Uh, I'm curious if, if you think that that is far off, that an AI can experience some- something akin to what a human feels with pain, uh, when they, when they, for instance, are s- set out on a task and don't get, don't accomplish it the way that they've been instructed.

Um, I always think it's funny that many people claim, even computer scientists claim, that AIs cannot feel pain because our AIs have felt pain at least since 1990, you know? So what do we do when we build an, an agentic AI, an artificial neural network that, you know, produces actions that change the world, and then new inputs are coming in from the environment, and, uh, again, there's an action and so on, and the a- agent, the neural network remembers what happened before and is trying to figure out how to behave such that the sum of, [coughs] of, of, uh, all, um, um, reward signals is, um, maximized and the sum of all pain signals is min- minimized.

What about these pain signals? The pain signals are the most natural thing because whenever we have a robot, we give it pain sensors.

Why? Because i- it's a learning robot, and so this learning robot ne- needs some sort of motivation to learn to protect itself.

So whenever the robot bumps against an obstacle, then, you know, the corresponding pain sensors Wake up and negative numbers are special inputs to the, to the robot brain, and the robot sees then these, these incoming negative numbers, and it is wired to, to avoid that, uh, so it's trying to learn to generate action sequences that, um, avoid the pain, and it's trying to learn to generate action sequences that lead to the rewarding events.

Maybe the robot has a, a charging station somewhere, and whenever the battery's low, the negative numbers are coming from the battery, hunger, pain, hunger.

And then the robot's goal is to reach the charging station without bumping into, into obstacles, and sit down there and, and, you know, and enjoy the pleasure, just positive numbers, as the, uh, battery is being recharged.

... tell me how you come to that.

And I guess that the dream is to have, you know, a fully AI-driven loop of that discovery, right?

That's true, yeah.

20 years ago, I wrote this paper about the formal theory of fun and creativity, which is about exactly that.

So what should an artificial scientist or any scientist or any artist or any comedian do when its main needs, like eating three times a day, are satisfied and it has extra time to do stuff.

What do you do? Well, you listen to music, or maybe you compose your own music, or you generate art, or you are a scientist who not only is trying to solve problems given to you by other people, no, you also try to invent your own new problems.

Ask your own questions, not just answer questions given to you by somebody else, but invent your own new good questions.

That's what scientists do.

And I'm wondering, you know, these past years, has anything changed your mind at all? Or are you kind of still in the same camp?

Or, or we- was that also created at the, at the pretty much the same moment in your opinion?

It was all created at the same time, yes.

Because what you really can do is you can say instead of using an autoencoder to make sure that the, um, current input is represented in a way that is informative, you can also use a technique which I call predictability minimization.

That was also in 1991.

That was one year before the predictability maximization that we are now discussing.

And the predictability minimization works like that.

You have an input, um, bunch of hidden units, a, a bunch of, um, input units, um, and maybe what comes in is an image, and then you have, um, a couple of hidden layers, and then you have the, the bottleneck layer.

And do you think we will see, uh, like how the, the market will look like? Do you think like we will see like, you know, like one company that will take all? Even when it will reach to AI, do you think like there will be one company that like the first one to reach AI, like even if it will be like five years from now, uh, or 10 years from now or like 20 years from now, do you think like this company will, um, will take all and stay at the race?

But what happened? Why do you think RL has only become better now? What got lost in translations? and needed to rediscover to get us our current algorithms.

Yeah, so reinforcement learning is more complicated than supervised learning.

What you see today in the most popular AI models, large language models, how do they work? They basically apply supervised learning to...

imitate all the data on the web.

So predict parts of the data from other parts of the data.

Predict the next word and a sentence from the previous words.

Predict the next pixel in an image from all the previous pixels that you have seen.

Yeah.