- il y a 3 semaines
Fostering European Innovation A Discussion on Open Collaboration
Catégorie
🤖
TechnologieTranscription
00:00Bonjour à tous, bienvenue à ce panel.
00:04Je suis très heureux, très heureux et heureux.
00:07C'est un panel sur l'innovation européenne et l'opportunité de collaboration.
00:12Je suis Frédéric Bardot, je suis le président et co-founder de Saint-Plon,
00:15et je suis très heureux d'avoir ce panel et cette discussion avec de nombreux speakers
00:20qui sont vraiment à l'intérieur du sujet que nous allons discuter.
00:24So, let me introduce Laura Bononcini, qui est le policy director, Southern Europe, at Meta.
00:32Zoé Tournant, co-founder of Fringant, qui est part de l'accelerateur programme
00:37of Meta, Scaleway, Hugging Face, et Station F.
00:41Marwan Elfites, head of le start-up programme at Station F.
00:46Et, last but not least, Doreen Pernell, qui est le chef sales officer at Scaleway.
00:54So, we are going to talk about Europe, but before talking about Europe,
00:58you know that Paris and France are going to be the centre of the world in the coming months.
01:03You know, it's usually the case, but it will be more and more the case for the Olympics.
01:09So, I've been told that there is a kind of trendy place and trendy view on French people on AI
01:18in the moment.
01:19So, I would like to have your view on it.
01:22And, what is your perspective on the special place that France and French people have in the ecosystem of AI?
01:29And, maybe Laura, you can take the floor.
01:32Very happy to.
01:34Yes, definitely.
01:36I mean, we really believe that France and Paris particularly play a key role,
01:40and it's really a tradition, I think, a culture of innovation that you guys have always had in this fantastic
01:46country.
01:46I mean, you have the best experts, our very own Yann LeCun.
01:51Obviously, you have the best master's degree with MVA, with NS LeClaire.
01:56I mean, you do have really the competence.
01:58And, what we're really proud about at Meta, and at the time it was Facebook,
02:03is that we were actually the first tech company who actually understood this potential
02:07and decided with Yann LeCun to open FAIR in 2015.
02:12And, this was really an idea.
02:14Like, it was really important for a U.S. company, actually, to open this in France
02:19and to actually start building an ecosystem that would go from France towards outside of the world.
02:26And, it was really based on two key principles.
02:28Like, the first one being openness, and I think we're going to talk about this a little bit more.
02:32Like, open systems developed by FAIR, etc.
02:35And, the second being collaboration.
02:37So, it's interesting to see that everyone who is working at FAIR actually works together
02:41with other research institutes, with other universities.
02:48We are very proud that many people who were working for FAIR
02:52have then founded their own company, like Mistral, etc.
02:56So, it's really like an idea of creating an ecosystem.
02:59And, I mean, even the fact, and going even further than that,
03:03what we've been doing with a lot of colleagues here at Station F, with the AI Accelerator,
03:08but also with Saint-Plon, in the training of 30,000 people that we want to train by the end
03:14of the year,
03:14for the open public.
03:16Like, it's all about building the ecosystem, but it's very important that it's starting from France,
03:20and we're really proud about this.
03:22Thank you, Lara.
03:23Marwan, what's your view?
03:24You are at the art of this ecosystem?
03:27Well, we're lucky because France is a kind of AI-ready ecosystem.
03:31So, it's not just from today, it's from quite some time.
03:34And because, I mean, as you said, we have great schools.
03:37I mean, and great French executives who used to work for the best, the state-of-the-art in AI.
03:42So, if you look at Mistral, if you look at H, yesterday, the last fundraising,
03:47I mean, if you look at the founder's backgrounds, they come from, I mean, the best of the AI world,
03:53and now they are able to launch their own businesses.
03:56And also, we, I mean, we have great unicorns.
04:00If you look at DataQ, for example, now we have a lot of people coming from these startups
04:04who are launching their own startups, so they truly have the knowledge.
04:09And, you know, so we have, we're lucky at Station F, we have a bunch of programs.
04:12And there is one also, in addition to Meta, that we manage us directly.
04:18It's called the Founders Program.
04:19And in that program, the last cohort that we unveiled last week,
04:23like 60% of the startups we have, they are in Gen.AI.
04:26And what is super interesting is to, again, to look at where the founders come from.
04:31So, some of them come from HEC, of course.
04:34Some of them come from ENS.
04:36They come from Inria, Sovy, or Stanford.
04:40So, it's super interesting to see how strong some founders can be,
04:46and that's why they are so good in launching businesses.
04:51And also, AI.
04:52So, you know, Station F, we launched in 2017, June 2017.
04:56And, well, so far, like, the biggest alumni we have,
05:00the biggest unicorn is Hugging Face, with whom you work on the program.
05:05And Hugging Face, you know, they were at Station F in 2017 and 2018.
05:09So, that, I mean, that is like seven and six years ago.
05:13So, now it's one of the biggest.
05:14The biggest also, the first, I mean, the first success we had at Station F,
05:18after a few months, it was recast.ai.
05:21They got acquired by SAP.
05:23So, they were in the Microsoft program.
05:25So, I mean, there is a huge track record in AI.
05:28So, machine learning in the days, now Gen AI.
05:30And AI and climate tech are the two biggest topics we're seeing at Station F.
05:35So, two biggest verticals and the biggest, I mean,
05:38the startups are raising the most in the community today.
05:42Right.
05:42And I've been told also that there is some U.S. companies
05:45that go in France, like Poolside, to set up.
05:48So, it's Hugging Face go to the U.S., Poolside go to France.
05:52So, it's, yeah.
05:53And also, I mean, just last week.
05:55So, for example, last week, we have a startup from Station F that was born in 2021
06:00that got acquired by Samsung.
06:02Samsung Medicine.
06:03So, the medical unit of Samsung.
06:06Sonio.
06:06So, it's a femtech startup.
06:08So, a startup that is working in female technologies.
06:10It's an AI assistant for gynecologists.
06:13And they got acquired for $92 million.
06:16So, for a startup that is only three years old,
06:18I mean, that shows how big the AI potential is in France.
06:24Zoé, you co-founded a startup in France.
06:26You are at the heart of this ecosystem.
06:28What do you feel about the French touch,
06:30especially in AI, but in tech in general?
06:33Yeah.
06:33So, to give you my perspective as a startup founder,
06:36we truly feel that France is at the heart of the AI boom.
06:40First of all, at Frangon, AI is at the core of our product.
06:44And it was built internally by one of our co-founders,
06:47who is a PhD researcher from a CEO at Saclay.
06:51So, again, we're very lucky to have access to this kind of talent.
06:55He was trained by top experts.
06:57So, it's a real resource.
06:59And second of all, we really feel that there is a strong ecosystem,
07:02which is pushed by Station F.
07:05For example, just a month ago,
07:06we were having a dinner at Station F
07:08where we could meet Yann Lockian, Xavier Niel, and top AI leaders.
07:12So, we really feel that there is something going on here in the conversations.
07:18And also from other countries,
07:20because we're quite frequently proposed to meet some ministers
07:24or investments coming from abroad.
07:28And in our case, we had an American fund
07:31who invested in Frangon,
07:33and they were actively looking to invest in AI in France.
07:38So, we really feel this attract from internationally.
07:43Thank you.
07:43And we are looking forward to discover more about Frangon just after.
07:46Doreen, born in France, Scaleway,
07:49and scale in Europe, in the world,
07:51what is your view on the French touch and the French ecosystem?
07:54Yes.
07:55So, before that Scaleway was international in 52 countries and 65 cities,
08:00we were born in Paris, in France.
08:02So, we are really proud of it.
08:03I think it's really important, the ecosystem and the sovereignty.
08:08And sovereignty can mean a lot of things.
08:10And I think the most important thing is,
08:12where were you funded?
08:13How are your investors?
08:15How is your tech built as well?
08:18Sovereignty is about a lot of things.
08:20And it starts at the head and really like the source of everything
08:24is where you start.
08:26And we chose to start in Paris exactly for those reasons.
08:30We know it's the heart of the ecosystem of Europe,
08:32and soon maybe of the world, we hope so.
08:35And we know, and that's something that Marwan just said with H.
08:38H is a client of Scaleway.
08:41It's where the clusters are going to be run with our technology.
08:44We're really proud of it.
08:46And it shows that the best startups work as well with cloud providers
08:51that are scalers that will be in Paris and in Europe.
08:55So that's something really important, I think, for the ecosystem.
08:58And apart from sovereignty and also talent,
09:02what is the character of the French Touch, Laura?
09:05Do you have a global view on Southern Europe?
09:08I mean, I would say openness.
09:11I mean, going back to what I was saying before, right?
09:13I mean, what's working here and what we have been pushing for as well,
09:17and that's the reason why we started focusing on openness,
09:20is that the French tech has been open, has been extending all over the world.
09:25And this open approach is actually something that characterizes
09:28everything that we've been doing around AI for two main reasons, actually.
09:33The first one is that we believe, and it's linked to your question as well,
09:37that it's actually democratizing access.
09:39Like, if you open up these systems,
09:41you have more people who can benefit from them and build upon them
09:45without having to spend on infrastructure, et cetera.
09:47The second one is related to safety and security.
09:50It can be a little counterintuitive, but it's actually fundamental, right?
09:55And so we have been building open systems since the very beginning, actually.
10:01Like, PyTorch was open.
10:02We've seen all the instances of Lama that have been launched in the last few years
10:07have been downloaded by millions of people.
10:09So it's actually working, and people are really using them
10:12and developing based on these large language models.
10:16And the interesting thing as well is that it's not only businesses and startups,
10:22it's not only academia who is using these models,
10:25it's actually also public institutions.
10:27I mean, your government has developed Albert, for example,
10:30to give easier access and easier information to citizens,
10:35like using something that was developed upon our Lama models.
10:40And so this is fundamental.
10:42But the reason why we're doing this, and I want to emphasize this,
10:45because, of course, we're an American company.
10:47We talk like Americans.
10:49And so everyone tells us, ah, you're saying you're being nice.
10:51You want to be nice.
10:52You want to appear nice.
10:53Yes, we are a nice company and an ethical company.
10:55But we're doing that also because it's in our interest, right?
10:58Because, as I was saying before, developing open systems means making these systems more safe,
11:05because people are going to be able to identify potential issues that are in those systems
11:10and are going to help us make sure that we can solve them.
11:14They're going to be able to identify bias.
11:17We're also going to be able, and we are able,
11:19to benefit from competences that are outside of our company to improve our system.
11:23And then if you think about hardware as well, if many people use our system,
11:28our hardware costs go down as well, which is also an advantage.
11:32And finally, and this goes a little bit larger,
11:35I think it's a benefit that access to these technologies
11:39and the capacity to develop on open systems is not only in the hand of a few tech companies,
11:45but is actually open.
11:46And I think this is a characteristic that we learned here in France
11:49and that we will continue pushing.
11:51Lauren, do you want to jump in on this?
11:53Because there is a close link between security, safety, and openness in Scaleway.
11:58Yeah, I think that's really key.
12:00And being open in your technology means not having dependency.
12:05And if you're not dependent, you are free.
12:08And I think that's really important now in AI, in technology,
12:12to really understand the responsibility that we have.
12:15At Scaleway, we build everything from scratch with openness.
12:20Our technology is completely open, was in-house built,
12:24and we have our own labs of innovation.
12:26We created RISC-V internally.
12:29And it's really something we are very proud to share with our customers,
12:34our partners, some AI startups as well.
12:37And we want to make sure to make it available to all.
12:39That's why we became international.
12:41And something that we all share in common,
12:44it's to help others to understand this
12:47and educate on the importance of being on open source now.
12:51Yeah, okay.
12:52So let's deep dive and talk about the common thread that you have,
12:56the fourth of you,
12:58this accelerator program,
12:59and the track that you made together,
13:02and you're part of, Zoe.
13:04Marwan, can you tell us more about this specificity of the track?
13:08Well, just to, yeah, exactly to explain,
13:11I mean, how StationF was structured.
13:13So StationF, 1,000 startups divided into 30 programs.
13:16So if you have some, if some of you are entrepreneurs,
13:19I mean, if you want to come,
13:20you need to apply to one of the program,
13:22Meta, and the others.
13:23So you go to the website, you see the whole list.
13:25And from the first days, we knew that we needed us,
13:29I mean, the staff.
13:30We wanted to be surrounded by the best.
13:31And we wanted to develop a kind of open innovation approach
13:36because if you look at the 30 programs,
13:38one third are managed by corporates.
13:40And we started to design StationF
13:43in assessing what startups needed.
13:46And of course, they had different needs,
13:50including funding, visibility, but also go-to-market.
13:53And for us, the best way to serve these needs
13:56was to have corporate, to onboard corporate at StationF
13:58so they can work end-to-end with startups
14:01and serve this kind of go-to-market approach.
14:05So at StationF, we develop this kind of open innovation approach
14:09through programs.
14:11And every program is looking for something different
14:14or is going to offer something different.
14:17And I will let Meta explain what they offer.
14:20There is a tech approach that is so unique
14:22with the fact that FAIR is in Paris.
14:25So it's, I mean, it's amazing to be able to connect
14:28entrepreneurs with this research lab.
14:30And all the programs, they develop POCs,
14:33they develop collaboration,
14:35or you have some corporates sometimes
14:37that are here to learn about the state-of-the-art
14:40from startups.
14:42So many startups now became service providers
14:47to corporates at StationF through the programs.
14:50So it's something key.
14:52We wanted it to launch.
14:53And all the time, every year,
14:55we launch between two to five new startup programs.
14:57Some of them are managed by corporates
15:00and the best in their field.
15:01And no one can be better than Meta, for example,
15:04for AI and open source.
15:06You are the head of all the programs.
15:09Each program is very special.
15:10But what is so special about this open source AI starter program
15:15within the other program?
15:17Well, I mean, the way the program was designed.
15:20So, you know, Meta and then Facebook.
15:22Facebook was the first partner we announced back in the day.
15:26So they were the first to trust us
15:27when we started to work on the project.
15:31And Meta is really strong in making the program evolve
15:35based on the needs.
15:36So the program is kind of new.
15:38I mean, you launched a few months ago in January.
15:42What is unique in the program is first the approach.
15:45I mean, AI and open source.
15:48Then the fact that you guys,
15:50you are surrounded by many partners,
15:52Hugging Face, Scaleway, for example,
15:54who are working on the program.
15:56So HLM is bringing unique resources.
15:58So you have an ecosystem within an ecosystem that is brilliant.
16:03And again, the fact that FAIR is in Paris is also a great asset.
16:09Laura, can you jump in on it?
16:11What is so special about this program?
16:13I mean, I think it's exactly that.
16:14It's the possibility for us also to learn from all these people
16:17and to make sure that we make these programs more important
16:20and hopefully that we can also internationalize them at some point.
16:23Like, let's see what's going to happen.
16:25But the idea is to see whether we can also,
16:27like I'm in charge of Southern Europe,
16:29whether we can also do something that goes further than France at this point
16:33and make sure that France can actually influence the rest of Europe
16:36with these activities.
16:39But also, like, yeah, as I was saying,
16:40it's a great way for an international tech company such as us,
16:44who usually focuses a lot on what's happening in the U.S., etc.,
16:48to actually have the opportunity with our engineers, with our experts, etc.,
16:51to learn from what is happening on the ground
16:54and to also maybe develop our products and our tools
16:57in a way that is more adaptable to the local realities,
17:00which is key for us because we have to localize while being international.
17:04So there are some close links between FAIR and the program,
17:08between your tech and your models?
17:11So that's great.
17:11Absolutely.
17:12Yeah.
17:12And what is the part of Scaleway in this program?
17:17Can you elaborate on what was the rationale for Scaleway
17:20to be part of this program with the GinkFace and Meta?
17:23Yeah, of course.
17:24So on top of everything that was just said by Laura and Marwan,
17:28so we help those startups on their architecture
17:31and their cloud needs and AI needs.
17:33So we offer to them all our products
17:37and we ensure that they know how to use them,
17:39that they have as well the right momentum
17:42to ensure that when they are out of the program,
17:45they are ready as well to use it freely.
17:48So it's as well a bit of education.
17:50And on top of that, I think there is a relationship with each of the startups
17:54is to help them with relationships,
17:56not only with what we do,
17:58but a bit of marketing, events as well.
18:01We want to make sure that we help them grow
18:03and that's really important because it's a tough and competitive market.
18:07So we want to make sure that the program goes beyond
18:09only sharing products and solutions,
18:11but really having this human aspect is really important for us.
18:15Great.
18:16So Zoe, is everything that has been said true, actually, or not?
18:20Because you're the insider, you're part of the program, so let's go.
18:23No, I'll give you the feedback.
18:25To be fair, we are very proud to be part of this program,
18:29which was sponsored by Meta, Hugging Face, and Scaleway.
18:32Open source, of course, is key for us.
18:34So we're very happy that this is the thematic of the program.
18:38Since the beginning in the company, it was really a choice.
18:41Exactly, since the beginning,
18:43because our technology is based on open source foundation models.
18:47So, of course, this is the base of our technology.
18:51Then we add some fine-tuning and some proprietary structures on top.
18:56But, of course, open source helps us going much faster.
19:00It's also...
19:01Please, Zoe, explain us what is Frangon and what you're doing.
19:04So, basically, to help to explain you Frangon,
19:07we do a personalized size recommendation for fashion brands online.
19:12So you click on Find My Size.
19:14This asks you for your height, your weight,
19:17and then we ask you for a selfie.
19:19And with a selfie, we're able to basically describe what the rest of your body is.
19:25And so we have your morphology, then we match it with a garment,
19:29and we tell you the ideal size for every garment that you pick in a brand.
19:33So this helps you find the right size, not return later.
19:38So that's what we do.
19:40And open source is really a way to have access to solid
19:45and quality foundation models, as it was said.
19:48So we also try to contribute to open source frequently.
19:51So you are using the model that was well-developed at FAIR
19:55and that are open on the table, and you get the model and the help to...
20:00One of our foundation models is the one from Meta called Dino V2,
20:06which we use for image recognition.
20:09So it helps us a lot.
20:11And the architecture also, it's powered by...
20:15Yes, and the monitoring, as Doreen was saying.
20:18We have access to lots of advices from very high-caliber tech companies.
20:24So that's a real advantage.
20:27And I imagine you can have some relation also with other colleagues in the program.
20:31You have other startups.
20:32Can you just mention them, or...?
20:35Yeah, so basically we are five startups, and we're all in the same space.
20:39Yeah, it's a cohort.
20:40Exactly.
20:40Yeah, okay.
20:41The cohort is based on AI, so we have very similar issues in our daily lives.
20:47So at the coffee machine, we're able to talk about AI, commercial problems, a bit everything,
20:53and it is very important to have some peers with whom we can share struggles
21:00and move faster together.
21:02Yeah, I think emulation will be not only at the launch part, but every time.
21:08Every time.
21:09Marwan, something to add on the...
21:13That's the...
21:13Well, I mean, it's the biggest value of Station F and the programs to...
21:16So, you know, the startups you will find on campus.
21:18So the 1,000 ones are between the pre-seed and Series A.
21:22So, of course, you guys, you have common challenges, common successes.
21:26So this is the biggest value prop we sell, basically, at Station F.
21:30Like, your neighbors are going to be other entrepreneurs.
21:33They're going to share their successes.
21:35They're going to share also their approach on their tech stack.
21:39And, I mean, everyone has its own approach.
21:41So we try to develop it.
21:44And, you know, Station F, we have, like, three priorities.
21:47So AI, quantum, and climate tech.
21:49So when I say that, it means we try to build the best programs on these topics.
21:55And we try to have the best ecosystem moments.
21:58So, for example, on AI, so we have, like, two programs on AIs with their own approach, Microsoft and Meta.
22:05You mentioned, for example, a month ago, indeed, we did a big dinner gathering the best of the AI ecosystem.
22:13And we're lucky to have Yann Lequin that day.
22:15But that was exactly for this purpose.
22:18So you can meet other entrepreneurs.
22:21You can meet key tech providers that day.
22:25So you were able to move in the right direction.
22:28And what are the evolution that are expected for this program?
22:32There will be another batch or something like this?
22:34Laura?
22:34I can't share anything specific at the moment.
22:37But as I was saying before, I think our objective is to make this as international as possible.
22:43But, I mean, I'm from Italy, as you probably heard.
22:45And we are so far behind, unfortunately, you guys in terms of development.
22:51And so hopefully, maybe some of our countries can take advantage of this as a model, but also as a
22:57potential integration.
22:58So stay tuned for more.
23:00Okay.
23:01And you have some programs that are equivalent, either in Southern Europe or in Italy or now?
23:07This is actually our best practice.
23:09I mean, we do have more programs, maybe a little bit similar to what I was talking about with Saint
23:15-Plon.
23:16Because another important part, I think, here is obviously to foster the ecosystem and the startups, etc.
23:23But it's also to make sure that AI becomes something that is relevant for the wider public, right?
23:28Because there is still an approach, a generalized approach related to AI that is based on fear of misunderstanding.
23:36And so it is important that we manage to explain to people how AI is actually something that is going
23:42to help them in their everyday life for their work, for their communication, for their safety, for their medical needs,
23:50etc.
23:50And so from this point of view, it is important not only to communicate about it, as I mean, we're
23:57doing in this fantastic event, but also to foster programs where we provide basic education to people in the working
24:03sector, but also in schools or in universities.
24:06And so those are the kinds of programs that we're doing, obviously, in France, but also across Southern Europe and
24:12Europe more widely.
24:14Great.
24:14As we are investigating the European ecosystem, we cannot avoid the elephant in the room, no?
24:21So we have to talk about the intense debate that's going on regarding maybe the trade-off between more investment
24:31or less regulation in Europe.
24:34So coming from a startup, is it something that you are thinking about?
24:39Is it touching you?
24:40Is it a problem?
24:41Are you discussing this kind of thing?
24:43Or is it too early for you?
24:45No, of course, we're discussing these topics.
24:47I mean, having access to funds is extremely important for us to scale and go fast.
24:54And it's true that sometimes in Europe we have access to great resources, great talents, but sometimes the funding lags
25:01behind.
25:02So this is important to compete globally.
25:06Of course, regulation is also needed because our product is AI and for people to want to use AI every
25:13day, you need to feel safe.
25:15And so AI has to be regulated.
25:18But it's true that if it was more harmonized between countries, for us it would make our lives way more,
25:26way easier.
25:26So is your product AI act compliant or?
25:30It is.
25:31It is, luckily.
25:33Yeah.
25:35Dorine, what is your view on it?
25:37You are an international and European actor and French one.
25:42I think it's really important to kind of have the elephant in the room token now.
25:46So on our point of view at Scaleway, regulation is important for fairness.
25:52We want to have competitiveness that will be fair for everyone to ensure that all the service providers play with
25:59the same game and cards in their hands.
26:02So that's one thing.
26:03And regulations for the startups.
26:05But funds are the most important thing.
26:07So, of course, we have clients that are enterprise.
26:11They don't have those issues.
26:12But a lot of our clients are startups.
26:15Even though they are in our programs, they know at some point they're going to exit and be alumni.
26:20So we want to make sure that they are in a jungle that is easier for them in the day
26:26-to-day because it's the case.
26:28And having access to funds with no thresholds would be very much easier for them.
26:34And making sure that whether you are in France, in the U.S. and other countries, you play kind of
26:40in the same sports and in the same league will be much easier and much fair, I think, to the
26:45world.
26:46Marwan, is it something that you are talking or thinking about every day when you will?
26:51Not every day, but, I mean, it's interesting because for startups, they need to work in a trustworthy environment.
26:59And I guess with this kind of regulation, that will help some startups to sell to big groups that can
27:07be, for some of them, afraid.
27:08And if you have the EU AI Act stamp on your solution, I guess that will definitely help for your
27:15go-to-market.
27:16And maybe this regulation will be a standard as the GDPR.
27:21So wait and see.
27:22But, you know, the French president mentioned like two or three days ago that governance is one topic when we
27:28talk about AI.
27:29You have other topics, as you said.
27:30So you have funding, you have infrastructure, you have usage, you have talent.
27:34You work a lot on talent on your side.
27:36So that's really important to train talent.
27:39And, you know, on our side, so to give you some ideas on how we are trying to act on
27:43every pillar.
27:46We were lucky to work with Entrepreneur First at Station F.
27:49So it's one of the 30 programs I mentioned.
27:51So EF, what they do, you know, they select people.
27:55They don't select startup.
27:56They select people so they can meet together and then create their own company.
28:01So they have a ton of biotech and deep tech startups.
28:03And researchers are doing projects there.
28:06And they did an AI school at Station F where they selected 100 people with the best AI talent.
28:12So they can learn more on the startup ecosystem and maybe, for some of them, launch a startup.
28:18So, for example, they did many master classes with Poolside and the CTO and founder of Poolside with the founder
28:24of Mistral, Arthur Mensch, and so on.
28:26And other alumni from EF.
28:28So it's one aspect.
28:29In terms of infrastructure, so we're lucky also to see the announcement of a hover hold to the Microsoft GNI
28:37program at Station F.
28:38So they have another approach.
28:39It's not open source.
28:40But it was interesting to see the investments they announced for the French markets.
28:46And Station F is part of it.
28:48So, yeah, you have all the aspects of the regulation.
28:50And for the funding, so we try, we do our best to attract many international funds so our startups can
28:57be connected to the right investors.
29:00So, Laura, as a policy director, what is your view on it?
29:04You work in a worldwide company, but you operate in France since decades.
29:08So what's your view on the, what is so special?
29:11Are we missing something in Europe?
29:12That's, that's actually one of the things that we're thinking about and working on and actually also discussing with institutions.
29:20I mean, by the way, the French government, together with other governments, is a government that is working very much
29:26to foster innovation and recognizes the importance of AI.
29:29I mean, there are lots of studies that demonstrate how AI has a huge economic potential.
29:35I think it's a Goldman Sachs study from 2023 that says that it's expected to increase it within the next
29:4110 years, the GDP by 7%, which corresponds to $1.9 trillion.
29:49Like it's, it's not even possible to imagine how much money that is, right?
29:53So there is a huge potential.
29:56What we're seeing, though, is that Europe is not yet at the level that other big players like the U
30:04.S., for example, is.
30:06I mean, the, we invest around one-fifth of what the U.S. invests in terms of research and development.
30:14If you look at research on AI and investment on the AI, the U.S. invests 50 times more than
30:21we do.
30:21So we are a little bit behind, and this is an issue that, that needs to be addressed, needs to
30:28be addressed also through regulation from our point of view.
30:32Because one of the, the issues that, that we're still seeing, even though, again, there is a recognition by European
30:37institutions that this is an issue.
30:39And there is, obviously, and that there has been an attempt to adopt, like, unified regulation on innovation, especially in
30:45the last few years.
30:46But we're still seeing a lot of fragmentation, like, if I'm a startup in France and I want to launch
30:52my business outside of France, very, within Europe, I have to deal very often with 27 different kinds of patent
30:59systems, of IP systems, of distribution systems.
31:03And this is an issue if I compare myself to a startup that is born in the U.S., which
31:08has just one system.
31:10And so this is something that needs to be really addressed and that we believe is really important to, to
31:15deal with.
31:16Yeah, we, we hope, and we see, though, that this is something that European institutions are seeing.
31:23These are discussions that we're having.
31:26We hope that this is going to be something of a priority for the new commission.
31:33I mean, everyone knows we have EU elections coming up.
31:37There were reports by my compatriots, former prime ministers, Draghi and Leta, about how to make competitiveness in Europe restart.
31:47These reports also focus on aspects related to innovation.
31:50So I think there is a recognition of the fact that this is an issue.
31:54I think further than the recognition, we need to take action.
31:57And this is something that also, yeah, as public policy people, we're also working on very closely with European governments.
32:04Is there another view in Italy from this perspective on this topic or not?
32:09What is special to France?
32:12I think, I think it's a generalized view.
32:14And it's actually, it's very interesting to see that, for example, the Italian government has been focusing as well on,
32:22on openness, by the way,
32:23as has the French government as well.
32:25But, for example, when I was talking about, about open systems, instead of trying maybe to build their own infrastructures
32:34and their own language models,
32:36why not take advantage of existing open systems that have been developed by big companies, both from the US, but
32:43also obviously from Europe,
32:44and build services upon that instead of trying to rebuild everything from scratch, right?
32:50These are the kinds of conversations that we're having.
32:52So, how can Europe catch up?
32:54Well, potentially by using what already exists instead of trying to replicate what has been created by other companies or
33:02other institutions that maybe are more advanced.
33:05And based on that, we can build the real European innovation that is probably, like, that will go much further
33:11beyond.
33:11And so, those are the kinds of conversations that we're having also with my government, but also with governments all
33:16around Southern Europe,
33:17and that are really starting to resonate, which is very good news for us.
33:22Okay, great.
33:22Before ending this panel, Zolée, do you have any things to announce or what's coming next for Fringan?
33:29So, it's a little different from regulation topics, but for us, we're happy to announce that we successfully raised our
33:37second round, so a seed round.
33:39Thank you.
33:41Great.
33:41So, as I was mentioning before, we have two French funds and one American fund, so we're very excited also
33:48to have these new perspectives, and we are launching several clients on our end.
33:54So, recently, we launched Maj and Sandro, two major brands in the whole Europe, so 11 countries.
34:04That was great.
34:04So, good news.
34:05Congrats, congrats.
34:07Thank you for these amazing insights.
34:10So, thank you for your attention, and please give a round of applause for these wonderful speakers.
34:17Make the most of VivaTech, and see you in VivaTech.
34:20Thank you.
34:21Thank you.
34:22Thank you.
Commentaires