- il y a 2 jours
The Future of Making, Moving, and Powering: How is AI Reshaping Industrial Operations?
Catégorie
🤖
TechnologieTranscription
00:02Good afternoon, and welcome to the session on the future of making, moving, empowering, how is AI reshaping industrial operations.
00:14So the AI and manufacturing market is projected to be worth over $6 billion by 2028, up from a mere
00:23billion dollars in 2023.
00:25And simultaneously, the energy sector is exploring AI-driven smart grids for optimized distribution, and the auto industry is racing
00:35towards AI-powered autonomous driving at scale.
00:40This panel of industry experts and our French minister will talk about the tangible realities of AI's impact on business.
00:55So let me start with you, minister, and ask, what is your view about the global AI race?
01:05This morning, the CEO of NVIDIA told this conference that we are in the middle of a new industrial revolution,
01:13one that is powered by AI factories.
01:17What's your view of the AI race, and what impact is this having on value chains?
01:24Well, first of all, thank you very much.
01:27And once again, congratulations to all the organization of this great VivaTech edition.
01:32Another record bit, I think.
01:35Very glad to be with you guys.
01:38From a trade minister perspective, first, I can say that the CEO of NVIDIA is perfectly right.
01:45It's more than a tech revolution.
01:47It's first an industrial and maybe more societal revolution that we are facing.
01:52So if you are looking at how countries, how continents are behaving facing this AI revolution, you can see first
02:02maybe the fact that value change are already challenged all the time, every day.
02:08And what is new, once again, from my perspective, when I'm traveling a lot, is that before geopolitical or let's
02:18say trade was leading and tech was following.
02:23I can say with AI quite the opposite.
02:25I think now about AI, tech is leading, and this is brand new because it changes everything in terms of
02:34priority, in terms of influence, in terms of investment, and in terms of ethic as well.
02:41So I would say, and it's not just a figure, you know, it's, I think it changes everything.
02:47When you are talking about reshaping global value chains, it can change everything because it's a question now of power,
02:56of influence.
02:57And now we are living in a world where trade is completely reshaped, mostly by United States today.
03:07We have, we have to think about how AI is reshaping itself, the global value chain within this new global
03:16context.
03:16So it's a kind of inception.
03:18It's a new reshape of global value chain within a global trade context, completely new.
03:24And this is what is maybe the most difficult to anticipate when you're a trade minister, is how to secure
03:31your own supply chain, for example.
03:35Who are your friends, if I may, in this new global context?
03:42With whom are you going to have some new partnership and new contracts?
03:46Of course, you can talk about chips.
03:48You can talk about models.
03:50You can talk about data centers and infrastructure.
03:52And we could talk a lot about what Europe and France has to offer to global investors, mainly in terms
04:00of models and in terms of infrastructure and data centers.
04:03But before, you know, competition and global value chain were about labor costs, were about capital flows.
04:13Now it's way more about access to compute.
04:16It's way more about chips.
04:19It's way more about models.
04:20It's way more about talents.
04:22And this is good news for Europe and for France.
04:24And we could also talk about attractiveness of Europe in terms of research and talent pool.
04:29So we have to take this situation as a great opportunity for Europe, if you allow me to say so,
04:40to have some advantages in this global situation.
04:43But to answer your first question, yes, we have a huge global change revolution, revolution, as in terms of ethics,
04:55in terms of industry, in terms of partnership, everything that we were used to is completely reshuffled.
05:04And this is quite exciting.
05:05Thank you, Minister, so clearly, AI is going to impact the competitiveness of countries and also companies, but only if
05:17we seize, you know, the potential of the technology.
05:22Now, if we look at studies from McKinsey, BCG, they all tell us that today, most companies are not reaping
05:34the full potential of AI.
05:36They're stuck in the pilot phase, or they're not focusing the AI in areas where they can reap the most
05:47for their businesses.
05:49How are Siemens and PwC avoiding this pitfall?
05:55Perhaps I start.
05:56I mean, the reality is that Siemens has been working on AI for 50 years, and it's one of those
06:01things.
06:01It's like DARPA net or the internet.
06:03For a long time, very little happens.
06:05And now with LLM, there's an acceleration, which basically means that for the next 50 years, on who's going to
06:12win and who's going to lose, and I fully agree this on the geopolitical, but also on a company size,
06:16is being defined with AI.
06:18Now, there's a mistake most big companies do is they appoint a chief AI officer.
06:23Worst thing you can do and say, I'm going to centrally sort of do it.
06:27You have to do it from the top and the bottom.
06:28So the first thing we've done is we basically set all of the data Siemens has, and 260,000 employees.
06:35We have lots and lots and lots of data.
06:37We're building a common data backbone.
06:39But on top of it, we have more than 400 AI initiatives to be able to see which one's going
06:44to work out.
06:45Because at the beginning of the internet, nobody knew where the sort of success is.
06:49So you have to build the capability to scale, and then you have to see which one of the elements
06:54are actually scaling.
06:55Is it in HR?
06:56Is it on IVRs?
06:58Is it on your own product development?
06:59And we're trying to do the same.
07:01So you have to lean in.
07:03You don't know what the future will look like.
07:05You have to have the flexibility, but you have to have the backbone to do this.
07:08And it meant that we have mandated that every information we have at Siemens is being shared, if there's not
07:15a good reason to do it, to be able to train those AI models.
07:18Because the key thing the minister was saying is you need capital, you need chips, you need data centers, but
07:23you need data, lots and lots of data.
07:26And making that available, I think, is what companies have to do, and then they have to put it to
07:31good use to be able to be differentiated.
07:33But I would love to see Mohamed on how you see it from your side.
07:41AI, speed matters, to your point, because we don't know what the answer is going to be.
07:47So it's very important to use AI, speed, and to try it out.
07:52It is not just a game of, you know, experimentation.
07:57It's a game of innovation.
07:58But when you think about innovation, you will actually fail.
08:01So the way we are thinking about it today is to redefine what is a return on investment in AI.
08:08Many companies are complaining about not seeing the productivity of AI.
08:13They're missing one point.
08:13They're missing the fact that even if they fail with a use case, they're learning something out of it.
08:20And the more they learn, so the return on investment in learning is very important, instead of just a return
08:27on investment on the financial return on investment.
08:29And many companies today are moving from the stage of using AI to gain efficiencies in the back office, whether
08:38it's HR, whether it's finance, to think about thinking about how do they use artificial intelligence basically to make money.
08:46artificial intelligence and how do you run factories, artificial intelligence, how do you embed it into your product and services.
08:54So there is an evolution.
08:55But the recommendation that we always had for ourselves and for any other company out there, speed matters.
09:02Innovation matters even most.
09:05So what sort of innovative products and services do you see emerging from AI in the industrial space?
09:14You know, in the industrial space, and that's what Cedric was saying, I will assume that whether you're thinking about
09:19any industrial product, whether it's cars, whether it's energy turbines, whether it's elevators,
09:27every single product would be AI-enabled, but as we just mentioned about the AI value chain, there is no
09:35such thing as using AI without microchips, without semiconductors.
09:41And then you need data and data centers.
09:44So that whole value chain will need to be improved.
09:48But the origins would believe that any product in the future and also any service that will be developed will
09:55have AI embedded into it.
09:56Now, Cedric is right.
09:59In 2001, when we had the mobile Internet, nobody ever talked about apps.
10:04Yes.
10:05Nobody knew about social media.
10:07What's coming next?
10:09New products and new services that will have AI embedded.
10:12The key thing is to try and to accept the fact that failing is okay as long as we learn
10:19about it.
10:23So, both of your companies are creating some new businesses around the AI boom.
10:31I think it's worth taking a minute to talk about that.
10:35So, in the case of Siemens, data centers has now become a booming business.
10:41Can you talk a little bit about that?
10:43So, it's interesting because you talked about the CEO, Jensen, and Jensen talks about AI factories.
10:50And I was asking, why do you talk about factories?
10:52And he says, it's a bit like factories in the past.
10:55You put energy in, you put chips in, and you have a product at the end which comes out, which
10:59is a model.
10:59So, the reality of actually what's happening is you have to see us not as data center but AI factories
11:06going forward.
11:07And so, what do we deploy?
11:09We build all the energy connection.
11:11The one thing I've never thought, look, I worked for Siemens now for eight years and I was in the
11:15energy part.
11:15I never thought that electricity would be scarce.
11:18That having nuclear reactors in France will be a big differentiator, right?
11:22Green or CO2 neutral energy.
11:25Most AI decisions are being done for these factories is, where do I get electricity from?
11:30So, we do the connection.
11:32We make sure that basically all of the data center, when there's a fire, don't burn.
11:36So, these are the things we do.
11:38But we also, most people don't know, there's only three companies in the world which have software which defines microchips.
11:44And we're one of them as Siemens.
11:45We bought Mentor Graphics.
11:46So, we help design microchips.
11:48We help sort of optimize them in the chip, in the rack, in the server, and in the data center.
11:56And we will optimize how do you do cooling.
11:58I mean, the racks of the future, you will have one rack consuming a megawatt, right?
12:04This is just insane of how much power will be used and needs to be cooled to be able to
12:09do this.
12:10So, as Siemens will look at all of it, how do I connect it, how do I keep it safe,
12:13how do I simulate the cooling and the warming, and also how do I build those in this direction.
12:18And it's an industry which I never thought would have these sort of acceleration.
12:23And the one thing I like is companies have to be fast, but also countries have to be fast.
12:28I'm half French, so I can be critical about France sometimes.
12:30But France is definitely moving on this AI thing, and I have to sort of give some compliment also on
12:35how fast this is going.
12:36So, Mohamed, I wanted to ask you about a new service that PwC is launching.
12:43And this is basically about assurance for companies to be able to audit their AI and make sure that it's
12:52doing what it's supposed to do and not doing things that it's not supposed to do.
12:56Can you talk about that as a business opportunity, but also, you know, talk about how this might help build
13:03business confidence?
13:05A couple of things.
13:06When you think about artificial intelligence, when companies decide to use AI, what does it mean?
13:11The people use AI.
13:13They have agents that they use within their organizations.
13:17They have AI in the products that they use.
13:20But also the change of business model, which means how do they run things on a day-to-day basis.
13:25All of these changes need to be audited to ensure that they don't create any collateral damage.
13:31A new service offering for us because for many of the companies that are using artificial intelligence to run the
13:36businesses, the control environment is going to change as far as who will make what decisions, who has access to
13:44what information.
13:45And it is very important for companies to understand.
13:48It's not just the wrong answers that can be provided by AI.
13:52It's the wrong decisions or activities that might be executed by AI also that we have to look at.
13:58So it's very tricky.
14:00We get a lot of efficiencies when we get to AI, a lot of productivity.
14:03But there will be some collateral damage.
14:05And as Cedric mentioned, there is no such thing as AI without data.
14:10That access to data is the most critical thing.
14:13What data will we give AI access to and what decision rights AI is going to be provided to ensure
14:18that we don't have any collateral damage.
14:20Every single company that is going to use AI in their business will have to go back and audit the
14:27fact that to solve one problem with AI,
14:30they have not created another bigger one, basically.
14:33That's what's happening today.
14:34Thank you.
14:35Now, keeping on the theme of data, data is so key, of course, to successful AI.
14:43Europe's companies, industrial companies, are sitting on a treasure trove of data.
14:48How do we make sure that these companies are able to keep the kimono closed, so not give away the
15:00data,
15:01data, but at the same time monetize it and progress and turn this into a real strategic advantage?
15:11So I'm going to ask also the minister also to help us.
15:13The one thing we need is, I'm a big European fan, but less regulation in this environment is something which
15:19we need on the AI side.
15:21Right?
15:21I mean, we have the AI Act.
15:23We have the Cyber Resilience Act.
15:24We have lots of acts at the moment.
15:26Omnibus is a good view, and I would love to have your view also on how you can help.
15:31But Jensen came, so the CEO of NVIDIA, and says, the one thing Europe has is that it has this
15:36immense industrial might.
15:38So in automation, so Siemens is one of the biggest automation player.
15:42So is Schneider.
15:43So is ABB.
15:44Industrial software, the two biggest industrial software companies in the world are European, right?
15:49It's Siemens and Dassault.
15:50So we're sitting on all of this industrial data.
15:53We need to make it available.
15:54So we need political help to be able to do this.
15:57And then we need to understand how do we make it available and how do we train models?
16:01And I'm going to give you a discussion I had for breakfast with the CEO of MISTA.
16:07And we talked about the next frontier will be to build models which are trained on AI.
16:13Models which are industrial.
16:15Why?
16:16Because the reality, most of our AI models we know at the moment are trained on the Internet, on all
16:21of the pictures, etc.
16:23But the value, I think it's 1.9 zettabytes of data, which is an industrial environment.
16:31Most AI models are now trained on it.
16:33And I'll give you an example.
16:34We are now capable.
16:35We trained with 2,000 data sets of technical drawings, an AI model.
16:40And we can now take a picture of something.
16:42We can say how it needs to be produced.
16:44We can optimize and already program the robots to do the CNC drilling.
16:50And all of this was an AI model which is driven on data.
16:54So my appeal is we have this huge treasure in Europe, which is industrial data.
17:00We need to make sure that we get the government support to train it.
17:05But at the moment, you would go to the U.S.
17:07or you would go to China to train the data.
17:08And we have the companies like Mistral or others to be able to train them on.
17:13And the outcome is so much more superior to the generic AI model.
17:17I think that's the future of which we need to focus on.
17:20Because imagine you could say to a modeling, I want to build a car which is greener, better battery, etc.
17:28Build it on all the last 30 years of designs I have, and it will give you three ideas.
17:33You would be faster to be able to drive it.
17:35And that's, I think, the big strengths of Europe.
17:37But perhaps, Mr. Minister, if you could sort of help us on some of that, that would be super, super
17:43important.
17:44So one of the keys to harnessing the data for AI is, of course, having the right AI talent.
17:53This is a challenge everywhere in the world.
17:56How is PAWC dealing with that issue?
18:00You know, when you think about AI talent, the talent has to come from schools, right?
18:07You only have a few countries that have university programs that are creating that talent.
18:13So the way we are looking at it, we are going different parts of the world for us to get
18:17access to that talent,
18:18to ensure that we can make sure that we use the talent from one country and leverage them globally, pretty
18:25much.
18:25Because we have to be able to share the talent.
18:27Because you just have to look at the education system to understand where do you have to find the talent
18:31when it comes to it.
18:33But I want to go back to what Cedric just said also.
18:36Because every time we think about AI, we think about AI talent.
18:41We also need industrial talent.
18:43We may need engineering talent for semiconductors.
18:47We may need engineering talent for batteries.
18:49The reason why every time we talk about artificial intelligence, it's not just one thing.
18:55It's not just a technology.
18:56It is the ecosystem.
18:58It's the data, the large language models, the semiconductors, the data centers, the real estate that you need to do
19:08this,
19:08the industrial equipment, energy, the energy equipment.
19:12So when we think about innovation, and we mentioned the value chain, like the minister just mentioned,
19:17it's not just a value chain of AI.
19:20The value chain of the whole ecosystem is evolving.
19:23And what we believe is that innovation is going to create it across all of the different segments.
19:30We will see innovation in renewable energy, in battery technology, in industrial equipment.
19:36So the type of people that we need are people for that whole AI ecosystem, not just people that understand,
19:44just artificial intelligence.
19:45You're absolutely right.
19:47And so I think we can talk some more about how do we marry the industrial talent with AI.
19:58But before we get to that, I want to ask you, Minister St. Mark, what can we, we're living in
20:04a very particular geopolitical situation right now.
20:09What, what could we, what could Europe and France do right now to capitalize on that situation and attract more
20:20AI talent?
20:22Well, first of all, I fully agree with what you both said about talents, because it's maybe the, at the
20:29end of the day,
20:30the main strength you can have about this race, because semiconductors are finite, compute can be scaled.
20:39But talents maybe are the best and most precious resource in this race.
20:46So attractiveness is about investment a lot.
20:51And by the way, it's very interesting to observe how from the beginning of the year,
20:55you, you can count on several and everywhere announcement of investment in AI.
21:03Of course, about what happened during the AI Action Summit in Paris in February with a record of 109 billion
21:13euros with the MGS project for Europe.
21:15I could get back on it because it's very, very important.
21:19Of course, the target project in the US, in China, et cetera.
21:22And within this kind of race of investment announcement, there is a question of the talent pool.
21:29Where is it going?
21:31And how can you, and that's your question, how can you attract more talents from everywhere?
21:37First of all, I think we have to be very proud in Europe to be honest about our universities
21:42and how we train AI talents today.
21:47And France alone has four universities out of the 101st in the world about AI, specifically about AI.
21:57And it's not a coincidence.
21:58You can see, of course, Mistral or other people, key people from France and from Europe in the top 10
22:07or top 20 startups and talents in this field.
22:12So after saying that, of course, you have to benefit from a global tension to attract more talents.
22:21This is what the president, Emmanuel Macron, did with the president of the commission, Ursula von der Leyen,
22:27with the true France for research and for science in Europe.
22:33And I'm saying Europe because it's not a French or German or Italian or Spanish or even UK matter.
22:44It is a European matter.
22:46We've done with Airbus a great example of how Europe can be of strength in terms of investment, in terms
22:55of industry, and in terms of talents.
22:58We can do the same with AI.
23:01What we didn't do at the very beginning of the century with Internet, it was because we were not between
23:08European people together and between European talents.
23:11So about attractiveness, we know the good recipe to attract talents.
23:16Of course, it's about money, but it's also about condition of research.
23:19It's about condition of...
23:21We have, to be very honest, a lot to do in France to make a lot of work to build
23:30bridges between universities, research and business.
23:33This is our main challenge.
23:36There is a lot of room for improvement to apply our own research into markets.
23:41And we have talents.
23:43We have to first keep them here in Europe.
23:46And I think it's key.
23:47And I want really to highlight all the speeches and all the willingness of great entrepreneurs, such as, of course,
23:58Mistral, but also other ones in France, in Germany and other countries that want AI to be in Europe.
24:05And they want Europe to be one of the key continents in terms of AI.
24:11So we need them to succeed.
24:12And we need to attract the other one.
24:14And, of course, in your question, you are mentioning what's happening in terms of tension between countries and continents and
24:23how talents can move from a continent to another.
24:27We know the recipes to do that.
24:28But most of all, most of all, we have to show them that Europe is a world continent with a
24:34common strategy in terms of AI.
24:37Otherwise, it won't be attractive.
24:39So it's not another answer.
24:42I think it's the main answer.
24:44Okay.
24:45Thank you.
24:46Please.
24:46Can I work on it?
24:47Because first, congratulations.
24:48I mean, one Europe is the only chance to be competitive.
24:51It can't be country-specific.
24:53But we have to be careful of one thing.
24:55Talent is super important.
24:56And the best talent is super important.
24:58But you should not lose the workforce.
25:00I think that we have a big, big political problem at the moment that a large part of the population
25:04feels that there's an elite and there's people with AI, which don't have access.
25:09And Mohamed said, we need the elite, but we need also to bring people on board.
25:12And there's this joke which we've been telling in industry for a while.
25:18And the joke goes as follows.
25:19How many people do you need in the future to run a factory?
25:25Two, you need a human and you need a dog.
25:28Why?
25:29Because you need the human to feed the dog and the dog to make sure that the human doesn't touch
25:33anything.
25:34That's the worst thing you can have.
25:36And the fear which we have at the moment is AI will take jobs away.
25:40Robots will take jobs away.
25:42The reality of what robots did is that the countries with the most robots was the most competitive in industry.
25:48And the country which will have the most AI needs to do this.
25:50But we need to bring on board all of the factory workers, etc.
25:56It's an anecdote, but it's very, very reassuring.
25:59We have an old factory out of the 80s.
26:02It's a World Economic Forum, and you probably also voted for it.
26:06Lighthouse for sustainability and digitalization.
26:09We have more than 100 AI models in there and 100 robots we put in.
26:15And the employees put them in themselves.
26:18And we gave them the chance to have an engineering co-pilot, which basically tells them what to do.
26:22The more junior the people were, the more they loved it.
26:25They felt more secure.
26:26They were faster, etc.
26:28And what we need to do is we need to attract the top talent.
26:31And we need to go to the broad population and say, it's going to save your job.
26:35It's going to save our way of life.
26:37And that's really sort of a role, which we have to do as companies.
26:40We have to do as countries.
26:41We have to do as people which help us to go in the right direction.
26:45And I think that's important.
26:46When we talk about talent, let's not talk about the top 1%, but let's talk about actually what makes a
26:52country.
26:52All great points.
26:54Go ahead.
26:54You know, one thing that I want to mention is like, again, you talk about what happened at the beginning
27:00of the Internet.
27:01A lot of people fear the Internet, new business models and all this.
27:05How many jobs were created since we have the last big technology disruption called the Internet, right?
27:13More than half of Silicon Valley didn't exist then.
27:17So we shouldn't fear the future.
27:19We should embrace it.
27:20Even if some jobs are going to be potentially replaced, the history of the world has shown us that every
27:27time we have technology disruption, it's net job creation.
27:31And people reinvent themselves.
27:34They get to solve actually bigger problems because the mundane problems can be used to solve technology so that humans
27:40can solve the big problems.
27:42And that is how we get to the new age of innovation because every time we have innovation, because somebody
27:48had an idea that they try out, now we will have the time to innovate even more because AI can
27:53take care of the mundane activities.
27:55But we shouldn't fear the future.
27:57So would you agree that, to Cedric's point earlier, that the challenge is really, you know, how to bring the
28:07entire workforce along with you, how to get them to use the technology, to learn it, embrace it, and even
28:15help co-develop it using their industry knowledge?
28:21My personal view, and that's what we have done at PwC, that you have to expose the employees, you have
28:27to expose everybody to AI so they understand what it is.
28:30Because honestly, people fear what they do not understand.
28:34I give you all an example, which is a little funny to me, because in 1994, I was an engineer
28:41at a company called Motorola.
28:42You don't know Motorola, right?
28:43Yes.
28:44I was in Chicago at the time.
28:46Motorola is the company that invented the cell phone.
28:49Yes.
28:50In 1994, as an engineer at Motorola, we received an email when Mosaic, who remembers Mosaic, the first antenna browser?
28:58Remember it?
29:00We also.
29:00We received an email from the CIO of the corporation saying, please do not use the internet.
29:07Right.
29:07The internet is dangerous.
29:09That was the same company that invented the cell phone nine years before.
29:14That's where we are today, where we actually have to show people that AI is something that they have to
29:21embrace.
29:22But for them to be able to embrace AI, we have to give them the tools for them to play
29:26with.
29:26So you mentioned earlier, Agenic AI.
29:31In Davos this year, the CEO of Salesforce said, this may be the last generation of CEOs who only manage
29:40a human workforce.
29:41So how do you see the collaboration between agents and human workers evolving?
29:52You know, I was actually in the room when Mark Benioff said that.
29:57And it might be true, but actually the caveat that we put out there is that it is true we
30:03will manage people and agents.
30:06You know why?
30:07It is actually already happening.
30:08It's already here.
30:10Now, where we are thinking about it, it's not just people to agent collaboration.
30:15It's also agent to agent collaboration.
30:17That's the next one here, right?
30:18You and I were talking about it in the back.
30:20Yeah.
30:20So the reality is we have learned on how to communicate with each other, different languages.
30:26We speak multiple languages and we know how to talk to each other.
30:28And it's real for us to speak French though.
30:29That's okay.
30:30We can speak French.
30:31But the key thing is, is how do agents talk to each other?
30:33Because what's going to happen is different agents will have to communicate and they will have to communicate to humans
30:38and to communicate to each other.
30:39Because the problems you're going to have are going to be linked.
30:42They're not going to be about my son is doing his essay on going to chat GPT or Mistral and
30:47putting it forward.
30:48It's I have a cat design.
30:50How do I get physical information?
30:51How do you get the production?
30:53How do you get sustainable?
30:54All of these agents will have to be linked one to each other.
30:57And we're working at the moment with different companies to be able to do that.
31:00Because you're correct.
31:01I don't know if humans will disappear, but there will be elements and agents which will be out there,
31:06which will solve problems for us proactively.
31:09And I can't wait to see what it is.
31:12I cannot predict how it looks like.
31:14But we definitely need to work that, A, they're ethical.
31:17That was one key point, which is important.
31:19But they also can talk to each other.
31:21So we don't have 20, 30 different sort of agent types, but there is one common language for all of
31:26them.
31:27So let me ask you, Cedric and Mohamed, to talk about what you would advise corporates who want to stay
31:39ahead of the curve on AI to do.
31:43Very simple.
31:44For me, speed matters.
31:46It matters actually more than anything else.
31:49Try.
31:50It's okay to fail.
31:50Now, the second thing it is to understand is the return on investment is not financial return on investment.
31:57It's when return on learning.
31:59And third, innovation is going to matter most.
32:02So using AI to do what we do today better, that will take you so far.
32:08It's to use AI to do some things really different in the future.
32:12And I will give a non-technical answer.
32:14When I talk to my team, I had all of my AI team together, and I say, three things you
32:19need to be curious.
32:20The world is moving so fast.
32:22Learn everything you can.
32:24Second one, be humble.
32:27You were at Motorola.
32:28I was at Siemens Telecom.
32:29My CTO put his arm around it and said, before those Americans put the voice over the internet, 100 years
32:35will go by.
32:36Be humble.
32:37This will move much faster.
32:39And the third one, be courageous.
32:42We will have to do courageous better.
32:43I loved what you said, is we will get some of it wrong.
32:46But if we're not courageous, we will be behind the curve.
32:50So my view on AI is be curious, be humble and courageous on this direction.
32:57Thank you.
32:59Minister St. Martin, you have the last word.
33:02What would you like the audience to take away from this session and from what your input?
33:09Well, AI is a race.
33:12You've said it.
33:12It's a race in terms of influence, in terms of control, in terms of resilience as well.
33:18And what I do believe is that geopolitics follow power, that power follow talent, and that talents go to places
33:32that invest in science, in freedom, in excellence.
33:38And this is the Europe we are building.
33:44And with that, I would like to thank our panelists.
33:50Please give them a nice round of applause.
33:52Thank you.
33:54Thank you.
33:57I think we go out.
33:59This way?
33:59Thank you.
34:00Thank you, James.
Commentaires