. That's one of the major lessons is that you don't need a

00:49:25.598 --> 00:49:29.698
very high complexity to produce what you are observing.

00:49:29.998 --> 00:49:34.238
But yet we have to think again about that. Well, it's a challenge we discussed earlier, right?

00:49:34.258 --> 00:49:38.478
You do constrain the degrees of freedom of the organism a lot.

00:49:39.150 --> 00:49:42.190
And then you can say, okay, it's a simple control system. But you already know,

00:49:42.330 --> 00:49:47.870
as you saw with the chicks and with the fish, suddenly life gets way more complicated

00:49:47.870 --> 00:49:50.490
and these simple rules already don't hold anymore, right?

00:49:50.510 --> 00:49:51.950
So I found it interesting that

00:49:51.950 --> 00:49:57.010
with the chicks, you actually observed this massive variability, right?

00:49:57.090 --> 00:50:01.450
That seems to be really systematic, right? So in that sense,

00:50:01.570 --> 00:50:08.490
maybe the notion of simple rules should also be critically analyzed because

00:50:08.490 --> 00:50:12.530
maybe a simple rule also is a simple rule that should imply the ability to generate

00:50:12.530 --> 00:50:14.910
huge variability across a population.

00:50:15.110 --> 00:50:18.510
And we're not really used to that, right? We want to think about gravitational

00:50:18.510 --> 00:50:21.930
forces and that's sort of pretty deterministic in that sense.

00:50:22.910 --> 00:50:26.830
So, but what you also did in the case, in the experiment with the chicks and

00:50:26.830 --> 00:50:32.890
also with the fish, is you start to automate much more of how you process this data, right?

00:50:32.950 --> 00:50:37.410
So you get sort of to an etonome, how would you call that? Atomics,

00:50:37.550 --> 00:50:39.050
it's not me. Atomics, okay.

00:50:39.570 --> 00:50:44.250
So that means you start to now capture massive amounts of data from behavior

00:50:44.250 --> 00:50:46.770
in an automated fashion, right?

00:50:46.830 --> 00:50:50.270
And in the case of the chicks, that really paid off because this allowed you

00:50:50.270 --> 00:50:54.390
to observe something most people had ignored, which is this variability, right?

00:50:54.430 --> 00:50:59.590
That you had sort of a larger portion of the animals indeed showing the imprinting,

00:51:00.550 --> 00:51:05.750
a smaller part actually not showing it, and an even smaller part actively avoiding

00:51:05.750 --> 00:51:08.290
this mother object, yeah?

00:51:09.010 --> 00:51:15.110
So do you see this sort of massive scanning of behavior in this automated fashion,

00:51:15.110 --> 00:51:20.610
session, including then inserting robots that now automatically get optimized

00:51:20.610 --> 00:51:24.150
to realize certain perturbations as the future of neuro-ethology?

00:51:24.430 --> 00:51:26.690
Yeah, that's one of the dreams, in fact.

00:51:27.050 --> 00:51:32.210
And still it's a lot of work. It's to try to automate as much as we can the

00:51:32.210 --> 00:51:36.790
theoretical and the experimental method.

00:51:37.755 --> 00:51:43.675
So to automate the experiments and to automate the data analysis and to automate the model generation.

00:51:44.155 --> 00:51:51.035
And it's a challenge. It's completely open. But it's sure if we make progress

00:51:51.035 --> 00:51:58.415
along that line, then we will be capable of maybe starting to do experiment,

00:51:58.795 --> 00:52:01.795
model production, embodiment of the model.

00:52:02.155 --> 00:52:07.535
Embodiment implies coding the controls of those robots.

00:52:07.755 --> 00:52:10.855
And all that, at the moment, take years.

00:52:11.955 --> 00:52:17.455
But, so take years, I mean, for simple, simple between quotes,

00:52:17.675 --> 00:52:20.675
but yet simple experiments like we have done.

00:52:20.835 --> 00:52:25.195
So we have to compress that time. So we, at the end of the day,

00:52:25.235 --> 00:52:30.175
the only thing that you cannot compress is the physical time of performing the experiments,

00:52:30.435 --> 00:52:34.715
right, because if you have to do experiments of one hour with some fish,

00:52:34.875 --> 00:52:37.515
you cannot compress that time. It's imposed.

00:52:37.735 --> 00:52:42.355
But then if you can analyze the data, produce some kind of model,

00:52:42.515 --> 00:52:47.135
produce some kind of controllers of the robot in the same lapse of time,

00:52:47.275 --> 00:52:49.215
you're saving a lot of time.

00:52:49.675 --> 00:52:54.555
And there are interesting theoretical questions out there. Can we automate model generation?

00:52:57.695 --> 00:53:07.355
It's again a holy grail of modeling. To what extent can we automate data analysis?

00:53:07.795 --> 00:53:13.035
It's again a holy grail in the system, the unsupervised machine learning that

00:53:13.035 --> 00:53:14.215
everybody's dreaming about.

00:53:14.515 --> 00:53:20.995
Because that's again a holy grail. So there are many fundamental questions behind that approach.

00:53:21.415 --> 00:53:24.775
Yeah, but then in your last set of experiments with the zebrafish,

00:53:25.075 --> 00:53:29.515
where also Also, the robots you insert have become much more sophisticated,

00:53:29.615 --> 00:53:33.395
and also, if you want the data analysis is more advanced and so on.

00:53:33.915 --> 00:53:39.815
However, you were not able to really systematically control or influence the

00:53:39.815 --> 00:53:42.775
aggregation behavior of these zebrafish.

00:53:44.327 --> 00:53:47.427
So what's the difference now going from the cockroach to the zebrafish?

00:53:47.527 --> 00:53:50.907
Why is it not working so easily? Why is the zebrafish so much more complicated

00:53:50.907 --> 00:53:53.747
to control? They have a completely different structure.

00:53:55.347 --> 00:53:59.407
They are completely different animals. Of course, they have vertebrates. They are not, etc.

00:53:59.827 --> 00:54:02.707
They live in a completely different environment. And they have a completely

00:54:02.707 --> 00:54:09.047
social and simple life, everyday life that's completely different.

00:54:09.187 --> 00:54:10.827
For instance, the cockroaches, they settle.

00:54:12.527 --> 00:54:16.307
After a while, they settle. They want to explore their environment.

00:54:16.547 --> 00:54:21.507
If they find a shelter, they settle there. The zebrafish are moving animals.

00:54:21.987 --> 00:54:24.847
They never stop moving during the experiment. They never settle.

00:54:25.027 --> 00:54:29.447
Okay, they stop from time to time, but they don't settle there after a while.

00:54:29.827 --> 00:54:36.707
We've been trying to give them some kind of shelter to see if they will select one.

00:54:36.787 --> 00:54:40.407
No, they don't do that. They oscillate between one to the other.

00:54:40.507 --> 00:54:44.427
So those animals are on the move all the time. But do they need that for their

00:54:44.427 --> 00:54:46.067
oxygen supply or for the gills?

00:54:46.287 --> 00:54:50.247
They need that because that's the way, exactly, that's the natural way of life

00:54:50.247 --> 00:54:52.947
in those ponds there in Nepal or India.

00:54:53.427 --> 00:54:58.907
They're always on the move looking for food, avoiding predators, for instance.

00:54:59.327 --> 00:55:03.927
And then they're in water, you have streams in water and stuff like that.

00:55:04.027 --> 00:55:05.647
So these are moving animals.

00:55:06.027 --> 00:55:11.307
They're all the time on the move. And so the question is, not only do you have

00:55:11.307 --> 00:55:14.827
to be accepted, you have to be part of the group, you have to be moving in the

00:55:14.827 --> 00:55:18.487
same time as the group, and then you have to try to influence,

00:55:18.767 --> 00:55:26.347
to modulate that on-the-move movement, perpetual movement during those experiments.

00:55:26.847 --> 00:55:29.307
So it's more challenging in fact. Right.

00:55:30.447 --> 00:55:38.107
So do you see this generalized to humans anyway? I'm not thinking too much about humans for the moment.

00:55:39.376 --> 00:55:44.356
Because I'm not well trained to understand the human behavior,

00:55:44.536 --> 00:55:50.936
but the methods, the scientific methods could be used in some cases.

00:55:51.116 --> 00:55:55.636
For instance, crowd movement, it's clear that many people are doing that kind

00:55:55.636 --> 00:56:00.676
of similar method and try to model crowd movement.

00:56:00.816 --> 00:56:05.696
Because crowd movement doesn't involve high cognitive capabilities.

00:56:05.696 --> 00:56:10.176
But I don't have a good knowledge about human behavior.

00:56:10.556 --> 00:56:14.956
And then again, I think this methodology can be also interesting,

00:56:15.116 --> 00:56:20.296
but for certain different kind of questions, let's say.

00:56:20.796 --> 00:56:27.476
But this is for you the outlook, so that your main thrust now is towards these

00:56:27.476 --> 00:56:32.456
more automated neuroethology experimental systems that autonomously do.

00:56:32.456 --> 00:56:39.056
I have two questions, let's say, if you want my roadmap of research until I'm retired.

00:56:39.496 --> 00:56:42.676
I have this set of methodological questions.

00:56:43.616 --> 00:56:51.416
How can we automate the process of data analysis and model generation and generating

00:56:51.416 --> 00:56:56.916
artificial systems that are capturing part of what we observe in the living system?

00:56:56.916 --> 00:57:03.516
I'm not the only one, it's a very broad question that is addressed by many people.

00:57:03.756 --> 00:57:07.416
And then there is another one which I find quite interesting,

00:57:07.516 --> 00:57:14.556
in fact, you know why, for reasons of sustainability again, living systems are

00:57:14.556 --> 00:57:16.596
going to be a very, very essential.

00:57:18.068 --> 00:57:23.528
They always, they've been always essential for humans, but we're back to think

00:57:23.528 --> 00:57:25.808
again about how essential they are.

00:57:26.128 --> 00:57:34.088
And in fact, all that research is about interacting and modulating the living

00:57:34.088 --> 00:57:35.788
system you're interacting with, right?

00:57:36.248 --> 00:57:39.988
And again, you find that at all level of living system.

00:57:40.088 --> 00:57:45.248
There are people who are interacting with cells and they want to modulate what the cells is doing.

00:57:45.868 --> 00:57:50.008
Like me, we want to modulate social behavior. You want to modulate metabolism.

00:57:50.128 --> 00:57:53.148
You want to modulate many things in living systems.

00:57:53.368 --> 00:57:57.868
And then you don't need, and that's one of the lessons of those models,

00:57:58.028 --> 00:57:59.808
Tony, if we go back to your question.

00:57:59.928 --> 00:58:05.028
You don't need necessarily to build a completely artificial living system to

00:58:05.028 --> 00:58:08.268
modulate the natural living system.

00:58:08.348 --> 00:58:13.128
You see what I mean? you can capture part of the living system,

00:58:13.328 --> 00:58:19.668
build an artificial one, a machine, let's say, that is going to interact with

00:58:19.668 --> 00:58:22.848
the living system and drive it to a state that you find interesting.

00:58:23.508 --> 00:58:27.668
It's very common. Breweries have been doing that for hundreds,

00:58:27.828 --> 00:58:29.208
thousands of years, I think.

00:58:29.508 --> 00:58:34.068
So the brewer, what is it doing? He's putting a living system in a tank and

00:58:34.068 --> 00:58:37.928
then it's driving this tank, controlling the temperature, the sugar level.

00:58:38.588 --> 00:58:44.588
To produce, to drive the yeast, to produce either wine or beer or something else.

00:58:44.888 --> 00:58:50.808
So that's a way to control a collective system, which is called yeast, right?

00:58:50.888 --> 00:58:53.548
And at the end of the day, you get the interesting product.

00:58:53.768 --> 00:58:57.748
In fact, we have to start to generalize that way of thinking,

00:58:57.888 --> 00:59:03.968
to have a system that drives a living system, But in an automated,

00:59:04.028 --> 00:59:05.968
in an autonomous way, in an automated,

00:59:06.088 --> 00:59:11.268
but also in an autonomous way, because of course, if you take brewery as one

00:59:11.268 --> 00:59:16.588
of the beautiful biotechnology that is centuries old,

00:59:16.748 --> 00:59:23.108
it's the brewer that is driving, keeping care of his tank fermentation and driving the system.

00:59:23.108 --> 00:59:31.048
Now, imagine you have a robot that is driving the tank and doing the beer. Can we generate a robot?

00:59:31.188 --> 00:59:35.728
Yeah, there's some automation for that kind of fermentation stuff.

00:59:36.308 --> 00:59:45.028
But can we find other systems where you want to drive living systems and you

00:59:45.028 --> 00:59:50.648
want it to be driven by another autonomous, an artificial autonomous system, a bio-rhebride system.

00:59:50.948 --> 00:59:55.148
And why do you want to do that? Because you want to explore the capabilities

00:59:55.148 --> 00:59:58.868
of the living system to produce interesting products for you.

00:59:59.428 --> 01:00:05.648
Beer and wine are extremely interesting products for us, but you can also dose-producing

01:00:05.648 --> 01:00:11.548
drugs or producing other kinds of materials that you want to extract from the living system.

01:00:12.469 --> 01:00:18.849
We're approaching a time in our history which is going to be a real step change

01:00:18.849 --> 01:00:26.929
when we're going to have robots, not just in factories, but in society and particularly in our streets.

01:00:27.029 --> 01:00:32.389
And I'm thinking perhaps most immediately we're going to have lots of vehicles

01:00:32.389 --> 01:00:39.069
on the roads which are actually robots interacting with other vehicles which are driven by people.

01:00:39.069 --> 01:00:45.329
So I'm wondering if these kinds of approaches that you're having is going to

01:00:45.329 --> 01:00:53.189
generalize up to, say, looking at the impact of driverless cars on how people will drive their cars.

01:00:53.309 --> 01:00:56.749
Because people are speculating, you know, once there are driverless cars on

01:00:56.749 --> 01:01:00.829
the road, other people will change their behavior because they will know that

01:01:00.829 --> 01:01:02.929
the driverless cars behave in predictable ways.

01:01:02.929 --> 01:01:06.869
But I think that there is going to need to be more of a science of this,

01:01:06.989 --> 01:01:09.329
because right now it is just speculation.

01:01:09.709 --> 01:01:14.349
Yeah, absolutely. I wouldn't dare by the methods I'm doing to apply immediately

01:01:14.349 --> 01:01:15.789
to bear, but that's a good question.

01:01:15.989 --> 01:01:21.349
When you introduce some autonomous system in contact with a living system, what's going to emerge?

01:01:21.629 --> 01:01:23.949
And we have no idea how to do that for the moment.

01:01:24.429 --> 01:01:31.669
We have still to work about that. How do you drive even a yeast cell? We have no idea.

01:01:32.229 --> 01:01:37.029
You say, how do you control, to what extent you can control its mechanism or

01:01:37.029 --> 01:01:38.009
its genetic regulation?

01:01:38.389 --> 01:01:43.249
It's not cleaner. But what your research shows is that we ought to be wary of

01:01:43.249 --> 01:01:46.809
the fact that there might be small perturbations you can make to the system,

01:01:46.869 --> 01:01:48.289
which could have very large-scale effects.

01:01:48.369 --> 01:01:51.269
That's selection from complexity and dynamical system.

01:01:51.669 --> 01:01:57.509
If you have nonlinear system, small change in the initial condition can lead

01:01:57.509 --> 01:02:01.109
to a big change in the system,

01:02:01.249 --> 01:02:06.949
the third lesson is a small change in the parameters of the system can lead to a bifurcation,

01:02:07.149 --> 01:02:11.409
which means a bifurcation means solutions that appear or disappear.

01:02:12.568 --> 01:02:20.088
And it's one of the major concerns if you think about ecosystem or the biosphere or the climate,

01:02:20.288 --> 01:02:24.848
it's that because those systems are most probably non-linear,

01:02:25.028 --> 01:02:29.628
if we start changing some of their parameters,

01:02:29.868 --> 01:02:34.388
like the quantity of carbon dioxide we're injecting in the atmosphere,

01:02:34.688 --> 01:02:40.828
we may have a strong non-linear effect and we cannot exclude that it can be a dramatic change,

01:02:41.648 --> 01:02:43.508
in the properties of the biosphere.

01:02:44.668 --> 01:02:51.228
We've seen collapse of some ecosystems that are robust to a certain extent,

01:02:51.328 --> 01:02:52.508
and then suddenly they collapse.

01:02:53.088 --> 01:02:54.848
So we've seen examples of that.

01:02:56.068 --> 01:03:01.048
And I'm sure that we don't have good ideas of how to deal with that or to control that.

01:03:01.288 --> 01:03:04.768
It's still a knowledge that is not well established.

01:03:04.988 --> 01:03:10.628
For instance, we have people that have been managing eutrophization of a lake, right?

01:03:10.828 --> 01:03:17.228
Suddenly you inject a fertilizer there you have a boom of algae and then you kill all biodiversity.

01:03:18.668 --> 01:03:21.468
There that is well known and then you have to manage how do

01:03:21.468 --> 01:03:27.048
I put back my lake in a state that I can have fish and biodiversity so this

01:03:27.048 --> 01:03:31.508
kind of general question I think has to be addressed also by the same kind of

01:03:31.508 --> 01:03:39.548
tools and again the non-linear dynamical system are good tools to address those questions but also that,

01:03:40.828 --> 01:03:44.008
type of model shows that the system will be intrinsically unpredictable.

01:03:44.908 --> 01:03:49.608
Certainly if this individual agents become more non-linear like in the case

01:03:49.608 --> 01:03:50.628
of having humans around.

01:03:51.348 --> 01:03:56.168
So if you talk about inserting, for instance, autonomous technologies in society, then.

01:03:57.588 --> 01:04:01.348
We should think very carefully about, indeed, what kind of reactions will trigger

01:04:01.348 --> 01:04:04.628
in the environment in which they have to operate, right?

01:04:04.708 --> 01:04:06.888
And these environments will, this will include humans.

01:04:07.268 --> 01:04:10.908
So now, so we look at this complex interaction between, in this case,

01:04:10.928 --> 01:04:13.008
artifacts, humans, and environments, right?

01:04:13.048 --> 01:04:16.368
And also in your model, you would say, well, the behavior we observe is indeed

01:04:16.368 --> 01:04:20.448
the function of, let's say, the controller of the agents, their morphology,

01:04:20.628 --> 01:04:24.008
as you showed us with the cockroach, and the environment itself.

01:04:24.148 --> 01:04:30.548
How many shelters are there, as an example, right? but juxtaposed to that,

01:04:30.628 --> 01:04:34.228
you're saying, but there's no hierarchy, right?

01:04:34.308 --> 01:04:38.528
Or you don't feel that the notion of hierarchy and hierarchical relations is

01:04:38.528 --> 01:04:40.988
helpful in trying to understand that system.

01:04:41.188 --> 01:04:47.948
So how should I be able to, how can I combine these two positions in a consistent framework?

01:04:48.648 --> 01:04:52.808
Yeah, but it depends what you mean by hierarchy. In the case of the cockroach,

01:04:52.828 --> 01:04:54.808
it's hierarchy at the social level.

01:04:54.808 --> 01:04:59.888
But then in terms of complexity, if you start to think that it's the system

01:04:59.888 --> 01:05:08.388
that is the system, you cannot cut part of it because the whole thing is a system.

01:05:08.488 --> 01:05:12.668
It's like people are doing now system science for Earth.

01:05:13.756 --> 01:05:17.856
With this controversial hypothesis of the Gaia hypothesis, but even if you get

01:05:17.856 --> 01:05:19.796
rid of that controversial hypothesis.

01:05:20.796 --> 01:05:26.316
Most people are now thinking about the Earth system as a whole.

01:05:26.576 --> 01:05:30.576
Because if you made a change in the climate, it's going to change biodiversity.

01:05:31.176 --> 01:05:34.876
But if you change biodiversity, you're changing also the local climate.

01:05:35.296 --> 01:05:38.696
So you see that you have to take into to call the whole system.

01:05:39.116 --> 01:05:44.396
And then what I'm claiming there is that, okay, the system is the system.

01:05:44.556 --> 01:05:49.036
There is no different level, but we have to work with that.

01:05:49.136 --> 01:05:54.776
So there is a methodology to work with that, is to select different level of description.

01:05:55.316 --> 01:05:59.716
But they have been chosen to answer a question.

01:06:00.176 --> 01:06:03.656
That doesn't mean that they have their own existence.

01:06:04.236 --> 01:06:07.316
They are not ontologically existent there.

01:06:07.576 --> 01:06:11.516
But it's a natural, sometimes it's a natural way to cut a system.

01:06:11.736 --> 01:06:16.356
For instance, if you're thinking about a solid, okay, you have the molecules, the atom.

01:06:16.496 --> 01:06:21.276
That's a very evident way, natural way to cut your system.

01:06:21.556 --> 01:06:28.116
But then if you have more complex way, more complex system than a crystal,

01:06:28.216 --> 01:06:31.516
a solid crystal, if you have the earth as a system,

01:06:31.656 --> 01:06:37.296
the way you're going to cut the system to answer a question is open.

01:06:37.576 --> 01:06:45.696
It's up to you to find the relevant elements in the system to build a model

01:06:45.696 --> 01:06:46.856
to answer your question.

01:06:47.096 --> 01:06:54.876
You're not stuck to a given, obviously given level of description because there

01:06:54.876 --> 01:06:57.416
would be the level of description.

01:06:57.816 --> 01:07:03.396
And it's one of the open questions in complex systems with emergence.

01:07:03.676 --> 01:07:06.496
You know, it's also a controversial term, emergence.

01:07:07.576 --> 01:07:09.116
Choosing,

01:07:10.147 --> 01:07:15.007
What you want to include in your description, your model, depends on the kind

01:07:15.007 --> 01:07:16.367
of question you're going to address.

01:07:16.547 --> 01:07:20.887
For instance, with the fish or with the cockroaches, if I just want to describe

01:07:20.887 --> 01:07:22.667
their social behavior, I don't

01:07:22.667 --> 01:07:29.107
need to go to discuss their sensory motor system or their neural system.

01:07:29.427 --> 01:07:36.307
I don't need that. But no, if I want to study the impact of the sensory motor

01:07:36.307 --> 01:07:41.307
system on social behavior, then of course I have to include that level also.

01:07:41.547 --> 01:07:45.627
But that's a different scientific question. I want to know the impact of the

01:07:45.627 --> 01:07:47.787
sensory motor system on social behavior.

01:07:48.607 --> 01:07:53.847
My other question was simply, I want to understand social behavior at the level

01:07:53.847 --> 01:07:58.827
of the population without taking into account other levels.

01:07:58.827 --> 01:08:06.687
If you want to now to study what is the metabolic impact on social behavior,

01:08:06.887 --> 01:08:10.087
because the state of the agent is changing depending on its metabolism,

01:08:10.467 --> 01:08:15.647
yet another question, then you start to include the metabolism.

01:08:16.047 --> 01:08:21.067
So we have to think as a system as a whole, but then as a methodology,

01:08:21.227 --> 01:08:25.847
we have to decide what are the relevant pieces that we have to take into account

01:08:25.847 --> 01:08:28.107
to have a good description,

01:08:28.647 --> 01:08:31.327
to get a good answer to the question we're addressing.

01:08:31.427 --> 01:08:34.827
And we have to be open-minded about the pieces we are including.

01:08:35.127 --> 01:08:42.067
So you had this talk about disease, mental disease and stuff like that.

01:08:42.527 --> 01:08:46.307
You have a whole system that is the human being, and then you can.

01:08:47.317 --> 01:08:51.397
Re-cut it in many different ways to take into account all the pieces that are

01:08:51.397 --> 01:08:55.037
relevant to answer the question, how do I cure that specific disease?

01:08:55.777 --> 01:08:59.897
One thing that's distinct about your approach is that, as you say,

01:08:59.957 --> 01:09:03.677
a lot of people will take an approach where they try, for instance,

01:09:03.777 --> 01:09:07.517
to understand social behavior in terms of sensory motor systems.

01:09:07.817 --> 01:09:12.377
So you're trying to describe what happens at the group level in terms of some

01:09:12.377 --> 01:09:13.857
process going on inside the individual.

01:09:14.497 --> 01:09:19.317
And most of the science you're describing, is saying, let's look at this level

01:09:19.317 --> 01:09:21.037
and see how it impacts on this other level.

01:09:21.157 --> 01:09:25.857
So how does metabolism affect social behavior? How do genes affect social behavior?

01:09:26.157 --> 01:09:29.837
But the dynamic systems approach doesn't seem to do that. It seems to say,

01:09:29.937 --> 01:09:33.377
let's try and capture behavior in terms of behavior.

01:09:33.577 --> 01:09:36.257
There is no crossing between two levels.

01:09:36.577 --> 01:09:39.437
Yeah, but that's, again, complex systems. Complex systems have,

01:09:39.537 --> 01:09:45.617
let's say, a level to simplify the discussion at which you can describe them.

01:09:45.617 --> 01:09:50.177
You can take into account the temperature of a gas.

01:09:50.257 --> 01:09:52.497
You don't have to take into account the kinetic energy.

01:09:53.477 --> 01:09:58.977
Simple temperature is good enough because you have this different way of measuring

01:09:58.977 --> 01:10:00.377
that or describing that.

01:10:00.477 --> 01:10:05.117
Sometimes you don't need all those details in terms of kinetic energy or vibration

01:10:05.117 --> 01:10:08.177
or quantum effects in the same way.

01:10:08.377 --> 01:10:13.157
The normal temperature that everybody knows with a thermometer could be good enough.

01:10:13.297 --> 01:10:17.317
But sometimes not. Sometimes you need to get into the other kind of details

01:10:17.317 --> 01:10:19.637
because you want a different answer.

01:10:19.717 --> 01:10:22.557
For instance, with the fish, what we're trying to do now is,

01:10:22.597 --> 01:10:27.157
okay, we have plenty of models of bird flocking or fish schooling and schooling, right?

01:10:27.497 --> 01:10:32.397
And the Vichek-like model that they've been done in physics in the 90s.

01:10:32.657 --> 01:10:37.917
But I say, okay, I have a different question. What kind of information do they

01:10:37.917 --> 01:10:42.377
take into account and how do they process that information, right?

01:10:42.437 --> 01:10:45.637
That's a different kind of question. Then I start, I have to,

01:10:45.637 --> 01:10:48.597
I have to start to open the black box of the individual.

01:10:49.317 --> 01:10:55.217
So I have to think about what is the vision in how vision works in fish,

01:10:55.317 --> 01:10:57.217
in that specific species of fish.

01:10:57.357 --> 01:11:01.097
And I have to kind of some kind of have a model of that, a minimal model.

01:11:01.257 --> 01:11:05.117
I don't have to take into account all the details of vision of the eye and stuff

01:11:05.117 --> 01:11:06.617
like that, a minimal model.

01:11:06.817 --> 01:11:12.197
And then what's the action we get in response to that perception that,

01:11:12.197 --> 01:11:13.497
and we did a model like that.

01:11:13.697 --> 01:11:19.417
You have the field of perception, the fish sees where are the individuals,

01:11:19.677 --> 01:11:24.477
and then it makes a probabilistic decision to go to certain direction.

01:11:24.797 --> 01:11:30.177
Then you have the other step. Okay. In that model, there is no processing of the information.

01:11:30.357 --> 01:11:35.517
There is this probabilistic description of the decision that the fish makes

01:11:35.517 --> 01:11:37.437
to go towards a certain direction.

01:11:37.657 --> 01:11:41.037
Now, if I want to understand how that information is processed.

01:11:42.483 --> 01:11:47.683
Right? Yet another question, yet another layer to add to the system.

01:11:47.903 --> 01:11:52.863
Yeah, but there's a challenge here, you know, because Poincaré already,

01:11:52.963 --> 01:11:58.343
you know, was talking about what concepts as scalpels in which you sort of carve reality.

01:11:58.783 --> 01:12:05.603
And Plato in his dialogue, Phaedrus, has this famous, this motto of that you

01:12:05.603 --> 01:12:09.323
carve nature by its joints. That means there's some intrinsic structure,

01:12:09.443 --> 01:12:12.523
and by accessing that structure, we can gain knowledge.

01:12:12.703 --> 01:12:18.583
But now, if we discuss hierarchy, right, across these many levels of organization,

01:12:19.223 --> 01:12:24.243
you seem to be saying, well, we should not commit ourselves too strongly about

01:12:24.243 --> 01:12:25.023
hierarchical relations.

01:12:26.603 --> 01:12:32.203
But does it imply that also intrinsically there is no hint of a hierarchical structuring?

01:12:32.203 --> 01:12:35.883
And don't you run the risk then of basically advocating, let's say,

01:12:35.983 --> 01:12:39.003
okay, there's an amorphous structure,

01:12:39.863 --> 01:12:44.483
with many elements that have no specific relation in a hierarchical sense and

01:12:44.483 --> 01:12:48.263
I can now arbitrarily group them in any way I like.

01:12:48.643 --> 01:12:52.063
Because, of course, that search space would be huge, right?

01:12:52.123 --> 01:12:56.903
So do you really believe that there is no intrinsic hierarchical organization in these systems?

01:12:57.363 --> 01:13:03.283
No, the patterns exist. I mean, living organisms are not amorphous stuff,

01:13:03.443 --> 01:13:05.103
a bunch of things interacting.

01:13:05.423 --> 01:13:11.563
The structure exists. The shape exists. The patterns exist in space and time and stuff like that.

01:13:11.763 --> 01:13:14.323
But the whole thing is a system.

01:13:15.639 --> 01:13:20.259
So some people say, okay, you take a thermostat, you cut the wire,

01:13:20.559 --> 01:13:24.359
you don't get the feedback, then it doesn't work. So you see, aha.

01:13:24.819 --> 01:13:28.259
Of course, because the thermostat is the system, you cannot cut the wire,

01:13:28.299 --> 01:13:30.139
otherwise you don't have a thermostat anymore.

01:13:30.499 --> 01:13:32.979
You have to consider everything together.

01:13:33.619 --> 01:13:37.859
But that doesn't mean that you don't have a wire, a temperature sensor,

01:13:38.039 --> 01:13:44.959
and some way to act on the radiator that you need. So, all those pieces do exist

01:13:44.959 --> 01:13:48.479
and they are displayed in a certain pattern.

01:13:48.659 --> 01:13:52.359
You have to put your temperature sensor at the right place.

01:13:52.559 --> 01:13:58.659
You need to act on the heating device in a certain specific way.

01:13:58.759 --> 01:14:05.259
So, the pattern exists, but you get regulation of the temperature,

01:14:05.419 --> 01:14:09.679
homeostasis of the temperature, if you have the whole system.

01:14:10.399 --> 01:14:14.459
And then if you want to understand how it works, okay, for a thermostat,

01:14:14.579 --> 01:14:18.659
it's pretty easy, but still you can get instability in thermostat in terms of

01:14:18.659 --> 01:14:21.279
negative feedback and you get oscillation and stuff like that.

01:14:21.279 --> 01:14:26.719
You see, then you start to analyze how it, and then you can recut your system

01:14:26.719 --> 01:14:32.299
any way you find it interesting to answer the question, how do I get a stable

01:14:32.299 --> 01:14:35.799
temperature, a chosen, given stable temperature?

01:14:36.019 --> 01:14:39.999
No oscillations in the temperature, no instability and stuff like that.

01:14:40.179 --> 01:14:42.759
So for living systems, it's still an open question.

01:14:43.099 --> 01:14:47.819
How do you shuffle the things? things, but you see that you always have an incomplete,

01:14:48.099 --> 01:14:52.699
again, talking this morning about this genome,

01:14:53.239 --> 01:14:59.039
proteome, neural nets, immune system, well, when you take into account,

01:14:59.199 --> 01:15:01.099
if you want a biomedical application,

01:15:01.519 --> 01:15:07.999
you have a disease, it can involve many, many things at the same time.

01:15:08.139 --> 01:15:09.079
So how do you choose that?

01:15:09.279 --> 01:15:14.319
Or you may decide, okay, it's a kind of cognitive dysfunction, so it's the brain.

01:15:15.162 --> 01:15:18.342
How do you know? It's a guess. I mean, it's an educated guess.

01:15:18.422 --> 01:15:19.402
It's probably involved.

01:15:19.802 --> 01:15:23.102
Okay, it's obvious. But how do you know it's only the neural nets?

01:15:23.202 --> 01:15:27.762
No, it's not Glyosense, it's not the other state, the physiological state.

01:15:27.882 --> 01:15:29.002
It's not a genetic defect.

01:15:29.242 --> 01:15:35.582
So you have many, many reasons to think that many other pieces can be involved.

01:15:35.982 --> 01:15:40.662
So some people are obsessed by the brain in itself and not even the brain in itself.

01:15:40.902 --> 01:15:45.422
You know, Paul, it's only a set of neurons in itself. Like there is only one

01:15:45.422 --> 01:15:49.022
type of cell in the cortical column.

01:15:49.202 --> 01:15:53.782
Some people are doing models where you have only one type of cell.

01:15:53.922 --> 01:15:57.602
And they pretend they are capturing some. Well, are they? Maybe.

01:15:58.062 --> 01:16:03.062
But you know there are all the different ways you can describe just a cortical

01:16:03.062 --> 01:16:06.062
column because there are many other things involved there.

01:16:06.382 --> 01:16:13.242
So taking the only way, the only methodology you may have is to have educated

01:16:13.242 --> 01:16:17.422
guess of what pieces to take into account because it's a system,

01:16:17.562 --> 01:16:22.662
it's a complicated system, not only a complex system, but it's a complicated system.

01:16:22.762 --> 01:16:27.942
And then you have to try to grasp the elements that are going to help you to

01:16:27.942 --> 01:16:30.562
have a model to answer a specific question.

01:16:31.897 --> 01:16:37.377
But do you believe, also as a physicist, that the math, the mathematics is all

01:16:37.377 --> 01:16:40.537
there, the framework is there, we just have not worked hard enough to apply

01:16:40.537 --> 01:16:43.297
them to this domain of biology and psychology?

01:16:44.277 --> 01:16:49.557
Or do you believe the math has just given us a starting point and most of the

01:16:49.557 --> 01:16:50.557
work still needs to be done?

01:16:51.277 --> 01:16:55.637
Well, that's again a big question. There are some people who think we don't

01:16:55.637 --> 01:17:00.437
have the math to describe a complex system because there is something difficult there.

01:17:00.437 --> 01:17:04.617
And some others are claiming, yeah, but we've done already a lot of,

01:17:04.637 --> 01:17:08.097
we have a lot of mathematical models to describe the system.

01:17:08.917 --> 01:17:11.777
I'm not sure. I have no specific answer to that.

01:17:12.057 --> 01:17:17.597
But clearly we are, if we look at the history, let's say again of physics only,

01:17:17.717 --> 01:17:20.937
because physics is clearly linked to mathematical methods.

01:17:21.257 --> 01:17:25.117
What physics has been doing since the beginning, if you place the beginning

01:17:25.117 --> 01:17:30.537
of physics with Galileo, Galilei, let's say, physicists and mathematicians have

01:17:30.537 --> 01:17:34.297
been working hand-in-hand to invent at the same time the physics models and

01:17:34.297 --> 01:17:35.317
the mathematical tools.

01:17:36.077 --> 01:17:40.517
So maybe we still need to invent new tools.

01:17:41.497 --> 01:17:48.637
And in physics, there is this famous problem of the N-core problems.

01:17:48.917 --> 01:17:53.737
When you have N bodies there, N entities,

01:17:53.917 --> 01:17:57.237
then it's becoming a mess because if you want to

01:17:57.237 --> 01:18:00.817
describe them that the so-called fundamental laws of physics as

01:18:00.817 --> 01:18:03.737
soon as you have three bodies there it's becoming

01:18:03.737 --> 01:18:07.097
a mess in terms of you cannot solve the equations and imagine

01:18:07.097 --> 01:18:12.437
you have avocado bodies there so so it's a it's a it's a big issue that's why

01:18:12.437 --> 01:18:16.757
statistical physics has been developed and those methods so now we have complex

01:18:16.757 --> 01:18:23.737
system with uh with a huge number of elements there so the question is.

01:18:24.867 --> 01:18:28.967
Will we need new models or new mathematical techniques? Why not?

01:18:29.147 --> 01:18:31.647
Why not? I'm not going to invent them myself.

01:18:32.187 --> 01:18:35.687
But the other thing is that you are in this tradition, you come out of this

01:18:35.687 --> 01:18:41.007
tradition that since the late 80s more or less was also advancing this link

01:18:41.007 --> 01:18:46.767
up of dynamical systems and life and also feels like artificial life and so on.

01:18:47.287 --> 01:18:51.627
So now we're almost 40 years later, I'd say 30 years later.

01:18:52.587 --> 01:18:57.867
How much progress did we then make on really understanding what life means so

01:18:57.867 --> 01:19:03.147
does this model you advance actually help us to understand and to define life.

01:19:04.707 --> 01:19:08.287
So little piece of it so we're still far,

01:19:09.047 --> 01:19:15.687
away from understanding well there was the famous quote a fine man on his blackboard

01:19:15.687 --> 01:19:21.267
which is you know it better than me what I don't understand I cannot construct No,

01:19:21.607 --> 01:19:26.067
I can only understand what I can construct, what I can build.

01:19:26.527 --> 01:19:31.347
So if you don't understand the system, you cannot build it, basically. Or the other way around.

01:19:32.507 --> 01:19:36.127
Actually, it goes back to Gian Battista Vico, the 18th century philosopher,

01:19:36.347 --> 01:19:38.587
who says that the truth and the fact are reversible.

01:19:39.347 --> 01:19:43.447
But then again, you think about the people in synthetic biology.

01:19:43.767 --> 01:19:45.867
There is no such thing as a synthetic cell.

01:19:47.347 --> 01:19:53.247
Purely synthetic, I mean, from scratch. So what people are doing for the moment

01:19:53.247 --> 01:19:58.007
in the field to try to understand how it works is taking piece apart and reassembling

01:19:58.007 --> 01:20:00.087
them or injecting pieces from there.

01:20:00.187 --> 01:20:03.687
You take the genome from one cell, you inject it in another cell,

01:20:03.807 --> 01:20:08.347
you take piece, you take a subset of a genetic regulation, you inject it.

01:20:08.447 --> 01:20:13.787
So we are tinkering still with what the cell should be.

01:20:13.987 --> 01:20:18.327
But we don't have a full understanding of just a cell, a bacteria,

01:20:18.547 --> 01:20:22.147
or a yeast cell. There are still things that we don't know.

01:20:23.427 --> 01:20:29.307
So now look, so we made the grand tour. Yeah, we made the grand tour of dynamical

01:20:29.307 --> 01:20:32.387
systems, behavior in our life.

01:20:33.467 --> 01:20:37.487
And you came also a long way in that whole adventure over the last decades,

01:20:37.847 --> 01:20:41.967
starting as a physicist, now trying to understand complex behavior.

01:20:42.547 --> 01:20:46.607
So if you would like to follow in that tradition, the José Alloy tradition,

01:20:46.967 --> 01:20:50.047
what would be José's law that we have to adhere to?

01:20:55.947 --> 01:21:01.047
It's more a way of thinking and a methodology than a law.

01:21:02.047 --> 01:21:07.507
I don't think for, and I'm skeptical that for complex, real complex systems,

01:21:07.747 --> 01:21:14.107
we will have the law or the simple set of equation that, like we have Maxwell's

01:21:14.107 --> 01:21:17.047
equation, Einstein equations, or Schrödinger equations.

01:21:17.167 --> 01:21:20.207
I think that doesn't work for real complex systems.

01:21:20.407 --> 01:21:22.747
It's a conviction. I cannot prove it.

01:21:23.570 --> 01:21:29.230
I have no theorem that proves it, but, um, when you're facing such system like

01:21:29.230 --> 01:21:32.810
living system or the earth system that includes living system,

01:21:32.990 --> 01:21:36.810
it's so complicated that we will not have a single set,

01:21:37.010 --> 01:21:44.190
simple, single set of laws, but we need a huge collection of good models capturing

01:21:44.190 --> 01:21:49.010
pieces of the system to, uh, interact with it.

01:21:49.190 --> 01:21:54.090
And so we have to be quite modest. That's why people are afraid of geoengineering.

01:21:54.410 --> 01:21:58.750
And there are people who want to interfere with climate by geoengineering.

01:21:58.870 --> 01:22:00.170
Well, that's a big question.

01:22:00.330 --> 01:22:05.590
You better think twice or even more than twice before starting to do that because

01:22:05.590 --> 01:22:07.990
we have no single idea what's going to happen.

01:22:08.170 --> 01:22:10.730
At the same time, because climate

01:22:10.730 --> 01:22:16.270
change is becoming such a big existence issue for the the humankind,

01:22:16.570 --> 01:22:22.150
well, you may start to think about how to interact with that complex system

01:22:22.150 --> 01:22:23.270
that is the earth system.

01:22:23.510 --> 01:22:28.130
Yeah, but look, we want to put your law, José's law, on a bumper sticker for Tony's car.

01:22:28.630 --> 01:22:30.990
So you cannot have like two pages.

01:22:32.630 --> 01:22:37.430
So what's José's law that we can fit on a bumper sticker that fits on Tony's car? Think again.

01:22:39.130 --> 01:22:44.330
Okay. Then three years from now, we're going to come visit you in Paris where

01:22:44.330 --> 01:22:45.310
you're going to take us out for dinner.

01:22:45.850 --> 01:22:50.190
And then we're going to interrogate you about a prediction you're going to make today.

01:22:50.350 --> 01:22:54.790
And the question will be three years from now, did you confirm this prediction?

01:22:54.970 --> 01:22:58.470
So what's the prediction you're going to make today that you're going to give

01:22:58.470 --> 01:23:01.130
us the answer, the confirmation to three years from now?

01:23:01.230 --> 01:23:05.570
What's the main hypothesis you would commit yourself to within your domain of research?

01:23:07.890 --> 01:23:12.170
It depends which domain you're talking about.

01:23:12.390 --> 01:23:16.150
The one about animal-robot interaction? The one we discussed today?

01:23:17.690 --> 01:23:25.250
I think my prediction is that we will not have made a lot of progress in the next three years.

01:23:26.910 --> 01:23:31.410
Well, José Aloy, thank you very much for this conversation.

01:23:36.630 --> 01:23:42.330
The CSN podcast was produced by the Convergent Science Network of Biometrics

01:23:42.330 --> 01:23:49.130
and Biohybrid Systems. a project funded by the European Sevens Research Framework Programme.

01:23:50.190 --> 01:23:55.590
For more interviews, recorded lectures or upcoming conferences in the field

01:23:55.590 --> 01:24:01.630
of biomimetics and biohybrid systems, go to csnnetwork.com.

01:24:01.840 --> 01:24:10.000
Music.