Good afternoon to all of you. It's for me a great pleasure to present Amoria Yatt, our

first speaker this year. So by the way, I wish you all the best for 2024. In this series

of lectures in which we try to bring different experts on the various areas and directions on

the interface between mathematics, artificial intelligence as one of the goals of the Center

of Mathematics of Data of the FAU. We thought that Amori Yatt could be a very appropriate

speaker. Amori is a present professor in Ecole de Pomp, the British school in Cernex, near Paris.

Yeah. It's actually exactly half between Paris and the Mandabert. Okay, yeah. So southeast of Paris.

And so Amori is now a professor in Ecole de Pomp. He was previously a student in Ecole Polytechnique

in the professors, Ecole Polytechnique of Paris, of France actually. And then soon after finishing

his studies in Polytechnique, he wrote a PhD thesis with Jean-Michel Coron in Sorbonne University,

in Paris. And he has developed several expertises. And this is why we thought he could be a good

speaker for this seminar series. One of them is, say, core applied mathematics hyperbolic systems

motivated by different kinds of applications like traffic and fluid mechanics. And then control,

stabilization, backstepping for these hyperbolic systems. That's, say, from the pure analysis

perspective. He was also engaged in some very interesting applications of this theory in

collaboration with some people here, like the team of Fluke and also Alis Kaima that are

collaborating also with Benedetto Piccoli in Radgers University and Pennsylvania and some other

groups on trying to understand to which extent these ideas can be really implemented in order to

regulate traffic. The controllers mean in those cases autonomous cars. And I remember seeing his

subunitation thesis recently. And I like very much this, say, first hint of really running

the experiments to see somehow evidences to be able to measure, take, really calibrate

the validity of the predictions done by the mathematical analysis of these models.

And I also like a lot this idea that, well, once you have done the experiment, you can do

a step further. And then rather than using control rules that are designed and computed

analytically by mathematical methods, just let them to be discovered to artificial intelligence.

I don't know whether you are doing it with ChatDBT or what. I mean, you can tell us later.

It would be like saying, OK, so rather than writing the exam, say, OK, so ChatDBT, please do it for me.

Right? Take the risk. And so these three avenues of work that he's emerging in his research,

I just found that particularly interesting. And I think this is more or less the spirit of his talk,

whose title is the role of artificial intelligence in the future of mathematics.

So we are particularly delighted that you accepted our invitation. And, well, we hope to see you soon

again. Hopefully next time there will not be any strike on trade. So you can do it right.

Otherwise, we have also the e-buy. That's a good alternative. Really good.

Thank you very much. Thank you very much, Enrique. It's a great pleasure to be in Alangun today.

I'm very happy to be in this amazing place. It's also a great privilege to be able to give the first

EMODY lecture of the year 2024. And when Professor Enrique Zueso invited me last November,

the two previous speakers were Lagrange-Brice-Lorraine and the Fields Medalists.

So this is either a very high bar or the beginning of a very long joke.

So I think the reason I'm here today is because I love mathematics.

I love its complexity. I love the kind of creativity you need in mathematics.

But I love also the fact that the more you study it, the more you start seeing things appearing

that you hadn't seen before. And this is the reason why when a few years ago we started

seeing jobs being replaced by machines and started discussing about what AI can do, can't do, will do.

I thought, you know, there's this one job that AI will never be able to do.

The job of doing mathematics, the job of mathematicians.

And today I'm going to try to convince you that I was actually entirely wrong.

So I'll start by saying that we are all using AI tools and we've been using AI tools for years.

The most easy examples are those ones, I mean, either through social networks, emails, apps, platforms.

Actually, even if you only have a smartphone, like a relatively modern smartphone, and you're taking pictures,

you might be using AI tools maybe without even knowing it.