- 15 hours ago
"Two teams of debaters. Considered jury members. One provocative motion. A lively exchange of ideas designed to challenge assumptions, surface competing perspectives and deepen understanding. The audience decides.
AI is reshaping the foundations of national power: economic competitiveness, defence, critical infrastructure, energy and the health of democracies themselves. AI sovereignty is the ability of a country or region to develop, control and govern AI without depending on foreign powers. It spans the full stack: the chips and computing power that train models, the models themselves, the talent and capital that build them, and the data, energy and rules that govern their use.
The United States leads on capital, scale and frontier models. China is building vast domestic capability and dominates the critical minerals that hardware depends on. Europe sits between them. It holds some strengths, including advanced chipmaking equipment, strong research and a distinctive, rights-based approach to AI regulation, yet it trails on investment, commercial scale and homegrown compute.
Can Europe close this gap, or has it now grown too wide for European AI sovereignty? Beneath this sits a harder question: is matching the United States and China even the right goal, or should Europe's task be to support a broader international ambition of building AI that is both capable of powerful outcomes but within the context of better governance? And crucially, could open-source AI offer Europe a credible path to compete without requiring the same capital and scale as its rivals?"
AI is reshaping the foundations of national power: economic competitiveness, defence, critical infrastructure, energy and the health of democracies themselves. AI sovereignty is the ability of a country or region to develop, control and govern AI without depending on foreign powers. It spans the full stack: the chips and computing power that train models, the models themselves, the talent and capital that build them, and the data, energy and rules that govern their use.
The United States leads on capital, scale and frontier models. China is building vast domestic capability and dominates the critical minerals that hardware depends on. Europe sits between them. It holds some strengths, including advanced chipmaking equipment, strong research and a distinctive, rights-based approach to AI regulation, yet it trails on investment, commercial scale and homegrown compute.
Can Europe close this gap, or has it now grown too wide for European AI sovereignty? Beneath this sits a harder question: is matching the United States and China even the right goal, or should Europe's task be to support a broader international ambition of building AI that is both capable of powerful outcomes but within the context of better governance? And crucially, could open-source AI offer Europe a credible path to compete without requiring the same capital and scale as its rivals?"
Category
🤖
TechTranscript
00:00:00Thank you. Thank you, thank you.
00:00:04Good afternoon and welcome to the New York Times debate curated by Debatable.
00:00:09I am Sophie Lombard. I'm the founder and CEO of Kite Insights and I'm also the curator of Debatable.
00:00:15This is the third debate that we're doing with the New York Times on the Vivatech stage
00:00:21and we are delighted to be back.
00:00:23And I would like to start by thanking Vivatech for once again creating a moment and a space
00:00:30to have this candid, if not robust, conversation on one of the most pressing issues of our time.
00:00:37Also, thank you for your willingness to jump headfirst in a format that is both innovative, albeit ancient, a debate.
00:00:50Today is a little experiment.
00:00:53We are looking to mix provocation with theatrics, with a bit of humor, but of course, a lot of substance,
00:01:01though you would be the judge of that.
00:01:04And the motion before us today is Europe does not have what it takes to achieve AI sovereignty.
00:01:13It's been a big topic today throughout the conference.
00:01:17And yes, this framing clearly suggests that we believe that the most important questions, problems of our time deserve a
00:01:25space for serious, constructive, productive disagreement.
00:01:30Because our goal is not to decide who is right, it's rather to explore and expose the strengths and weaknesses
00:01:40of the arguments on both sides of the argument and thereby looking to reveal the nuances, the trade-off, and
00:01:47hopefully enabling you, the audience, to leave this room with a bit more clarity.
00:01:53And with that, it's my great pleasure to welcome to the stage my co-moderator, New York Times journalist, Jason
00:02:01Carrion.
00:02:10Hello, thank you very much.
00:02:11Thank you very much.
00:02:12Hello, everybody.
00:02:13Welcome to the debate.
00:02:14I'm Jason Carrion.
00:02:15I'm a business editor at the New York Times.
00:02:17I'm based in London.
00:02:19And I'm delighted to be here.
00:02:21Thanks for having me.
00:02:22And it's going to be great.
00:02:24One of the sessions I was in earlier, the European Commissioner for Startups acknowledged, said, yes, Europe has challenges, but
00:02:31let's be positive.
00:02:32I don't know what else she would say.
00:02:35But on that note, I will just reiterate the motion.
00:02:38Europe doesn't have what it takes to achieve AI sovereignty.
00:02:42Not necessarily the most positive notion, but we'll see which side wins.
00:02:46So, Jason, if I'm not mistaken, you spent most of your business career working in the business world, understanding finance,
00:02:54capital markets, the inner working of companies, and companies that move markets.
00:03:00I can't think of a better person than yourself to explore where the power lies on this issue of AI.
00:03:07AI, it's really about exploring who has the capital, the technology, the political will to lead an AI.
00:03:15What is your take on this framing?
00:03:18Yeah, just for a very short sort of scene setting, you know, I'm a journalist.
00:03:23I'm synthesizing the things I hear, the people I speak to.
00:03:26And just over the past few days, there's been, just to underscore how urgent and important this debate is, you
00:03:35know, the end of the G7 today,
00:03:37and elsewhere in France, you know, ended with a discussion of AI, some of the biggest AI companies.
00:03:44The CEOs were there meeting with world leaders.
00:03:46It's kind of a coda to their meeting that was obviously covering a lot of other very weighty and heavy
00:03:51issues before Macron and Trump go to Versailles.
00:03:55And so that's just one thing that's happening.
00:03:59Looking through the VivaTech schedule today, just on the main stages, I counted 20 different sessions that had sovereignty in
00:04:06the title.
00:04:06So it's clearly in the air somehow and many, many more elsewhere out there in the, in the booths and
00:04:13whatnot.
00:04:13And a lot of that had this kind of gist that, you know, Europe is being squeezed between U.S.
00:04:20dominance in technology and money and China's strength in manufacturing.
00:04:25The European Commission says that the EU relies on foreign providers for over 80% of its digital products, services,
00:04:32infrastructure, and intellectual property.
00:04:35And that extends, I think, more or less to AI as well.
00:04:39What someone said in just a session I was, I was in before was that it's a hemispheric problem for
00:04:44Europe, which I think puts it in this very grandiose terms.
00:04:47But you can just see, like, there's been some extremely stern warnings from leaders around the continent about the risks
00:04:54of this reliance, of this lack of sovereignty, if you will.
00:04:59And a lot of that spawns, I think, from what's happened with Anthropic recently and how the U.S. was
00:05:04able to sort of pretty quickly throttle that for reasons that were not only or not even at all commercial.
00:05:11We need to do more reporting on that.
00:05:13So a few recent quotes that caught my eye, just, again, to set the scene a little bit.
00:05:17So Prime Minister Cornu here said yesterday that France must have its own tools.
00:05:23Very direct.
00:05:24We cannot rely on the goodwill of certain partners.
00:05:29Ursula von der Leyen said earlier this month, we cannot afford to depend on others.
00:05:33The German Chancellor today said, to put it in kind of the other side of the coin,
00:05:38these new technologies should be available to all countries.
00:05:42And this is him saying that he hoped to work with the U.S. on Anthropic.
00:05:46And then, if I may, I'm based in London to take a bit broader view of Europe to something from
00:05:51the U.K.
00:05:52The Minister for AI and Online Safety there said that the most significant question for Britain's national security is the
00:05:59future of AI.
00:06:00AI that we treat any other threat to our sovereignty with deadly seriousness, but we haven't learned to treat this
00:06:08one in the same way.
00:06:10That underscores, I think, the gravity of this debate.
00:06:13And then just a final word about definitions, because, you know, we're going to be arguing this notion of AI
00:06:19sovereignty.
00:06:20And I just want to underscore it really quickly.
00:06:22As we've laid it out, although I'm sure the definition will get a little squishier as the debate goes on,
00:06:28AI sovereignty is the ability of a country or region to develop, control, and govern AI without depending on foreign
00:06:32powers.
00:06:33It spans chips, computing power, models, talent, money, data, energy, et cetera, et cetera, et cetera, and the rules,
00:06:41which is a very strong specialty of Europe, that govern their use.
00:06:45And that is the heart of today's debate.
00:06:48Thank you, Jason, for setting the scene.
00:06:50And in essence, this conversation is really about power, but it's also about freedom.
00:06:56Will Europe shape AI or will AI be shaped for Europe by others?
00:07:01And as you said, is sovereignty simply catching up with what the U.S. and China is doing?
00:07:07Or is sovereignty, is Europe, does Europe have an opportunity to define a new conception of sovereignty that is different
00:07:17altogether?
00:07:17And as you say, the debaters will definitely play with that definition.
00:07:21Yes, let's leave it to them, and we'll bring them up very shortly.
00:07:24We did promise some fun.
00:07:26We did.
00:07:26And some interactivity.
00:07:28So now is your turn to set the scene with how the room is split at the moment, and you
00:07:35explain how that works.
00:07:36Indeed, indeed.
00:07:37In the grand tradition of ancient debate, think the Greek agora, the Roman forum, the great whole of medieval disputation,
00:07:45the audience was never a passive actor.
00:07:49They were asked, their job was to hiss, to clap, to cheer, even to stump their feet.
00:07:56Well, this is going to be your job today, and I count on you very much to make that work
00:08:01for us today.
00:08:02And remember, the energy you're going to inject into this debate is nearly as important as the words of the
00:08:08debaters.
00:08:09So please, don't hesitate to clap, to whoop, when you hear something that the debaters say that you happen to
00:08:17love.
00:08:17That's right.
00:08:18Okay, so you, the audience, in the end, and we'll explain more about that, will decide who wins.
00:08:25Again, there's lots of sessions out there about sovereignty today, but none of them will come to a conclusion and
00:08:31an actual answer.
00:08:33And so let's get the debaters up.
00:08:35Enough from us, actually.
00:08:36Shall we get the debaters up?
00:08:38Audience.
00:08:39Yeah.
00:08:39Yes?
00:08:40Okay.
00:08:43So, debaters, be ready.
00:08:45Let's bring you up.
00:08:47On the foresight, I'm going to...
00:08:49Good.
00:08:50Welcome.
00:08:51Good.
00:08:51Be ready.
00:08:52Dakota Carey, Leticia Beyanoun, Massa Alimardani, and Thomas Lingard.
00:08:58Please join me in welcoming them to the session.
00:09:01The foresight.
00:09:03Foresight.
00:09:12Great.
00:09:13And it is my pleasure to introduce the against side.
00:09:17Again, the motion doesn't have what it takes to achieve AI sovereignty.
00:09:20It's kind of the pro-European side.
00:09:22It'll become clearer.
00:09:23It's a bit of a mind twist.
00:09:24You can use it.
00:09:25And that is Nayana Prakash, Christopher Porter, Eric Kazan, and Clara Shappaz.
00:09:30Please come up.
00:09:30The against side.
00:09:42All right.
00:09:43Welcome.
00:09:47So, Jason and I, without losing any of our neutrality, right, Jason?
00:09:52Journalist.
00:09:53Strictly neutral.
00:09:54We will temporarily hold allegiance to one of the team for the sake of the debate.
00:09:58So, I'll be standing with the foresight.
00:10:01Yes, and I'm representing the against.
00:10:04And so, this is how it's going to work.
00:10:06After our debaters speak in pairs, you will decide the winner.
00:10:10Based on the arguments that you hear today and which side you find most compelling in support of their position.
00:10:18And so, let's see how we're split up again.
00:10:21I mentioned that earlier.
00:10:22Using this incredibly scientific, carbon neutral, human centered, sustainably sourced measure of support.
00:10:30And that is clapping.
00:10:32Let's do a practice run.
00:10:34Shall we do a practice run?
00:10:35So, people, this is your one and only time where you can actually tell us what you think.
00:10:41Because, as Jason said, the next vote would be for the team that put the best debate.
00:10:46So, I'm going to ask you to tell us what you think and tell us where your allegiance actually lie.
00:10:52So, if you think that Europe does not have what it takes to achieve sovereignty, please clap and cheer now.
00:11:04We're not looking who's clapping.
00:11:07Okay.
00:11:08And if you disagree, if you think Europe does have what it takes to achieve AI sovereignty, clap now.
00:11:18Uh-oh.
00:11:20Partisan.
00:11:21Woof.
00:11:22Okay.
00:11:23Okay.
00:11:24I think we're warmed up and I think the gauntlet has been thrown.
00:11:28Yeah, yeah.
00:11:28But don't worry.
00:11:30Our team have been switching rooms around many times over.
00:11:34Anyway, again, I want to reiterate that this debate motion is not meant as a true statement.
00:11:40It is a provocation.
00:11:42And our debate are not necessarily, our debaters, sorry, are not necessarily arguing their personal conviction,
00:11:50even though we've asked them to lean into the theatrics of the moment.
00:11:53So, it's good that you know that.
00:11:56And they have, in many ways, bravely agreed to defend a specific position for the sake of the debate.
00:12:02And in doing so, they might very well go beyond their own conviction.
00:12:06And as you can imagine, that takes a mix of courage, intellectual agility, creativity, empathy, you name it.
00:12:14You wouldn't want to be here now.
00:12:15They are here, have done the work.
00:12:17They're ready to take the challenge.
00:12:20So, remember, support the debaters on both sides throughout the process.
00:12:26And please, if you feel like capturing some of those magic moments on your mobile phone for social media,
00:12:33don't post that out of context.
00:12:35We don't want anyone to get in trouble for whatever they end up saying on this stage.
00:12:39Yes.
00:12:40And one other little twist here.
00:12:42After the debaters have spoken, we will turn our attention to a jury who are sat in the front row
00:12:47here.
00:12:47We'll bring them up later.
00:12:48Later, Martin Houlin, Louise Agersnap, and Bruna Poisson.
00:12:56We'll hear from them later, and we can give them more applause then.
00:13:01They will not decide who wins, but they will present challenges to the debaters after they've heard each of their
00:13:07arguments.
00:13:08They'll shine a light on what they liked, what they didn't like, what they feel debaters might need to flesh
00:13:12out a bit more, etc., etc.
00:13:14And so debaters will need to take those comments on board, and then there will be a very short break
00:13:20where they confer amongst themselves to come up with their final rebuttal, their final retort, closing argument, if you will.
00:13:29And then they will come up and do that, and then we will have another, again, very scientific poll of
00:13:36the room with great granularity and accuracy, and we will see whether you have been swayed one way or the
00:13:42other, and whether it stayed that way or changed, and it will be very exciting.
00:13:46Great. I think we should start.
00:13:48Are we ready to start?
00:13:51Okay.
00:13:52I want a bit more energy in the room.
00:13:55All right.
00:13:57All right, all right.
00:13:58So, we're going to have the first two debaters up.
00:14:02On the fourth side, I would like to welcome Dakota Carey.
00:14:06Yep.
00:14:09And on the against side, we have Nayana Prakash.
00:14:16So, we asked Dakota, whose credential is on the screen, to tell us a fun fact about himself,
00:14:23and he wanted us to let you know that he's fluent in Mandarin.
00:14:29Not French, I'm really sorry about that.
00:14:31Well, Dakota, we see straight through you, because this is a debate about European AI sovereignty, and you're playing the
00:14:37China card.
00:14:39Right.
00:14:39I'm going to have a lot of those.
00:14:41And for Nayana, the fact about her is, before moving to the UK, she grew up in India, Bahrain, Saudi
00:14:47Arabia, and the UAE, so she knows that no nation, you know, exists in isolation.
00:14:52And so, she's coming from that point.
00:14:54Great.
00:14:54So, debaters, you have the clock here.
00:14:56Three minutes.
00:14:56We're going to start with you, Dakota.
00:14:58You hold this as if it was a lollipop, but you don't lick it.
00:15:02Excellent.
00:15:02And when you're ready, you go.
00:15:03Good luck.
00:15:04Excellent.
00:15:05You know, it took me so long to get to the peripheral states of the American empire, and I'm glad
00:15:11to see all of you here.
00:15:12I'd like you to raise your hand if you can access Fable at the moment.
00:15:18No hands.
00:15:19Excellent.
00:15:19So, no access to current frontier models.
00:15:22And I wanted to start here because the size of the AI that you will make sovereign or will say
00:15:29is sovereign matters a great deal.
00:15:32And the resources that the EU can marshal to make an AI of the size that they're interested in are
00:15:38exactly the reasons that Europe cannot achieve sovereign AI.
00:15:44If you would like to acquire the talent that is capable of producing leading-edge models, you have to be
00:15:51able to lay out millions of dollars in either equity that's currently good,
00:15:56so you have access to Meta, Google, any number of large tech companies where you can set aside cash.
00:16:02You have access to cash flow from investors.
00:16:06Europe does not have a robust capital market for VC.
00:16:09And, in fact, because of this, the people who can produce those leading AI models end up in countries like
00:16:17the United States,
00:16:18where we do not cap their effective income.
00:16:21We allow them to become outrageously rich, and as a result, we're very effective at acquiring talent from abroad,
00:16:28including from Europe and other parts of the world.
00:16:33At the end of the day, we are fundamentally in competition over about 2,000 competent AI researchers,
00:16:43in addition to all of the hardware and energy components related to operating AI models and making them accessible to
00:16:52the people who are going to use them.
00:16:54So I would like to ask everyone in this room, if AI sovereignty is fundamentally about the control over the
00:17:01design,
00:17:02the implementation of AI systems, Europe can do that.
00:17:06But I would ask you to consider the capabilities that Europe is actually able to achieve under this condition of
00:17:14AI sovereignty.
00:17:17That bar, I think, is unfortunately quite low.
00:17:21So, I would ask you to consider, as the last part, if you're only choosing which components of foreign technology
00:17:30to integrate,
00:17:31which components of foreign AI stacks to allow within your borders, how they operate, etc.,
00:17:38you are fundamentally choosing an AI tech stack in the same way you have chosen an IT tech stack,
00:17:45and all of you are sending emails across American technology right now,
00:17:49your payments are transacted on Visa and MasterCard, payment systems, AI is going to be exactly the same.
00:17:58Well done.
00:18:00Okay.
00:18:02Liana, three minutes. Off you go.
00:18:05Thank you very much.
00:18:06So I'm just going to start by saying, Europe has what it takes for AI sovereignty,
00:18:10especially if we stop defining sovereignty incorrectly.
00:18:14No country in the world can achieve true AI sovereignty throughout the stack,
00:18:19not even the US and not even China,
00:18:21but I think it's a European principle that we're better together.
00:18:26Europe is ready to redefine what sovereignty means,
00:18:29to reduce its dependence on these tech superpowers,
00:18:32and to remake sovereign technology in the European image,
00:18:35by which I mean to have sufficient control, leverage, and resilience to pursue its economic and political objectives.
00:18:43We aren't talking about owning every layer in the stack,
00:18:47but rather the potential and the current advantages that Europe has.
00:18:51Deep talent pools, world-class research institutions, advanced manufacturing,
00:18:57major cloud infrastructure providers, major AI firms,
00:19:00and crucially, the world's most important company in ASML for lithography.
00:19:06Europe also has a growing wave of self-belief and ardent backers from across the continent,
00:19:11and I'm sure I'm looking at some of these people in the room right now.
00:19:14If the UK partners with Europe, I'm biased because obviously I am British,
00:19:19but I very much hope it does,
00:19:21then this talent and research pool becomes even stronger.
00:19:24We can't achieve sovereignty overnight,
00:19:26but we have what it takes to get there in the medium term.
00:19:29If sovereignty is about reducing dependence, not completely eliminating it,
00:19:33which isn't possible for any country,
00:19:36then Europe has been preparing for many years to build the Eurostack.
00:19:39This isn't a new idea that we're debating here today.
00:19:42What it needs now is the final push.
00:19:44We're obviously having this debate because of the sheer urgency
00:19:47in reducing dependency on the US,
00:19:49and the very real threat of choke points of tariffs,
00:19:53of supply chain shocks in AI development.
00:19:55Europe has already shown it can absorb short-term costs
00:19:58to reduce strategic dependence.
00:20:00After Russia's invasion of Ukraine,
00:20:02many said ending dependence on Russian energy was impossible,
00:20:05but Europe did it anyway.
00:20:06It can be done.
00:20:08And some of America's failures,
00:20:10like the attacks on US universities,
00:20:12the attacks on research funding,
00:20:14the kind of inequalities with visa holders,
00:20:17can be Europe's gains.
00:20:18If it opens its doors to talent leaving the US,
00:20:22Europe can learn from America's mistakes.
00:20:25Critics of European sovereignty,
00:20:26which maybe some of my competitors here might be,
00:20:30often reduce sovereignty to building the next OpenAI,
00:20:33or Google, or Microsoft,
00:20:36or maybe even Fable or Mythos.
00:20:38They're probably a little bit less after the weekend.
00:20:40But Europe isn't trying to follow the American model
00:20:42of technology development,
00:20:43or create the next Mark Zuckerberg.
00:20:46This is not what sovereignty is about.
00:20:48It's about Europe building technology
00:20:50for its own governments,
00:20:52for public institutions, and for people.
00:20:54European sovereignty will carve its own distinctive path
00:20:57without seeking billion-dollar valuations
00:20:59or world domination.
00:21:01Europe is ready,
00:21:02even if the US doesn't like to hear it.
00:21:08Well done.
00:21:09Great.
00:21:10Excellent first round.
00:21:11Please take your seats back.
00:21:13You can leave the mic.
00:21:15Pretty good first round.
00:21:17Thank you for that.
00:21:18I think Dakota, in a sense,
00:21:19is saying that, you know,
00:21:21Europe is far too reliant on American tech,
00:21:24and even through regulation,
00:21:25it can't, you know,
00:21:27build its way into relevance.
00:21:30Pretty damning.
00:21:31Well, on this side, listen,
00:21:33just have some self-belief
00:21:34throughout the entire stack.
00:21:36There's nobody can be sovereign across all of that.
00:21:39It's about reducing dependence,
00:21:41not doing necessarily everything within your walls.
00:21:44And quite powerfully,
00:21:45Europe can learn from the US's mistakes.
00:21:48Great.
00:21:49Love that line.
00:21:50Next round.
00:21:51On the fourth side,
00:21:52we have Leticia Benayoun.
00:21:57Yes.
00:21:59And on the against side,
00:22:01Christopher Porter.
00:22:07Great.
00:22:08So, as you can see from Leticia's credentials,
00:22:10she's a pretty formidable woman.
00:22:12And while studying her master's,
00:22:15Leticia suddenly often had to sneak out of lecture
00:22:18to set up her company, Dealstone.
00:22:20So she has to slip out.
00:22:22Don't take it personally.
00:22:23She is a footnote on her path to world domination.
00:22:27We are a footnote in her path to world domination.
00:22:30Sorry.
00:22:31And a little bit about Christopher.
00:22:33He has six children, five boys,
00:22:35and just welcome the sixth is a girl.
00:22:37So he's building his own sovereign state as we speak.
00:22:41All right.
00:22:42Leticia, are you ready?
00:22:43The clock is here.
00:22:44Keep the mic close to your mouth.
00:22:46You go.
00:22:47Okay.
00:22:47First and foremost,
00:22:48sorry for my French accent.
00:22:49And let's go.
00:22:50So I would like to give you four main problems I step up every single day
00:22:55as someone who just graduated in AI in 2026
00:22:58and who's someone who built an AI native company.
00:23:00So my first problem is in Europe,
00:23:03we don't really have frontier model to build on.
00:23:06And this is something I really encounter
00:23:10because when I decided to draw my AI tech stack,
00:23:13when I built my AI tech stack,
00:23:15I couldn't find one European model better than the U.S. one.
00:23:18And this is a big issue
00:23:20because I rely on a foreign company
00:23:24and foreign model for my AI tech stack.
00:23:26And just last Friday,
00:23:28the U.S. government cut me off
00:23:31from the two best models in the world.
00:23:32So if it's not a sovereignty issue,
00:23:35I don't know what it is.
00:23:37The problem number two
00:23:39is that in Europe,
00:23:40it's very difficult to scale.
00:23:42Actually, there is a real cruel irony
00:23:46because we have a big market than the U.S.
00:23:49But it's only on the paper.
00:23:51In reality, we have 27 countries
00:23:54with different cultures,
00:23:56different languages,
00:23:57and different work habits.
00:23:58So it's very hard to scale.
00:24:00And AI needs to scale to get better.
00:24:03And the conclusion of that
00:24:05is that 64 European unicorns
00:24:08have relocated into the U.S.
00:24:11Problem number three,
00:24:13it's very difficult to keep our talent in our home.
00:24:16And I can tell you
00:24:17because I just graduated last week,
00:24:20last Friday,
00:24:21and most of my classmates...
00:24:26And most of my classmates
00:24:28already decided to go to San Francisco
00:24:30to launch their AI startup
00:24:32or even join an AI startup there.
00:24:35And this is very sad
00:24:36because we have good master,
00:24:38we have good training,
00:24:40but we can't keep our own talent in our home.
00:24:44Problem number four.
00:24:47So it's also a big problem
00:24:49because it's very difficult
00:24:51to raise money in Europe
00:24:52and we don't have enough capital.
00:24:55And AI also needs a lot of capital
00:24:58to get better.
00:24:59And this is something I can tell you
00:25:01because my main competitor in the U.S.
00:25:03just raised in Precid
00:25:0550 million of dollars.
00:25:06And you can imagine
00:25:08that it's very difficult in Precid
00:25:09to raise such amount of money.
00:25:13In the U.S.,
00:25:13you can raise four times more
00:25:15than in the Europe.
00:25:16And this is a very big issue.
00:25:18And even on the exit part,
00:25:21we lack of capital
00:25:22because when it comes to exit,
00:25:26in the majority,
00:25:28the acquirer is even Chinese
00:25:30or even American ones.
00:25:33And we can't keep our own company in Europe
00:25:36because we can't acquire them.
00:25:38So it's also a big issue.
00:25:43Well done.
00:25:47I'm keeping excellent time.
00:25:49I congratulate all of the speakers so far.
00:25:51Christopher, the floor is yours.
00:25:52Three minutes.
00:25:53No pressure.
00:25:54We've heard a lot of arguments so far
00:25:56about why it might be difficult
00:25:57for Europe to develop its own AI,
00:25:59but we haven't really heard any arguments
00:26:01for why you can't.
00:26:03And I think it's important to understand
00:26:05that on many of the important measures,
00:26:08Europe is already an important player.
00:26:10It's a counterfactual.
00:26:11The U.S. can no more develop
00:26:13its own AI systems without ASML
00:26:15or, for that matter,
00:26:17memory from SK Hynix in Korea.
00:26:20In the case of Google,
00:26:21our frontier AI lab is in the U.K.
00:26:25So the same concerns you have about,
00:26:27oh, we don't have access to this capital
00:26:29or that intellectual, whatever,
00:26:30those are the same concerns
00:26:32the U.S. government has
00:26:33about its own companies.
00:26:35So I just want to say, first off,
00:26:37this is a little bit of a counterfactual
00:26:39to say that you can't do it
00:26:40just because it would be difficult.
00:26:42The concerns we've heard raised about Fable,
00:26:46first off, there's nothing special about Fable.
00:26:49Every AI model that is trained
00:26:50at a trillion parameters or more
00:26:52will do the same thing,
00:26:53including the sovereign AI
00:26:54developed in France by Mistral,
00:26:57including the Chinese open-weight models.
00:26:59None of these things are prohibitive.
00:27:02In France, I think,
00:27:03it's just an especially illustrative example.
00:27:06The physical infrastructure
00:27:08that undergirds AI, cloud,
00:27:10France has decided
00:27:11that it was important to them
00:27:13that that be sovereign,
00:27:14that someone like me,
00:27:15who used to work in U.S. intelligence,
00:27:17who's in America,
00:27:19not be able to control or access
00:27:20or deny access to that infrastructure.
00:27:23So it's run by Thales in France.
00:27:25They use our technology,
00:27:27open-source technology.
00:27:28We provide engineering input.
00:27:30But to the degree that it's important,
00:27:32Europe already chooses
00:27:33to be sovereign in these areas
00:27:35or there's interdependence.
00:27:37We can't develop our models
00:27:39without Europeans
00:27:40contributing scientifically.
00:27:41So I think the important question
00:27:43isn't if you can,
00:27:45you've seen within the last week
00:27:46why you have to,
00:27:47but what does sovereignty mean to you?
00:27:50And I think the important part
00:27:51is twofold.
00:27:52One, can you ensure
00:27:54that you'll always have access,
00:27:55that no one anywhere else
00:27:57in the world can shut it down?
00:27:58That physical infrastructure
00:28:00is already available in France
00:28:01in an undeniable way
00:28:02that nobody else can shut down
00:28:04because it's run by French companies,
00:28:06manned by Europeans,
00:28:07but it still benefits
00:28:09from a global research funding
00:28:11that comes, yes, it's U.S. centric,
00:28:14but it's ultimately scientific research
00:28:15that benefits everyone.
00:28:17The second part,
00:28:18I think it's very important
00:28:19to understand why,
00:28:20why am I on this side?
00:28:22Why do I want you to succeed?
00:28:24You know, the Pope has spoken,
00:28:26I think, very importantly
00:28:27about not just AI being used
00:28:30in an inclusive manner,
00:28:32but being designed and built,
00:28:33that it's not a morally neutral tool.
00:28:35You know, Europe is a leader
00:28:37on ideas about protecting people's privacy
00:28:39and your ownership of your data
00:28:40and so forth.
00:28:42The world needs Europe
00:28:43to infuse those values
00:28:45into AI development.
00:28:46That's not something
00:28:47that's naturally going to happen
00:28:48based on capital allocation.
00:28:50So I want to just encourage you
00:28:52that no country is an island
00:28:53in that sense,
00:28:54but to the degree that you need to,
00:28:56much like nuclear weapons,
00:28:57France has maintained
00:28:58a totally independent
00:28:59strategic weapons capability.
00:29:01Just because it's hard
00:29:02doesn't mean that it's not possible.
00:29:04That's simply not true
00:29:05and we haven't heard
00:29:05any arguments otherwise.
00:29:07Great.
00:29:10Excellent.
00:29:11Thank you very much.
00:29:12No, it's fine.
00:29:13Very good round.
00:29:14And Leticia, really well done.
00:29:16Great French accent,
00:29:17but I'm obviously biased.
00:29:19But, you know,
00:29:20you made your case very clearly
00:29:21in your four or five points
00:29:23that you don't have
00:29:25the capital to scale
00:29:27and the consequence of that,
00:29:29you're losing critical talent,
00:29:30which you feel very acutely
00:29:33in the setup of your startup.
00:29:34Yes.
00:29:35A few of the takeaways
00:29:36I had from Christopher's comments
00:29:37that there's nothing special
00:29:38about even frontier AI models
00:29:40all over the world.
00:29:41You need to know
00:29:42what sovereignty means for you.
00:29:43Access is already undeniable
00:29:45in different places,
00:29:46including in France,
00:29:47including in Europe.
00:29:48And on this side,
00:29:50on our side,
00:29:50we invoke the Pope
00:29:53and to infuse,
00:29:55you know,
00:29:55to be sovereign
00:29:56is to infuse values
00:29:57and Europe is sort of
00:29:59leading the way on that.
00:30:00Great.
00:30:01So it's hotting up.
00:30:02Round three.
00:30:03We're ready.
00:30:04On the fourth side,
00:30:05we have Masha Ali Mardani.
00:30:08Sorry.
00:30:09I'm just slightly
00:30:11massacred that name.
00:30:13I'm sorry.
00:30:14Yes.
00:30:15And on this side here,
00:30:16we have Eric.
00:30:16How's on?
00:30:20So,
00:30:21Masha's fun facts.
00:30:22She told us that
00:30:24she once slept through
00:30:25a hotel fire
00:30:26in Brussels
00:30:27with noise-canceling
00:30:30headphones.
00:30:31A few hours after,
00:30:32the fire brigade
00:30:34came in her room
00:30:34and said,
00:30:35why have you not evacuated?
00:30:36And she said,
00:30:37well,
00:30:37I haven't heard anything.
00:30:38Anyway,
00:30:39if you can bring
00:30:39that level of concentration
00:30:40and focus to the debate,
00:30:42you will do a very good job.
00:30:44Yes.
00:30:44And Eric has written
00:30:46several books
00:30:47on French political history,
00:30:48including about
00:30:49the French Revolution.
00:30:50Now,
00:30:51the era,
00:30:51the dawning era
00:30:52of the trillionaire
00:30:53is like out of the scope
00:30:54of this debate,
00:30:55but maybe after this
00:30:56we can talk about that.
00:30:58Great.
00:30:59Masha,
00:30:59are you ready?
00:31:00Yes.
00:31:00Keep the mic close to the...
00:31:02Perfect.
00:31:02Three minutes.
00:31:03Good luck.
00:31:04So,
00:31:05my fellow debaters,
00:31:08Dakota and Leticia,
00:31:10have done a really good job
00:31:11of already arguing
00:31:14what the issue is.
00:31:16Europe doesn't have
00:31:17a frontier model.
00:31:18It doesn't necessarily
00:31:20have a way to scale.
00:31:21It doesn't have
00:31:22all the capital
00:31:22that it needs.
00:31:24And in a lot of cases,
00:31:25it doesn't even have
00:31:26the energy to run
00:31:27all the data centers
00:31:28that you need for AI.
00:31:29So,
00:31:30I'm not going to soften
00:31:31what they've already said.
00:31:32I'm just going to give you
00:31:34a new point,
00:31:36which is a week ago.
00:31:39The European Union
00:31:41published the Code of Practice
00:31:43on transparency for AI,
00:31:46part of the EU AI Act.
00:31:48We were lucky
00:31:49within my team at Witness
00:31:51to be part of the drafting
00:31:52of this code.
00:31:54So,
00:31:55what that code
00:31:57actually shows you
00:31:58is that
00:31:59it doesn't actually
00:32:01in its final output
00:32:03require
00:32:03a single chip
00:32:05to be made in Europe.
00:32:06It doesn't require
00:32:07a single model
00:32:08to be trained in Europe.
00:32:09And frankly,
00:32:10it doesn't even try
00:32:11to require these things.
00:32:13What it does
00:32:14is write rules
00:32:15that bind every AI system
00:32:16sold into the European market,
00:32:18whether they're American
00:32:19or Chinese,
00:32:21and requires them
00:32:22to mark their content
00:32:23through open standards,
00:32:25something we push for a lot.
00:32:27So,
00:32:27the major US providers
00:32:28are expected to sign
00:32:30within weeks
00:32:31before the rule
00:32:32takes into effect
00:32:33in August.
00:32:34But,
00:32:35what does this tell us?
00:32:37When Europe reaches
00:32:38for sovereignty,
00:32:40this is what it actually produces.
00:32:42Not a stack of its own,
00:32:44but a rule book
00:32:45for everyone else's stack.
00:32:47So,
00:32:48even EU's flagship AI law
00:32:51governs what others build.
00:32:55and it's frankly
00:32:57because it hasn't really
00:32:58built a competitive model
00:33:00of its own.
00:33:01So,
00:33:01I argue
00:33:02this motion is right
00:33:04and Europe doesn't have
00:33:05what it takes
00:33:06for sovereignty.
00:33:08And,
00:33:08what does Europe actually have?
00:33:11Because we're not here
00:33:12to beat Europe down.
00:33:13Europe offers us
00:33:15a lot of great things,
00:33:16but it's not achieving sovereignty.
00:33:18It is doing something different
00:33:20and maybe even something better
00:33:21for the world.
00:33:22It's protecting people
00:33:24in the face of technology
00:33:26it may never own.
00:33:28So,
00:33:29that's not sovereignty.
00:33:30Sovereignty asks us
00:33:32whose flag is on the infrastructure,
00:33:34on the server.
00:33:36But,
00:33:36this is actually
00:33:37maybe something else
00:33:39that we should focus more on,
00:33:41which is self-determination.
00:33:43And,
00:33:43self-determination asks
00:33:45what an ordinary person
00:33:47has rights to,
00:33:48what recourse they have
00:33:50to protect themselves.
00:33:51and it doesn't really matter
00:33:53who built the machine,
00:33:54whether it's American
00:33:55or Chinese.
00:33:56So,
00:33:57sovereignty is really about states
00:33:59and something like self-determination
00:34:02is about people.
00:34:04And,
00:34:04Europe can't win the first.
00:34:06Europe can't have sovereignty.
00:34:08And,
00:34:08that's
00:34:08my contribution here.
00:34:10Well done.
00:34:11Thanks.
00:34:16Eric.
00:34:17So,
00:34:18I will argue that
00:34:20Europe does not need
00:34:21to become America.
00:34:23By the way,
00:34:23it does not need
00:34:24to become China also.
00:34:26It needs to become
00:34:27Europe at full power.
00:34:29And,
00:34:29in that sense,
00:34:31it needs to become
00:34:32an industrial,
00:34:34an open,
00:34:35a trusted,
00:34:35a sovereign,
00:34:36and fast AI.
00:34:37This is why
00:34:38I oppose the motion.
00:34:40Europe does have
00:34:42what it takes
00:34:42to achieve
00:34:43AI sovereignty
00:34:44and I try to demonstrate that.
00:34:46It needs to build
00:34:47the future
00:34:48on the strength
00:34:49that it already has.
00:34:51So,
00:34:52first,
00:34:53so,
00:34:54we thank the Americans
00:34:55for the chatbots.
00:34:57But,
00:34:57the next wave of AI
00:34:58will not be about chatbots.
00:35:00It will be about
00:35:01industrial AI.
00:35:02And,
00:35:03industrial AI,
00:35:04to do that,
00:35:05you need factories.
00:35:06You need energy grids.
00:35:07You need hospitals.
00:35:08You need banks.
00:35:09You need defense systems.
00:35:10You need agriculture.
00:35:11You need aerospace.
00:35:13You need mobility.
00:35:13You need construction materials.
00:35:15Pharma.
00:35:16Supply chains.
00:35:17And,
00:35:17that is
00:35:18Europe terrain.
00:35:20Europe has
00:35:20Airbus,
00:35:21Schneider Electric,
00:35:22Siemens,
00:35:23Dassault System,
00:35:24Thales,
00:35:25Leonardo,
00:35:26Stellantis,
00:35:26Mercedes,
00:35:27BMW,
00:35:28and a few other companies
00:35:29that you may know.
00:35:30These are not
00:35:31marginal assets.
00:35:32They are exactly
00:35:33the industrial clients
00:35:35that are owners
00:35:36and deployment
00:35:37environments
00:35:37where AI
00:35:38can create productivity.
00:35:40This is where
00:35:41resilience
00:35:42and sovereignty
00:35:43for Europe
00:35:45actually relies.
00:35:46There is
00:35:47another reason
00:35:49why I reject
00:35:50the idea
00:35:51that the gap
00:35:51is not unbridgeable.
00:35:53Actually,
00:35:53the current AI
00:35:54paradigm
00:35:55is not the final
00:35:57paradigm.
00:35:58World models
00:35:58will be the next
00:35:59frontier.
00:36:00and this next frontier
00:36:03is actually
00:36:03being built
00:36:04in Europe
00:36:04by Naline Nucca
00:36:05and a few others.
00:36:07So,
00:36:07Europe may have
00:36:08been late
00:36:08to the first wave
00:36:10or large language model,
00:36:11namely the chatbots,
00:36:12but it can still be early
00:36:13to the next wave,
00:36:14the world models,
00:36:15physical AI,
00:36:16and industrial autonomy.
00:36:18Then,
00:36:18there is the defense
00:36:19industrial ecosystem
00:36:21that we are currently building.
00:36:22And we thank
00:36:23really deeply
00:36:25Donald Trump
00:36:25for letting us
00:36:27building a military
00:36:29or industrial complex
00:36:30and withdrawing
00:36:31for Europe.
00:36:31It will enable us
00:36:32to rebuild
00:36:33demand at scale
00:36:35in cyber security
00:36:37in drones
00:36:38and the real defense demand.
00:36:39The war in Ukraine
00:36:40has shown
00:36:41that the future
00:36:42of defense
00:36:42is actually inseparable
00:36:44from AI.
00:36:45And we can be
00:36:47at the forefront
00:36:47of this.
00:36:48We can create
00:36:49an AI flywheel
00:36:50rooted in security,
00:36:52in autonomy,
00:36:54and in industrial capabilities.
00:36:56You got the point.
00:36:56It happens
00:36:57that we are also finally
00:36:58at the right scale.
00:37:00We're investing
00:37:01with the Invest AI initiative
00:37:03200 billion euro
00:37:05across the next years
00:37:07and we have
00:37:08a 20 billion euro fund
00:37:10for AI gigafactories.
00:37:11This is why
00:37:13the question is
00:37:14whether Europe
00:37:16can become
00:37:17a late Silicon Valley.
00:37:18The question is
00:37:20whether Europe
00:37:20can build,
00:37:21control,
00:37:22deploy,
00:37:22and govern
00:37:23AI
00:37:23in the areas
00:37:24that matter
00:37:25most for its economy.
00:37:27Well done.
00:37:32All super disciplined
00:37:34debaters.
00:37:35A bit of convergence
00:37:36of arguments
00:37:37in this round
00:37:37interestingly
00:37:38but I think
00:37:39Masha,
00:37:40I think you made
00:37:40a very good point
00:37:41that sovereignty
00:37:42might not just be
00:37:43the right goal
00:37:43for Europe.
00:37:44Self-determination
00:37:45is
00:37:46and some
00:37:46very interesting
00:37:47reflection
00:37:48on Europe
00:37:50ability
00:37:50to write
00:37:50a rule book.
00:37:51I usually
00:37:51implied
00:37:52I'm American.
00:37:53Yeah.
00:37:55Okay.
00:37:56And on this side
00:37:57what I took away
00:37:58is that
00:37:59European industry
00:38:00is not marginal.
00:38:02There's this flywheel
00:38:03between security,
00:38:03autonomy,
00:38:04industrial capability.
00:38:05Maybe Europe
00:38:06has missed
00:38:06the first wave
00:38:07of chatbots,
00:38:08consumer-facing AI,
00:38:09but it is well-placed
00:38:10for the next wave.
00:38:12Great.
00:38:13Final round.
00:38:14Ready?
00:38:15So,
00:38:16on the fourth side
00:38:17please join me
00:38:18in welcoming
00:38:18Thomas Lingard.
00:38:20All right.
00:38:22And on the against
00:38:22we have
00:38:23Clara Chappaz.
00:38:26All right.
00:38:30So,
00:38:30last but by no means least.
00:38:32So,
00:38:33Thomas,
00:38:33you might not know
00:38:34that about him
00:38:35but he loves
00:38:37karaoke.
00:38:38He's actually
00:38:38a very good singer
00:38:39but unfortunately
00:38:41we just put a mic
00:38:42in front of him
00:38:42so hopefully he knows
00:38:44he's only had
00:38:44three minutes.
00:38:46Anyway,
00:38:47Thomas,
00:38:48the last thing
00:38:48we need
00:38:49in an AI debate
00:38:50is someone
00:38:50who is stone death.
00:38:52Ouch.
00:38:54Clara has been
00:38:55pretty intimately involved
00:38:56with the G7
00:38:57negotiations
00:38:58this week
00:38:59and there's probably
00:39:00a lot that she
00:39:01could tell us about it
00:39:01and none of it
00:39:02that she's allowed
00:39:03to tell us
00:39:04but maybe after
00:39:05you know.
00:39:06All right.
00:39:07Thomas,
00:39:07with all your card
00:39:08laid out,
00:39:09over to you.
00:39:10You go.
00:39:10I only said yes
00:39:11because it involved
00:39:12a microphone
00:39:12I thought it might
00:39:13turn into karaoke.
00:39:15Thank you very much.
00:39:18Look,
00:39:19we started today
00:39:20with a brilliant
00:39:20setting out
00:39:21of the definition
00:39:22and it's very important
00:39:23that you,
00:39:24the audience,
00:39:25pay attention
00:39:25to the definition
00:39:26when deciding
00:39:27how to vote
00:39:28and sovereignty
00:39:30has been defined
00:39:31as I think,
00:39:32correct me if I'm wrong,
00:39:33but the ability
00:39:33to develop,
00:39:34control and govern
00:39:36AI
00:39:36without dependence
00:39:38on foreign powers
00:39:39and I want to ask you
00:39:40to listen to what
00:39:41you've heard
00:39:41from actually both sides
00:39:43because I'm going to argue
00:39:43that we've made
00:39:44a convincing case
00:39:45and that these guys
00:39:46have also made
00:39:46a convincing case
00:39:47for us
00:39:47that it's not possible,
00:39:49it's not possible
00:39:51for Europe
00:39:52to have sovereign AI.
00:39:55So you've heard Dakota
00:39:56talk about the fact
00:39:57there's not enough money
00:39:59and it's not just
00:40:00that there's not
00:40:01enough money now
00:40:02but look where Europe is,
00:40:03it's fighting a land war,
00:40:04it's got rising
00:40:05social security costs.
00:40:06Where is this money
00:40:07going to come from?
00:40:08Even the very recent
00:40:10quite positive announcement
00:40:11about what Europe
00:40:12wants to do on AI
00:40:13from some governments
00:40:14is very clear
00:40:15it contains no new
00:40:16public spending commitments.
00:40:17Europe is broke,
00:40:18it does not have the money
00:40:19to build the capabilities
00:40:21that we would need
00:40:23in the time
00:40:24that we would need them.
00:40:25You've also heard this
00:40:26really personal story
00:40:28from Letitia
00:40:29about trying to build
00:40:30a startup in Europe
00:40:31and we have so many
00:40:32great startups,
00:40:33we actually generate
00:40:34more startup per capita
00:40:36than the US.
00:40:37We have an amazing
00:40:38startup ecosystem
00:40:38and then everybody
00:40:40goes over the pond
00:40:41to the US
00:40:42to get the capital
00:40:43that they need
00:40:44and the ecosystem
00:40:45that supports the scale-up
00:40:46of those organizations.
00:40:48So, you know,
00:40:50we know that
00:40:50stronger interconnection
00:40:52is actually what Europe needs,
00:40:53not building defensive walls
00:40:55and this illusory sovereignty
00:40:58that has been talked of.
00:41:01But I want to talk also
00:41:03about the kind of
00:41:04political will challenge
00:41:06facing European governments.
00:41:10You could imagine, right,
00:41:11you could imagine,
00:41:12we've been invited to imagine
00:41:13by the other team
00:41:14that this might be possible
00:41:15if we got all our ducks in a row,
00:41:18if we marshaled all the capital,
00:41:20if we got everybody to agree
00:41:22that we could do this.
00:41:24But the idea that we would get
00:41:25the full tech stack,
00:41:27the chips,
00:41:28the lithography,
00:41:29the software,
00:41:30the hardware,
00:41:30owned, operated,
00:41:32under the control
00:41:33of European governments
00:41:34is ridiculous.
00:41:37I think even they would agree that.
00:41:39They've kind of already
00:41:40conceded that.
00:41:41If you imagine
00:41:42that we could do it
00:41:42with foreign technology,
00:41:44but we could survive
00:41:46a situation where we say,
00:41:48okay, something's gone wrong
00:41:49over there,
00:41:50a very close,
00:41:51friendly country
00:41:52is now not so friendly,
00:41:53we want to be able
00:41:54to switch that off,
00:41:55it's very hard to imagine
00:41:57us being able
00:41:58to have the political will
00:41:59to switch off technology
00:42:01from any country
00:42:02and suffer politically
00:42:04the consequences of doing that,
00:42:06which could be quite devastating
00:42:07in any country
00:42:08at any one time.
00:42:09We are locked into
00:42:10these relationships.
00:42:11The idea that we could be
00:42:12entirely sovereign
00:42:13is an illusion.
00:42:17And even if you imagine
00:42:18that somehow we could do that,
00:42:20there is still,
00:42:22I can see the timer running out,
00:42:23so I'm going to jump a card,
00:42:25there's still the question
00:42:27of how fast we could go
00:42:30and what we could achieve.
00:42:31But let me just leave you
00:42:33with the simple thought
00:42:34that, you know,
00:42:39sovereignty,
00:42:40European AI sovereignty
00:42:41is something that we can't build,
00:42:43we can't afford,
00:42:44and we can't keep.
00:42:45And if we try and do something
00:42:46that looks like that,
00:42:48it won't actually be sovereignty,
00:42:49it will be a flag
00:42:50on a rented building.
00:42:52Thank you very much.
00:42:53Ouch.
00:42:55All right.
00:42:59Your time begins now.
00:43:02Good.
00:43:03Well,
00:43:04the first way
00:43:05to answer this question
00:43:06and the reason
00:43:07I'm completely against emotion
00:43:09is
00:43:09where about are we?
00:43:12In Europe.
00:43:13And I don't think
00:43:14anyone was forced
00:43:15to come to Vivitech.
00:43:17You came here
00:43:17because you think
00:43:18Europe has value,
00:43:19let you be an investor,
00:43:21an entrepreneur,
00:43:22someone working in policy.
00:43:24I mean,
00:43:25even a Google person
00:43:26is on our side.
00:43:28I let you think through that.
00:43:31Second,
00:43:32well,
00:43:33you know,
00:43:34we've been helped,
00:43:35you talked about G7,
00:43:37we've been helped
00:43:37by a few friends lately
00:43:39to realize that sovereignty
00:43:41is not just
00:43:42some marketing word
00:43:43because we don't really
00:43:45have a choice
00:43:45and I think
00:43:46everyone has understood
00:43:48that this week
00:43:49because who wants
00:43:51to be reliant
00:43:51on technologies
00:43:52that some people
00:43:54in the world
00:43:54can decide to switch off
00:43:56from one day
00:43:57to the other
00:43:57when those technologies
00:43:59are actually
00:44:00saving lives
00:44:01in the hospital,
00:44:02helping our children
00:44:03get educated
00:44:04or doing all the steps
00:44:05you need to do
00:44:06every day at work.
00:44:07No one.
00:44:08Two,
00:44:09three,
00:44:10sorry,
00:44:10we do have
00:44:12everything it takes.
00:44:13I heard it,
00:44:14oh yeah,
00:44:15you need to have it
00:44:16really big.
00:44:17We're going to come
00:44:18to that in a minute.
00:44:19But we have the data,
00:44:20industrial data,
00:44:21it's all here
00:44:22in Europe,
00:44:23it's nowhere else.
00:44:24We have the energy.
00:44:25Someone was not
00:44:26breathed perfectly.
00:44:28Where is the nuclear energy
00:44:29in this country
00:44:31and where are investors
00:44:32coming in this country?
00:44:34There is no single
00:44:36basis in France
00:44:37where I hear
00:44:38lenders saying,
00:44:39hey,
00:44:39we are going to put
00:44:40a morandum
00:44:41on data centers
00:44:42because we don't have
00:44:43access to energy anymore
00:44:44so we cannot train
00:44:45those models.
00:44:46We exported the equivalent
00:44:47of the entire
00:44:48Belgium consumption
00:44:49in nuclear power
00:44:50just last year.
00:44:52So come along,
00:44:53we can train your models,
00:44:54I know you need us.
00:44:56It's green
00:44:56and it's sustainable
00:44:57so that's also pretty good
00:44:59and it's predictable.
00:45:00Talents,
00:45:01who are building
00:45:02those models
00:45:03everywhere in the world?
00:45:05Europeans,
00:45:06even in San Francisco,
00:45:08you rely on us.
00:45:09Capital,
00:45:10again,
00:45:11wrong,
00:45:11we have more savings
00:45:12in Europe
00:45:13than everywhere
00:45:14in the US.
00:45:15Okay,
00:45:16we don't know
00:45:16how to invest it properly
00:45:17but we are going
00:45:18to solve that.
00:45:20And yeah,
00:45:21the rule book,
00:45:22the good old rule book,
00:45:23well,
00:45:23it's also called
00:45:24predictability
00:45:25and in a world
00:45:26where everything changes
00:45:27by the minute,
00:45:28predictability is not bad.
00:45:30So maybe there's one thing
00:45:31we don't have.
00:45:33We always feel sorry
00:45:35for ourself.
00:45:36Like honestly,
00:45:37I heard Europeans
00:45:37on this stage
00:45:39just saying,
00:45:40we are bad,
00:45:41we are ugly,
00:45:42we don't know
00:45:43how to do things,
00:45:44this is so bad,
00:45:45Europe is so shitty.
00:45:46Do you think
00:45:47you're selling
00:45:47and building
00:45:48the greatest technology
00:45:49going around
00:45:49and saying you smell?
00:45:50No.
00:45:51So what we need
00:45:52is a good bunch
00:45:53of marketing
00:45:54and I think
00:45:55we can do that.
00:45:56I mean,
00:45:56we need to do that
00:45:57because the world
00:45:58is not waiting
00:45:58and if we want
00:45:59to build for ourself,
00:46:00we need to believe
00:46:01in ourself.
00:46:02That's the only thing
00:46:03we're missing.
00:46:04So I hope I convince you
00:46:05to believe in yourself
00:46:06because we're here for you.
00:46:12Wonderful.
00:46:14Wonderful.
00:46:16A very,
00:46:17very strong
00:46:18final round
00:46:19and Thomas,
00:46:20you're basically saying
00:46:21we can't afford it,
00:46:22we need to align
00:46:23too many players,
00:46:24we are too interconnected
00:46:25to meet the relentless
00:46:27pace of AI development.
00:46:30Yeah,
00:46:30and listen,
00:46:31why else are you here,
00:46:33right?
00:46:33Don't feel sorry
00:46:34for yourself out there.
00:46:36Rules are there
00:46:37for predictability.
00:46:38It's not this like
00:46:39bad, scary thing
00:46:40and this is where we are,
00:46:42you know,
00:46:42and there's all the,
00:46:44there's Belgian energy,
00:46:45you know.
00:46:46Belgium.
00:46:46Many Belgians
00:46:47are available
00:46:48for training your models.
00:46:50All right,
00:46:51so we're now moving
00:46:51to the next round
00:46:52which is inviting
00:46:53jury members
00:46:54to come to the stage.
00:46:56So a reminder,
00:46:57jury members
00:46:58are not yet the judges.
00:47:00They're really here
00:47:01in this sort of
00:47:02enviable position
00:47:04of stir things up
00:47:04a little bit
00:47:05and we're asking
00:47:06jury members
00:47:07to think about
00:47:07what they like,
00:47:08about what they heard
00:47:09on both sides
00:47:10of their arguments,
00:47:11what they might have
00:47:12wanted to hear
00:47:14from either side
00:47:15of the argument,
00:47:16what is missing
00:47:17and maybe what
00:47:18the debaters
00:47:19should think about
00:47:19when they come
00:47:20for their final retort.
00:47:21So please,
00:47:22join me
00:47:23in welcoming
00:47:24to the stage
00:47:24our three jury members,
00:47:27Martin Hullin,
00:47:28Brune Poisson
00:47:29and Louise Agusnap.
00:47:40So who is going first?
00:47:42I guess I've been
00:47:43volunteered.
00:47:44Good, Martin,
00:47:45you go.
00:47:46Such a pleasure.
00:47:46Up to two and a half minutes.
00:47:47Yes, thank you so much.
00:47:49Thank you for the
00:47:50very, very convincing
00:47:51arguments on both sides.
00:47:53Thank you for the
00:47:53four motion perspectives,
00:47:55also the
00:47:56against motion
00:47:57and I would like
00:47:57to start with
00:47:58the four motion
00:47:58kind of reflection quickly
00:48:00on what I felt
00:48:01was still missing.
00:48:03One observation
00:48:04we are currently having
00:48:05is this moment
00:48:06of a tech clash
00:48:07in the US
00:48:08which is real.
00:48:09Eric Schmidt being booed
00:48:10during the graduation
00:48:11and wondering
00:48:12where this could actually
00:48:14then fit into
00:48:15the narrative
00:48:15that you've been sharing
00:48:16around the question
00:48:17of the ones
00:48:18that are catching up
00:48:19if this is the narrative
00:48:20that is there.
00:48:21Observeers also say
00:48:22that dependence
00:48:23is coercion in waiting.
00:48:25Hence, you know,
00:48:26the first examples
00:48:27we've heard about
00:48:27Fable and other
00:48:28geopolitical weaponization.
00:48:30Should Europe
00:48:31just accept that
00:48:32or any other clients
00:48:33outside of the United States?
00:48:36And the final point also,
00:48:38Fable-like capability
00:48:39as was mentioned
00:48:39by the colleague
00:48:40from Google
00:48:40will be available
00:48:41in China in three months
00:48:43and as open source
00:48:44models in 12.
00:48:45Don't we have the time
00:48:46to wait
00:48:46to actually see
00:48:47what works out there?
00:48:48And on the against side
00:48:50that says that it's possible,
00:48:52I was really wondering
00:48:53to what extent also
00:48:54you could clarify further
00:48:56in what parts
00:48:57of the stack
00:48:58Europe could actually
00:48:59make itself indispensable
00:49:00because ASML
00:49:01is of course
00:49:02a wonderful example
00:49:03but we use it
00:49:05quite extensively
00:49:06and there are plenty
00:49:07of other areas
00:49:08I think that this argument
00:49:08could actually stick.
00:49:10And also,
00:49:11the narrative
00:49:11on the other side
00:49:12actually has unleashed
00:49:14lots of investments
00:49:15and political influence
00:49:16that we are all affected by
00:49:17at this stage.
00:49:18And I'm wondering,
00:49:19is Europe successful enough
00:49:21in countering it
00:49:22with a proactive vision
00:49:23that aligns also
00:49:24the markets
00:49:25and how can we find a way
00:49:28of not just using
00:49:29the kill switch framing?
00:49:30And I very much liked
00:49:31the courage
00:49:32and also the marketing angle
00:49:33but I think we can do better
00:49:35in order to actually
00:49:37also copy
00:49:38the other side
00:49:39in this regard.
00:49:40So I leave it there
00:49:41and yield some of the time.
00:49:42Great set of comments.
00:49:43Thank you, Martin.
00:49:44Please.
00:49:50Louise.
00:49:51Thank you so much
00:49:52and thanks to the two
00:49:54debater teams
00:49:54for great energy
00:49:56and wit
00:49:57and creativity
00:49:58and some humor.
00:49:59I think what resonated
00:50:01with me was
00:50:02in general
00:50:03when you got concrete
00:50:04on both sides.
00:50:05I think that really
00:50:06speaks volumes
00:50:07when you can bring it down
00:50:08to a specific law,
00:50:10a specific experience.
00:50:11I think this founder experience
00:50:13over here
00:50:14stood really out for me.
00:50:16I felt really sorry
00:50:17for you
00:50:18in trying to build
00:50:19your company
00:50:20and not being able
00:50:21to find the ingredients
00:50:23you needed.
00:50:24And so those kinds
00:50:26of examples,
00:50:27a piece of advice,
00:50:28if you can find more
00:50:29of that on either side,
00:50:31the personal experience,
00:50:32the concrete experiences
00:50:33and examples
00:50:35from law,
00:50:36investments
00:50:37that France is making,
00:50:39et cetera,
00:50:40those things
00:50:40speak volumes.
00:50:44There's been a lot
00:50:45of sort of debate
00:50:47here about
00:50:47what is sovereignty
00:50:49and I think
00:50:50the importance here
00:50:52is to not make it
00:50:53a meta debate
00:50:54but make it
00:50:54more concrete.
00:50:56I'm wondering
00:50:57whether there is
00:50:58a sort of
00:50:59a third way
00:51:00because right now
00:51:01we're defining
00:51:02sovereignty
00:51:02as sovereignty
00:51:05from the US
00:51:07or China
00:51:08but what if Europe
00:51:09took on a third way
00:51:11where it's Europe
00:51:13bringing the world along,
00:51:15being a steward
00:51:16of innovation,
00:51:17really rallying
00:51:18the innovation ecosystem
00:51:20globally
00:51:21and showing
00:51:22that there is
00:51:22a third way
00:51:23where we can build
00:51:25something that's better
00:51:26for global social purpose,
00:51:28better for global human health
00:51:30and really leveraging
00:51:31the capabilities of AI
00:51:33which is a gift
00:51:33to us all
00:51:34if we can get it right.
00:51:37So I'll stop there.
00:51:38Thank you, Louise.
00:51:44And Brune,
00:51:45over to you.
00:51:46Thank you very much.
00:51:47It was a great debate.
00:51:48I think that echoes
00:51:50really well
00:51:50the discussions
00:51:51that are currently
00:51:52going on.
00:51:54Perhaps
00:51:56I would have loved
00:51:58and maybe this is
00:51:59what we're going to have
00:51:59in the rest of the debate
00:52:01a bit more about
00:52:03I think there's
00:52:04one elephant in the room
00:52:05and we barely heard
00:52:06that word
00:52:07which is China
00:52:09and I think
00:52:10you touched on that.
00:52:11How really do we
00:52:12how can Europe
00:52:13position itself
00:52:14and beyond that
00:52:15maybe the US
00:52:16also will need
00:52:17some allies
00:52:18finally
00:52:19and eventually
00:52:20and this could be
00:52:22a great opportunity
00:52:23to strengthen
00:52:23its link
00:52:24and relationship
00:52:25with the EU
00:52:26needing it more
00:52:27than it thought initially.
00:52:30So first,
00:52:31China.
00:52:31It would have been great
00:52:32also to hear
00:52:33maybe a bit more
00:52:34about the energy
00:52:35that it takes
00:52:36the water
00:52:37that it requires
00:52:38and the critical
00:52:39raw materials
00:52:40and all of that
00:52:41perhaps not the raw materials
00:52:43but at least
00:52:44energy
00:52:45and our ability
00:52:46to manage water
00:52:46this is something
00:52:47that we have.
00:52:49Another thing also
00:52:50is a social license
00:52:51to operate.
00:52:53Maybe
00:52:54it would have been great
00:52:55to have the US
00:52:57point of view
00:52:58on that
00:52:58as we know
00:52:59a lot of the data
00:53:00centers
00:53:01are now
00:53:02being very difficult
00:53:03to build
00:53:05in the US
00:53:06and maybe
00:53:07this is something
00:53:07Europe is very good at
00:53:09or better at
00:53:09public debates
00:53:10building consensus
00:53:12not always in France
00:53:13not these days
00:53:14but sometimes
00:53:15and maybe finally
00:53:16and this is also
00:53:18something
00:53:18maybe you can
00:53:19keep on discussing
00:53:20in the rest of the debate
00:53:21is really the finance part
00:53:23and the billions
00:53:25especially in France
00:53:26obviously
00:53:26but in a lot of
00:53:27European countries
00:53:28as well
00:53:29of private savings
00:53:30that we have
00:53:31and that we could mobilize
00:53:32if we learn
00:53:33how to channel it
00:53:34to the right directions.
00:53:37So these are points
00:53:38that would have been great
00:53:39and that may be great
00:53:40to hear more about.
00:53:42It doesn't mean
00:53:43that the debates
00:53:43were not great.
00:53:44It doesn't mean
00:53:45that we're facing
00:53:47a very complicated
00:53:48I think job
00:53:49in the jury right now.
00:53:54And thank you
00:53:55to members of the jury
00:53:57for your excellent
00:53:59contributions.
00:54:00You can return
00:54:00to your seats there.
00:54:01So debaters
00:54:02you've heard
00:54:03what the jury
00:54:04has had to say.
00:54:05I think there's
00:54:06some very pertinent
00:54:06points there
00:54:07for both sides.
00:54:07So now this is
00:54:08your opportunity
00:54:09to respond
00:54:10to them.
00:54:12We have three minutes
00:54:14for each group
00:54:15to confer
00:54:16and prepare
00:54:17your final rebuttal.
00:54:18This is your last chance
00:54:19to convince us
00:54:21well to convince them
00:54:23and to clinch victory.
00:54:25So audience,
00:54:26while they confer,
00:54:28we invite you
00:54:29to discuss
00:54:29amongst yourselves
00:54:31or a chatbot
00:54:32of choice
00:54:33or the prediction markets
00:54:34or whatever.
00:54:35What do you think?
00:54:37Which side do you think
00:54:38won the day?
00:54:39What would you,
00:54:40if you were on the jury,
00:54:41what points would you bring up?
00:54:42What would you want to hear
00:54:43in this final rebuttal?
00:54:45The clock starts now
00:54:46and we'll be back
00:54:47in three minutes.
00:54:48And I think we have a video
00:54:49during that moment
00:54:50as well
00:54:50from the New York Times.
00:55:36This is the end.
00:55:56Talk amongst yourselves.
00:55:58Four.
00:55:59Against.
00:56:00Maybe you...
00:56:01Against.
00:56:01Four.
00:56:02A few more minutes.
00:56:05Anyway,
00:56:05do you remember,
00:56:06that was not an easy job.
00:56:07So thank you
00:56:08for coming on stage
00:56:09reacting like that.
00:56:10They didn't prepare
00:56:11any of that.
00:56:12So I don't know
00:56:12how easy it is to do
00:56:14but I admire
00:56:15what you've done.
00:56:16So thank you.
00:56:17We're giving the speakers
00:56:18another minute.
00:56:22Some heavy debates
00:56:24amongst the debaters here.
00:56:25It's getting meta.
00:56:26Okay.
00:56:52Are we ready?
00:56:54Are we about ready?
00:56:56All right.
00:56:57Final furious scribbling.
00:57:00Pencils down.
00:57:01We're going to switch
00:57:02things around.
00:57:03We're going...
00:57:04Clara meant to say...
00:57:05Sorry if I forgot
00:57:06but we're starting
00:57:07with the against side.
00:57:09Yes.
00:57:10So...
00:57:10Yes.
00:57:11We're swapping it around.
00:57:13You have a minute and a half
00:57:15to final rebuttal
00:57:17on the against side.
00:57:18This house believes
00:57:18that Europe does not have
00:57:19what it takes
00:57:19to achieve AI sovereignty.
00:57:21Over to you.
00:57:23Hold on.
00:57:24Let's make sure the...
00:57:25Are you ready?
00:57:26This is when it gets
00:57:27a little raggy,
00:57:27a little chaotic.
00:57:28That's part of the fun.
00:57:30Yeah.
00:57:31Yes.
00:57:31You can go afterwards.
00:57:33Clara, over to you.
00:57:34Cool.
00:57:35So we heard you
00:57:36and we are going to convince you
00:57:38that Europe
00:57:39has a heavy thing it takes.
00:57:40First,
00:57:41we didn't talk
00:57:42about acceptability
00:57:43and this is a big word
00:57:45behind sovereignty
00:57:46and we don't have people
00:57:48who are, you know,
00:57:49going in the streets
00:57:50and fighting against AI
00:57:51in Europe
00:57:52because we know
00:57:53how to get our people
00:57:54on board
00:57:54and we have all the resources
00:57:56we need
00:57:56to make it
00:57:57in a smooth way.
00:57:58Data centers are welcome
00:58:00and you need data centers
00:58:01to build AI.
00:58:02Two,
00:58:03sovereignty doesn't mean isolation.
00:58:06Sovereignty doesn't mean
00:58:07closing doors.
00:58:08It means being,
00:58:10you know,
00:58:10partnering with allies
00:58:12and being reliable
00:58:14as a partner,
00:58:15as a friend
00:58:16is extremely important
00:58:17because this conversation
00:58:18about AI sovereignty
00:58:19is not only
00:58:20a European conversation.
00:58:22From India
00:58:23to Korea,
00:58:24Japan,
00:58:24I was in Kenya
00:58:26just a few weeks ago,
00:58:27we organized a big summit
00:58:28and the conversation
00:58:30we had with African
00:58:31head of states
00:58:31was exactly the same.
00:58:33Guess what?
00:58:34They don't want to be locked
00:58:35in east side
00:58:36or west side.
00:58:37They want to look forward
00:58:39and they want to look forward
00:58:40with us
00:58:40because we also want
00:58:41to look forward.
00:58:42So we are this third voice
00:58:44that everyone
00:58:45is looking for
00:58:46and that's very attractive
00:58:47to build sovereignty.
00:58:48And last,
00:58:50this conversation
00:58:51in this group
00:58:51is only about
00:58:53capital and money.
00:58:54But this is not
00:58:56either
00:58:57something that
00:58:58should make us
00:58:58believe it's not possible
00:58:59because first,
00:59:00we do have
00:59:01private savings.
00:59:02We have more
00:59:02private savings
00:59:03in Europe
00:59:04than in the US.
00:59:06450 million people,
00:59:07yes,
00:59:08size matter
00:59:08and we're bigger.
00:59:10And on top of that,
00:59:12you know,
00:59:12we've been through
00:59:13hard times
00:59:14where we've actually
00:59:15been broke.
00:59:16World War II
00:59:16and guess what?
00:59:17We built nuclear weapon
00:59:19out of that
00:59:20so we can do it.
00:59:26All right.
00:59:27Excellent.
00:59:27A very poised performance.
00:59:29So 90 seconds
00:59:30to seal the deal
00:59:31on the foresight.
00:59:32Why don't you all go?
00:59:33Come on
00:59:34and you decide how you...
00:59:35We decided
00:59:36we needed more
00:59:37than one person
00:59:37to go up
00:59:38against a minister.
00:59:39And to optimize time,
00:59:41you can have
00:59:42more than one mic.
00:59:43Okay, you go.
00:59:44So very quickly,
00:59:45I just want to say
00:59:46yes,
00:59:46France has a great
00:59:48power grid
00:59:49so congratulations
00:59:50on that.
00:59:52But what is really
00:59:53being built with it,
00:59:54there's a supercomputer
00:59:56France uses
00:59:56that's built
00:59:57by the British
00:59:58running on American ships
00:59:59funded largely
01:00:00by foreign capitals
01:00:01serving models
01:00:02that have been
01:00:03trained elsewhere.
01:00:05So France supplies
01:00:06the socket
01:00:07and congratulations
01:00:08on being the world's
01:00:09cleanest landlord.
01:00:13And something we need
01:00:14to say that
01:00:15we are whole
01:00:16European committed
01:00:17people.
01:00:18We want Europe
01:00:19to be the best
01:00:19in the world.
01:00:20It's a simple fact.
01:00:22But there is something
01:00:23we need to do today.
01:00:25It's to being honest.
01:00:26And being honest
01:00:27says we need
01:00:28to take action
01:00:29to do better tomorrow.
01:00:32So actually,
01:00:32it's not true.
01:00:33Capital is a problem
01:00:34in Europe.
01:00:34We raise four times
01:00:35more in the US
01:00:36than in Europe.
01:00:37There is a problem
01:00:38in also data.
01:00:40We don't have
01:00:40much data center
01:00:41than in the US
01:00:42or in China.
01:00:43We don't have
01:00:43the same system
01:00:44that China.
01:00:45We can't today
01:00:46have what it takes
01:00:47to have high sovereignty.
01:00:48But we do have
01:00:50we do need
01:00:51to have it tomorrow.
01:00:52So we need
01:00:53to be honest today.
01:00:55And finally,
01:00:57the point about the money.
01:00:58I mean,
01:00:59all of these wonderful
01:01:00things that Europe
01:01:00could do.
01:01:01The European Capital
01:01:02Markets Project
01:01:02was started in 2015
01:01:04and is still in progress.
01:01:06We're just not very good
01:01:07at moving these things
01:01:08beyond committee stage
01:01:09into the real world.
01:01:11But I want to come back
01:01:12just finally
01:01:12for this definitional point.
01:01:14You know,
01:01:14we started with
01:01:15the definition
01:01:15of what sovereignty means.
01:01:16The other team
01:01:17have just seeked
01:01:18to redefine sovereignty.
01:01:20And downgrading
01:01:21downgrading
01:01:22the definition
01:01:22is not an argument
01:01:24that Europe
01:01:24has what it takes.
01:01:25It's a concession
01:01:26by this team
01:01:28that it doesn't.
01:01:30Nice.
01:01:32Punchy ending.
01:01:33A punchy ending.
01:01:34I like it very much.
01:01:35Wonderful.
01:01:36So Jason had warned you
01:01:37that at some stage
01:01:38in this debate
01:01:39the gloves come off.
01:01:40And I'm glad
01:01:41we got to that stage
01:01:42of this debate.
01:01:43So this is
01:01:44the moment of truth.
01:01:45I know we are
01:01:45a bit past time
01:01:46so we're going to
01:01:46speed through that process.
01:01:48But audience
01:01:49this is your moment
01:01:50to vote
01:01:51and remember
01:01:52you have your own
01:01:53personal opinion
01:01:54which you came
01:01:55into the room with
01:01:56and then you have
01:01:57an appreciation
01:01:57for how the debaters
01:01:58I'm not being biased
01:01:59how the debaters
01:02:00on either side
01:02:01of the argument
01:02:02have argued
01:02:02and who has made
01:02:04the most compelling argument
01:02:05in a very complex
01:02:06set up.
01:02:08So
01:02:09I'm going to ask you
01:02:10to clap
01:02:11and cheer
01:02:11for the team
01:02:13that you thought
01:02:13won the debate
01:02:15put the best debate
01:02:16forward.
01:02:16If you think
01:02:17that the four team
01:02:18put the best debate
01:02:19forward
01:02:20please clap now.
01:02:22Pretty good.
01:02:24Pretty good.
01:02:27And hey
01:02:28guess what
01:02:29if you think
01:02:29the against team
01:02:31made the strongest
01:02:32argument
01:02:32please clap now.
01:02:36Very close.
01:02:37Very close.
01:02:40I don't know
01:02:41about you Jason
01:02:41but there's a fair
01:02:42amount of double
01:02:42clapping involved
01:02:43but we would
01:02:44allow for that.
01:02:45But yes
01:02:46I think
01:02:46we have a winner
01:02:47right?
01:02:48I think we do.
01:02:50Well we have a winner.
01:02:51Bravo.
01:02:52Bravo
01:02:53against team.
01:02:56As you have
01:02:57worked it out
01:02:58by now
01:02:58who wins
01:02:59who loses
01:03:00the argument
01:03:00is somehow
01:03:01completely secondary
01:03:02to the process
01:03:03that we're trying
01:03:03to do.
01:03:04We told you
01:03:05that this team
01:03:05had obviously
01:03:06to push the argument
01:03:07beyond their own
01:03:08conviction
01:03:08for the sake
01:03:09of the argument
01:03:09and they've done
01:03:10a wonderful job
01:03:11so bravo
01:03:12to the four team.
01:03:17Obviously
01:03:18bravo
01:03:19to the against team
01:03:20you won the debate.
01:03:21Thank you to the
01:03:22jury members as well.
01:03:23Thanks to the jury.
01:03:27And I think
01:03:28you should hug it out
01:03:29actually teams
01:03:30in the spirit
01:03:31of the European
01:03:32spirit of solidarity
01:03:34you should hug it out.
01:03:35But I really want
01:03:36to close by
01:03:36thank you Jason
01:03:38for being
01:03:39a wonderful
01:03:39co-moderator.
01:03:41Thank you very much
01:03:41and thanks to you Sophie.
01:03:43Thank you as well.
01:03:45And thank you
01:03:46the audience
01:03:47for your engagement.
01:03:49Thank you very much
01:03:49everyone.
01:03:50Bye.
01:03:51Thanks a lot.
Comments