TheChek News Flow

Momentum has been slowing but the country boasts enviable technical skills and a burgeoning entrepreurial culture

Contrasting with the morose mood in much of France nowadays, optimism and ambition fill the air at the cavernous Station F business incubator in Paris. Since opening in 2017, the world’s biggest start-up campus has nurtured 7,000 businesses including two unicorns: the AI company Hugging Face, now based in the US, and the healthcare insurer Alan.

Talk with the founders of AI companies at Station F and it is hard to resist their enthusiasm for the technology’s potential and the attractions of France as a place to launch a company. Of the incubator’s 40 best-performing start-ups, 34 have AI at the core of their business. The rapid emergence of Mistral, the Paris-based AI start-up now valued at $6bn that has developed one of the world’s most impressive foundation models, has also given them much to cheer about. 

“Europe can create competitive AI models today,” Xavier Niel, the French investor in both Station F and Mistral, recently told the FT. “I think we can create big things with a few hundred million euros.”

Much is going right in France’s start-up world. The country’s education system trains an endless chain of talented engineers. Paris is vying with London as Europe’s top AI hotspot. France’s business culture has been transformed over the past two decades, making it acceptable, even fashionable, to become an entrepreneur. Venture capital is more readily available than ever before. In spite of his troubles elsewhere, President Emmanuel Macron has been an active champion of the sector.

Unlike most big US AI companies, French AI start-ups favour open-source models that encourage greater collaboration and broader access to the technology. That, they hope, will give them a competitive edge in applying AI to almost every sector of the economy.

But the question remains: can France’s vibrant tech sector overcome the political mess and economic uncertainty that are blighting the rest of the country?

Station F’s young start-up founders have few doubts. Historically, French entrepreneurs have been far more successful at building companies in the US than in France itself, but that is changing now, says Thomas Le Corre, chief executive of the edtech start-up Rakoono. He has studied at HEC business school in Paris and the University of California, Berkeley. “I strongly believe in European tech,” he says.

The country’s abundant technical skills are perfectly matched to the AI industry, making France a great place to build a tech business, adds Joel Belafa, chief executive of Biolevate, an AI-enabled therapeutic research company. “For a long time, France has built a culture of engineering,” he says. Similarly qualified engineers in the white-hot US market, he reckons, might cost five to eight times as much.

Still, the momentum in the French tech sector slowed last year, partly as a result of the political turmoil resulting from divisive parliamentary elections. Data from Sifted, the FT’s sister publication, showed that French start-ups raised just €3bn in the second half of 2024, down from €5.9bn in the first six months. The latest Global Startup Ecosystem Index ranks France as the eighth most successful start-up nation in the world, up from 12th in 2020 but still behind the UK, Sweden and Germany in Europe. 

No matter how much progress the French tech sector has made, the US still exercises a powerful gravitational pull. The Parisian AI start-up Pathway announced last month that it was moving its headquarters to the US to be closer to its biggest customers. “We need to be in the room where it happens — and it happens in the Bay Area,” Zuzanna Stamirowska, Pathway’s co-founder, said. 

Rumours swirl around Paris that Mistral will itself have to sell out to a giant US company if it wants the resources to become globally relevant, just as Britain’s DeepMind was bought by Google in 2014.

Unlike their competitors in the post-Brexit UK, France’s AI start-ups must contend with the higher regulatory burdens of the EU’s AI Act. But some entrepreneurs argue the legislation can help build trust and boost creativity. “This is not just negative for Europe. It can drive better innovation,” says Samuel Bismut, co-founder of Corma, a software licence management company.

Little can be achieved without such optimism and ambition. But having benefited from some helpful tailwinds over the past few years, the French tech sector is now facing stiffer headwinds. This year will test France’s entrepreneurial mettle as never before.

Other news:

Select index ETFs showing strength:

Other news:

Donald Trump’s ‘Maganomics’ will damage growth, economists tell FT polls

On both sides of Atlantic, most surveyed claim protectionist shift will overshadow benefits of other policies

Donald Trump’s vision to reshape the world’s largest economy through protectionist policies that put “America First” will damage growth, according to Financial Times economists’ polls that contrast with investors’ bullishness over the US president-elect’s plans.

Surveys of more than 220 economists in the US, UK and Eurozone on the economic impact of Trump’s return to the White House showed most respondents believed his protectionist shift would overshadow the benefits of other elements of what the president-elect has dubbed “Maganomics”.

Many economists in the US, who were polled jointly by the FT and the University of Chicago’s Booth School of Business, also believe a new Trump term will spur inflation and lead to more caution from the Federal Reserve on cutting interest rates.

“Trump’s policies can bring some growth in the short term, but this will be at the expense of a global slowdown which then will come back and hurt the US later on,” said Şebnem Kalemli-Özcan, a professor at Brown University who also sits on the New York Fed’s economic advisory panel. “His policies are also inflationary, both in the US and the rest of the world, hence we will be moving to a stagflationary world.”

However, most economists — including at the IMF, the OECD and the European Commission — forecast stronger growth in the US than in Europe in 2025.

The US economy has consistently outgrown its counterparts across the Atlantic since the coronavirus pandemic, expanding at an annualised rate of 2.8 per cent in the third quarter of last year.

Trump has yet to lay out a comprehensive economic policy prospectus, leaving analysts to base their outlooks on pledges and threats made on the campaign trail.

Those include plans to impose blanket tariffs of up to 20 per cent on all US imports, mass deportations of undocumented workers, slashing red tape and making tax cuts introduced in 2017 permanent.

Trump, a self-described “tariff man”, has a long-standing and deep-rooted belief that the US needs to close its trade deficit and boost homegrown production.

“The announced policies include substantial tariffs and deportations of immigrant workers,” said Janice Eberly, a former Obama administration senior US Treasury official now at Northwestern University. “Both tend to be inflationary and likely negative for growth.”

Overall, more than half of the 47 economists polled specifically on the US economy expect “some negative impact” from the Trump agenda, and another tenth forecast a “large negative impact”. On the other hand, a fifth of those surveyed expect a positive impact.

The gloom among economists contrasts with investors’ optimism over Trump’s second term.

The US S&P equity index surged in the weeks following Trump’s win, though it pared some of those gains in December after US rate-setters signalled they would make fewer rate cuts this year than previously anticipated.

In its best two-year run this century, the benchmark index ended 2024 up 23.3 per cent, following a similar gain in 2023.

Benjamin Bowler, a Bank of America strategist, said this week that Trump’s “laissez-faire economics, tax cuts and deregulation”, coupled with a potential “AI revolution”, meant the rally was likely to continue into 2025.

A separate survey by the FT showed that Eurozone economists were even more pessimistic about the impact of Trump policies in their region than those in the US, with 13 per cent of analysts saying they expected a large negative effect and another 72 per cent forecasting some negative repercussions.

For the Eurozone the main concern is about manufacturing production, especially in Germany, the region’s biggest economy. 

Martin Wolburg, senior economist at Generali Investments, highlighted the possibility of the country’s car industry being “especially targeted” by Trump.

Trump’s threat of a 60 per cent levy on China “could further challenge European industries,” said Christophe Boucher, chief investment officer at ABN Amro Investment Solutions, as it would raise the prospect of Beijing flooding the region with cheap products.

While the UK is seen as better insulated from tariffs, thanks to its large services sector, Alpesh Paleja, lead economist at the CBI, warned that the country would be exposed to the “second-round impact” should tariffs weigh on Eurozone growth.

In the UK, more than 56 per cent of almost 100 respondents expected some negative impact, with many speaking of the drag on sentiment from the prevailing climate of uncertainty ahead of Trump’s inauguration on January 20. Just over 10 per cent forecast some positive impact. 

“The Trump administration will be an ‘unpredictability machine’ which will dissuade business and households from taking long-term decisions with ease,” said Barret Kupelian, chief economist at PwC UK. “This will inevitably have an economic cost.”

Regulators, ethicists and nation-states will be all over robotics in the workplace from every conceivable angle

It has been a while since humanoids stepped out of the pages of sci-fi writer Philip K Dick. Sophia, the work of Hong Kong-based Hanson Robotics, debuted in 2016 and gained (Saudi) citizenship a year later. Tesla’s Elon Musk plans to have his humanoids manning production lines next year.

Viewed from the C-suite, there is a lot to like. Human robots do not get sick, require pensions or join unions. Health and safety departments need not fret about how much stooping and carrying they do. There are no pesky demands to work from home either. 

Strength, dexterity and a total lack of squeamishness make them useful across industries, from burger-flipping in eateries to bed-bathing in hospitals. Laser-sharp precision, not to mention an absence of toilet breaks, makes them perfect factory fodder.

All told, Citibank analysts reckon there could be 648mn humanoids by 2040. That is four times the size of today’s US workforce but still looks modest beside predictions from futuristic devotees such as tech entrepreneur David Holz, who puts the number at 1bn, and Musk, who concurs.

Payback is swift. Even on a $35,000 price tag and (US) minimum hourly wage of $7.25, upfront costs — based on a 96-hour working week — are recouped in under a year, says Citi. Raise that to a $16 hourly wage and it only takes a few months.

Like their flesh-and-blood peers, humanoids will learn as they go. Large language models, as used in AI, are the easy bit. Teaching behaviour is a newer science. But Toyota Research Institute, for instance, is championing large behaviour models. So far, it has used demonstrations to “teach” robots more than 500 skills.

What could go wrong? Red tape for starters. Regulators, ethicists and nation-states will be all over this from every conceivable angle — data privacy, labour and national security among them.

Costs could mount when downtime for charging and maintenance is factored in. Agility Robotics’ humanoids, already busy on Amazon shop floors, require one hour of charging for every two they work. The company is working to lift that ratio to one to four next year; its next version will operate on 1:10. Still, that is a lot of tea breaks. 

Maintenance costs are about 20 per cent of total cost annually. Upgrades are not simply software but will affect hardware too, such as sensors, mobility functions and weight-bearing capabilities. Lifespans in a nascent industry are hard to judge but will probably be closer to cars than humans, maybe about 10 to 15 years. 

Perhaps the biggest issue is application. Industry, healthcare and logistics all offer scope for robots able to sense, think and react — matching the barcode, say, for Do Androids Dream of Electric Sheep, and packing it in the appropriate-sized box up in a fulfilment centre. Less obvious, however, is whether these robots really need to be bipedal and designed with Hollywood rather than out-of-town warehouses in mind.

Chaque mois, nos dilemmes moraux intimes sont passés au crible. Entre ceux qui appellent au boycott des créations d’« hommes monstrueux » et ceux qui veulent « séparer l’œuvre de l’artiste », le malaise est profond lorsque l’on continue à aimer l’art de ceux qui nous indignent.

On pourrait raccrocher ce dilemme à l’actualité – celle, par exemple, qui a vu l’annulation de la projection du Dernier Tango à Paris (1972), de Bernardo Bertolucci, par La Cinémathèque, à la suite des protestations d’associations féministes, qui par ailleurs n’en demandaient pas tant (mais plus simplement une contextualisation du film, qui comporte une scène de sodomie simulée, tournée sans que l’actrice Maria Schneider en ait été prévenue). « Personne n’aurait pris satisfaction à voir ce film dans ces conditions-là. Je pense que même les cinéphiles n’ont pas nécessairement envie de se retrouver dans une ambiance de guerre civile. Personne n’a rien à y gagner », déclarait au Monde le président par intérim du Centre national du cinéma et de l’image animée, Olivier Henrard. Mais toutes les annulations, mises en garde, tous les procès et condamnations ne règlent pas – pas toujours – le malaise qui se pose quand on aime, encore, l’art de ceux qui nous révulsent moralement.

L’écrivain et rappeur Gaël Faye parlait le 4 décembre au micro du journaliste Mouloud Achour, qui lui posait la question : que faire de ces « idoles », R. Kelly, P. Diddy, Michael Jackson, ou plutôt que faire de leur musique, maintenant qu’on sait ce que l’on sait ? « On brûle nos idoles peu à peu, se désolait l’écrivain. On ne peut plus accepter que ces choses-là passent, se dire “ce n’est pas grave”. » Ce n’est pas de boycott dont Gaël Faye parlait, mais du rapport intime que l’on a avec un artiste : « Je n’arrive plus à écouter Michael Jackson. Et je le dis avec peine. Quelque chose s’est brisé en moi. » Pour lui, la question ne se pose pas comme un dilemme : c’est presque à regret qu’il n’arrive plus à aimer ce qu’il a aimé.

C’est justement l’amour qui est au centre de l’excellent essai de l’écrivaine et critique Claire Dederer Les Monstres. Séparer l’œuvre de l’artiste ? (Grasset, 2024), consacré à cette question. L’amour de l’art quand il se teinte de culpabilité, de dégoût, mais quand malgré tout il persiste. L’amour des films de Polanski, en ce qui concerne l’autrice américaine. Claire Dederer admet que, même après avoir appris les crimes commis par son réalisateur préféré, elle prit conscience qu’elle les aimait toujours autant. « Je voulais être une consommatrice vertueuse, une féministe dans les actes, mais, dans le même temps, je voulais être une citoyenne du monde des arts, l’opposé d’un philistin, écrit-elle dans son dernier ouvrage. Pour moi, la question, l’énigme, consistait à savoir comment répondre à ces deux injonctions similaires, en apparence contradictoires. »

Le récit suit les déambulations de l’autrice dans cet espace où se rencontrent une œuvre, son auteur et son public. On partage les allers-retours de Claire Dederer lorsqu’elle se demande où commence la monstruosité ; si abandonner son enfant pour créer – comme Doris Lessing ou Joni Mitchell – range ces femmes du côté des monstres ; si elle n’a pas adoré ces femmes précisément parce qu’elle les trouvait un peu monstrueuses. S’il peut exister un rapport « vertueux » à l’art. Si elle-même – ancienne alcoolique – n’a pas fait des choses un peu monstrueuses. Si elle n’aurait pas été meilleure autrice si elle l’avait été davantage. On y trouve de drôles d’idées. Celle de Martha Gellhorn, par exemple, journaliste, écrivaine et accessoirement troisième épouse de Hemingway, qui pensait plutôt qu’« un homme doit être un grand génie pour compenser le fait d’être aussi répugnant en tant qu’humain ». Que, autrement dit, c’est le monstre qui a besoin de se transformer en artiste, et non l’artiste qui peut tout se permettre.

« On ne décide pas la beauté. Elle nous frappe »

Et Claire Dederer se demande donc « que faire de l’art des hommes monstrueux » : « La première pensée qui nous vient à l’esprit est de boycotter cet art – la solution progressiste qui consiste à retirer notre argent et notre attention. » Mais nous ne devrions pas aborder ce dilemme en consommateurs, prévient-elle, ni exercer notre moralité par ce que nous choisissons de « consommer » ou non. En fait, « l’art que vous consommez ne fait de vous ni une mauvaise ni une bonne personne. Il faudra trouver une autre manière de vous accomplir ».

A une consommation « vertueuse » de l’art, elle oppose un autre paramètre : la beauté. Principe fragile, si on l’oppose à la moralité. « Et pourtant, la beauté compte. Et on ne décide pas la beauté. Elle nous frappe. »

En 1990, la dramaturge américaine Pearl Cleage publiait l’essai Mad at Miles (« en colère contre Miles », Cleage Group, non traduit). Elle y examine sa relation – car c’est cela qui nous lie à certains artistes – à la musique de Miles Davis, qui était très violent avec les femmes. Partant de sa propre expérience de victime de violences, elle s’interroge : « Est-ce qu’on peut faire l’amour en écoutant un “vieux disque de Miles”, alors que le jour où il l’a enregistré, il a peut-être passé sa matinée à mettre des baffes dans la gueule de l’une de nos sœurs ? » L’amour de Pearl Cleage est personnel, sa haine l’est également.

Il se passe avec ces œuvres que l’on aime avec tourment la même chose qu’avec les gens que l’on continue d’aimer, même si parfois ils nous blessent. « “Le problème, c’est que vous l’aimez toujours.” Cette phrase décrit des relations si fréquentes avec notre famille, notre partenaire dans la vie, parfois même avec nos enfants, écrit Claire Dederer. Que faire des gens terribles que nous aimons ? La plupart du temps, nous continuons à les aimer tout de même. » Prétendre que cet amour n’existe pas ou qu’il ne devrait pas exister, cela n’aide en rien. Aimez ce que vous aimez, nous dit-elle. Cela n’excuse personne.