FT : Why precision-guided chemotherapy has driven a dealmaking boom

Why precision-guided chemotherapy has driven a dealmaking boom
Weight-loss treatments are not the only source of excitement in the pharmaceutical research world

Investors’ interest in anti-obesity drugs sometimes seems all-consuming. Yet weight-loss treatments are not the only source of excitement in the world of pharmaceutical research.

A class of molecules known as antibody-drug conjugates (ADCs) is generating buzz. These use antibodies to deliver chemicals directly to tumours, making them the guided missiles of the cancer world. Invented decades ago, they have recently improved so much that developers such AstraZeneca talk about them replacing conventional chemotherapy. When the UK company and its Japanese partner Daiichi Sankyo revealed impressive breast cancer trial results for their ADC Enhertu in 2022, oncologists broke into a standing ovation.

Daiichi Sankyo is expected to be the ADC leader in 2028 with nearly $10bn of sales, according to Evaluate. It trades on a price/earnings multiple of 50 times, the kind of hefty multiple normally associated with obesity stocks. Novo Nordisk is on a 37 times multiple and Eli Lilly on 59 times. Some estimates of the long-term market potential invite parallels with obesity drugs. Morgan Stanley reckons the ADC market could eventually be worth more than $140bn.

That bullish estimate is based on a one-for-one switch from conventional chemotherapy, which accounts for more than 37 per cent of US cancer prescriptions. Healthcare funders might resist this on cost grounds; most chemotherapies have lost their patent protection and are relatively inexpensive. In April, access to Enhertu in England was blocked on value-for-money grounds although it got a green light in Scotland.

Investors should also worry that the market is getting overcrowded. Roche, Johnson & Johnson and Denmark’s Genmab have all inked deals this year, following nearly $100bn of ADC-focused M&A and partnership transactions last year. This included a Merck agreement with Daiichi Sankyo worth up to $22bn, Pfizer’s $43bn acquisition of lossmaking Seagen and AbbVie’s $10.1bn acquisition of ImmunoGen. The latter represented a 95 per cent premium to the undisturbed price.

The ImmunoGen deal came four decades after the company started working on the first generation of ADCs. Some of the current hopes for ADCs — which includes using them to treat bacterial infections and autoimmune diseases — could fail or take a long time to come to fruition.


The money pouring into ADCs is a reminder of just how important oncology is to the pharmaceutical sector, even as obesity steals the limelight. It is the industry’s biggest source of sales, with projected 2028 revenues of $440bn, says data provider IQVIA. Drugs that precisely target cancerous cells look set to play a crucial role in treating the disease, unleashing another wave of oncology growth.

FT : Robinhood warns of SEC lawsuit threat over crypto business

Robinhood warns of SEC lawsuit threat over crypto business
Retail brokerage says ‘Wells notice’ comes after years of attempts to seek regulatory clarity

Robinhood has warned of an impending lawsuit from the Securities and Exchange Commission over its cryptocurrency business, in a sign that a US regulatory crackdown on digital assets continues.

The retail brokerage said in a filing on Monday that the SEC had over the weekend sent its crypto unit a so-called Wells notice, which warns a company that it faces legal action. SEC staff have made a “preliminary determination” to recommend enforcement action against the company, which could lead to civil litigation, monetary penalties and limits on business activities.

The SEC has taken a hard line on crypto enforcement, arguing that many tokens constitute securities and should be regulated as such.

Robinhood, founded in 2013, says it has a customer base of more than 23mn investors and $119bn in assets under custody. While also brokering equities and options, it offers commission-free trading of 15 cryptocurrencies in most US states, according to its most recent annual report. Cryptocurrencies accounted for $135mn of its $785mn in transaction-based revenues last year.

The company said Monday it had “made difficult choices not to list certain tokens or provide products, such as lending or staking,” to avoid fall foul of the SEC’s policy stance on cryptocurrencies.

“After years of good faith attempts to work with the SEC for regulatory clarity including our well-known attempt to ‘come in and register,’ we are disappointed that the agency has decided to issue a Wells notice related to our US crypto business,” said Dan Gallagher, Robinhood’s chief legal, compliance and corporate affairs officer.

“We firmly believe that the assets listed on our platform are not securities and we look forward to engaging with the SEC to make clear just how weak any case against Robinhood Crypto would be on both the facts and the law.”

The SEC said it “does not comment on the existence or non-existence of a possible investigation.”

The SEC’s efforts to exert control over the growing digital assets industry have intensified after the failure of FTX in 2022, which culminated this year with founder Sam Bankman-Fried’s 25-year prison sentence on fraud charges. The agency has also sued leading crypto platforms such as Coinbase, Binance and Kraken.

TD Cowen analyst Jaret Seiberg said in a policy note that the SEC’s warning to Robinhood should not come as a surprise and noted that the broker had little incentive to settle such a lawsuit given that the term of SEC chair Gary Gensler was due to expire within two years. His departure could lead to a change in the regulator’s position on cryptocurrency enforcement.

“We see this as consistent with the SEC’s approach of forcing crypto fights to the courts,” Seiberg said in the note. “We also see little reason for Robinhood to settle given the political and legal changes that are possible in the coming years.”

Robinhood’s stock price was unchanged on Monday and is up about 47 per cent since the start of 2024. The company is set to report its first quarter financial results on Wednesday.

The company has previously paid out large sums to resolve regulatory cases, including a $65mn settlement with the SEC in 2020 over charges that it had failed to provide its customers with the best prices for trades, and more than $70mn in penalties in 2021 from industry regulator Finra over alleged harm to customers.

TechCrunch : Quora CEO Adam D’Angelo talks about AI, chatbot platform Poe, and w

Quora CEO Adam D’Angelo talks about AI, chatbot platform Poe, and why OpenAI is not a competitor

Last November, Adam D’Angelo found himself at the epicenter of one of the biggest controversies in the tech industry. The board of OpenAI — the $80 billion startup leading the AI bandwagon — had abruptly booted its CEO, Sam Altman, only to reinstate him just days later. D’Angelo was on the board that dismissed Altman… and he was (and remains) on the board that brought him back in. In fact, he was the only person who kept his seat amidst the ensuing restructuring that saw a lot of the original board leave.

It was certainly a rocky time for OpenAI, but it was perhaps doubly so for D’Angelo, since the drama was playing out while his own company, Quora, was taking big steps towards AI.

Quora, the crowdsourced Q&A site D’Angelo co-founded and leads as CEO, had been building an AI platform of its own while also fundraising (a $75 million round that valued it at $425 million, per PitchBook). The company in February 2023 had launched Poe (short for Platform for Open Exploration), which lets users ask questions of and talk to a variety of chatbots, lets developers build their own bots, and offers a bot monetization program and marketplace similar to OpenAI’s GPT Store.

Quora’s core Q&A service was facing some big questions, too. Incumbent search engines like Google and Bing were beginning to use AI to produce more fluid results and answer questions, and with tools like ChatGPT and Perplexity being widely available, what could Quora do to secure a position as one of the top websites where people could get their questions answered? More crucially, does anyone actually want or need crowdsourced Q&A anymore?

For D’Angelo, those questions are intrinsic to his pursuit of AI, which he sees as an important tool that people can use to tap the Internet’s collective knowledge. An important, if understated, figure in tech for years, he’s been involved in efforts to tap the Internet’s store of knowledge for a long time — he was friends with Mark Zuckerberg in high school, where in 2002 the pair built a digital music suggestion service called Synapse that, according to this vintage piece from the Harvard Crimson, beat off acquisition offers from Microsoft and more. Later, he became CTO at Facebook when it was just starting out, and then eventually co-founded Quora.

All of that was seemingly a long road toward building AI tools for him, it appears. I recently caught up with D’Angelo about the challenges and opportunities in AI today, how to build and support a developer community, and what role humans can play when it comes to sharing and accessing knowledge. Here are a few highlights from our conversation:

Humans are better at answers than AI — for now
The hype around AI seems to be having less of an impact on the search for information than you might think. D’Angelo said that Quora is seeing record numbers of users despite the proliferation of AI tools — although he declined to update the 400 million monthly active users figure it disclosed last July.

Still, there is a bridge between what Quora set out to do and D’Angelo’s interest in AI. Recently, in a conversation with David George, a general partner at a16z, D’Angelo said he was drawn to social networking because he was actually interested in AI. The latter was hard to develop at that time, but he saw social networks as an alternative architecture for achieving the same idea: People, assembled in a social network, in his view, almost played the role of living, large information models, as they could provide news, entertainment and more to each other.

He worked on that concept when he was with Facebook, and later, founded Quora to distill the role social networks could play in answering questions. Now, AI is taking over that role.

“In the past, humans were substituted for AI to provide answers. You could ask a question like, ‘What is the capital of California?’ and humans would answer that on Quora. Now, you can use AI tools to get that answer,” he said.

But AI, at least in its current shape today, cannot provide answers to all the questions people can have. That, D’Angelo believes, helps people retain a lot of value.

“Quora has always been founded on the idea that humans have a lot of knowledge they have access to in their heads that’s not on the internet anywhere. And AI will not have access to any of that knowledge,” D’Angelo said.

He acknowledged that AI still has a hallucination problem, which makes it hard to rely on such answers, even if newer, more advanced models are slowly making progress in tackling that issue.

Supporting developers on Poe
Quora opened up Poe to all users last year after a few months of closed beta testing. Since then, the company has introduced tools to create and browse the bots on its marketplace.

The company’s pitch is that consumers get to use all the different kinds of models or bots on the platform. For developers, the allure lies in the possibility of reaching millions of users without having to worry about distribution across platforms. And developers can earn money on Poe in two ways: The first is through a referral when a user becomes a Poe premium subscriber via their bot; the second is by setting a per-message rate, so they get paid based on how often people use their bot.

In essence, Poe offers developers and users access to different large language models, but its functionality is similar to OpenAI’s ChatGPT and GPT Store.

But that means both platforms face some of the same challenges. They make it easy for anyone to create bots with prompts, which makes it hard for developers to stand out. D’Angelo told me that there are already a million bots on the platform, compared to 3 million custom GPTs on ChatGPT. For reference, it took Apple’s App Store more than five years to cross the million-app mark.

Both Poe and GPT Store also suffer from a ton of spam, similarly named bots, bots claiming to escape plagiarism, and even ones that flirt with copyright law. Poe has also released a feature that lets users chat with multiple bots in one conversation. All that noise makes it hard to choose a bot that will do the job well.

Despite these challenges, D’Angelo says that Quora wants to help developers earn sustainable money by improving bot discovery.

“One of our goals with developers is to be able to make a living [out of making AI bots] and cover their operational costs,” he said. “We have taken a big step forward with the pay-per-message feature, but we also want to help developers get distribution inside the platform as much as possible. So, we are working on improving our recommendation system so more people can find out about the bots.”

No ads on Poe just yet
Poe is growing steadily, but it is still a lot smaller than ChatGPT. Market intelligence firm SimilarWeb suggests Poe has 4 million monthly active users in the U.S. (iOS and Android) and 3.1 million monthly active users worldwide (Android only). Compare this to ChatGPT users, which now averages 100 million users a week.

D’Angelo said that the company will stay away from ads, instead relying on Poe’s $19.99 per month subscription product to generate revenue. That is in contrast to some of the other AI-powered tools on the market: Perplexity, Bing Search, and Search Generative Experience (SGE) by Google all feature ads.

Quora and D’Angelo declined to disclose revenue figures, but data from analytics firm Sensor Tower indicates that Poe users have spent $7.3 million on subscriptions since its launch, amounting to close to 40,000 paid users. In comparison, ChatGPT has more than 1 million paid subscribers, according to Sensor Tower.

More AI tools for Quora and Poe
Despite stating the importance of human answers, Quora is already experimenting with answers written by Poe. The site surfaces the AI-written answer to some questions with a link that lets you chat with Poe if you have further questions.

D’Angelo said that Quora had already deployed systems to rate different human answers. Now, it is applying techniques like asking users through a survey if an AI-generated answer is useful.

“My goal is for the AI-written answers to be fairly ranked and only to be above a human answer if they are more useful than the human answer,” he said.

D’Angelo also wants to avoid having Quora tagged as an “answer engine.”

“I think we never really saw Quora as an answer engine. That term kind of implies that there are AI-only answers. Quora is really about human knowledge, and we’ll have AI enhance it,” he said

Quora is also working on AI tools that users can use to write answers and hopes to release them soon. D’Angelo noted that one of the tools it is testing allows users to generate an image based on their answers.

The company is using AI in a few other ways, too. One involves trying to catch bots or users using automation to answer questions on Quora. D’Angelo didn’t share details about the project, saying that the company would give a heads-up to perpetrators who are trying to game the system.

A few outlets and users have recently pointed out that the answer quality on Quora has plummeted. To that, D’Angelo said people feel that the overall standard of answers has decreased because low-quality answers have more visibility. He said AI is helping the company determine the difference between different quality of answers, and the early results look promising.

On Quora’s relationship with OpenAI
D’Angelo declined to discuss any of the OpenAI drama — “I just can’t talk about any of this stuff,” he said. “I’m not here to represent OpenAI. I can just represent Quora.” But he did say that he doesn’t see OpenAI as a competitor, because the bigger startup has, well, bigger ambitions.

“There is some sense of overlap in terms of what users can do on the GPT Store and what they can do on Poe. But that’s minor in the grand scheme of things. OpenAI is working towards this big mission to build AGI [Artificial General Intelligence]. And at Quora, we are looking to make AI products available to the world — including OpenAI’s products.”

Quora also continues to be a “big customer” of OpenAI and D’Angelo expects more collaboration with the company than competition.

“We spend a lot of money as a customer with OpenAI, because OpenAI is the biggest source of models for Poe,” he added.

While D’Angelo did mention that Quora pays “tens of millions” to developers on Poe and companies whose models the platform uses, he didn’t explicitly detail how these payments compared to the payout to OpenAI.

Quora currently doesn’t have any data licensing deals with any of the major companies, and it is not thinking about building its own model either, D’Angelo told TechCrunch.

“We are not in a rush to license our data. We want to make sure our rights and users’ rights are respected. Right now, there is not a lot of clarity around how all of this (AI landscape) will play out. So right now, we are just waiting before taking any steps in this direction,” D’Angelo said.

The company’s also relatively fresh out of its last fundraise, so it is focused on building AI across the business and improving revenue growth on its existing products. He said that Quora will go public “at some point,” but that is not the focus right now.

L'Express : Le cri d’alarme d’Elisabeth Badinter face au déclin démographique :

Le cri d’alarme d’Elisabeth Badinter face au déclin démographique : "S’il continue à ce rythme…"
Idées. Education positive, GPA, RN… La philosophe, qui publie "Messieurs, encore un effort…", s’inquiète de la chute de la natalité qui pourrait entraîner un retour de bâton conservateur et religieux contre les femmes.


"Certains courants féministes me verront comme une nataliste enragée…" soupire-t-elle. Mais à 80 ans et après bien des combats, Elisabeth Badinter ne se soucie depuis longtemps plus du qu’en-dira-t-on. Messieurs, encore un effort… est une mise en garde dont la publication a été repoussée par la disparition de son mari Robert. La philosophe y alerte sur la chute de la natalité qui, craint-elle, pourrait se retourner contre les femmes à travers des politiques natalistes menées par des partis conservateurs et religieux, comme on le voit déjà en Hongrie, en Italie ou aux Etats-Unis. Pour l’icône du féminisme universaliste, c’est au contraire aux hommes d’alléger la charge qui pèse sur les mères actuelles, prises entre leur carrière, les tâches familiales et le culte de l’enfant roi.

Dans un long entretien accordé à L’Express, Elisabeth Badinter explique comment il est facile de culpabiliser les femmes, mais aussi pourquoi la question de l’IVG "pourrait coûter cher à Donald Trump". Elle réitère son engagement en faveur de la GPA, qui fait aujourd’hui l’objet d’une attaque des médias conservateurs. Aux électeurs tentés par le vote RN, elle rappelle "les origines détestables" du parti, mais estime que Gabriel Attal répond mieux qu’Emmanuel Macron au "désir d’ordre" exprimé par une partie du pays.

L’Express : Votre livre évoque le déclin démographique auquel font face de nombreux pays. Mais la baisse de la natalité n’est-elle pas une bonne nouvelle pour les femmes, en sachant que leur accès à l’éducation en est l’un des facteurs importants ?

Elisabeth Badinter : Je ne dirais pas que la dénatalité est une "bonne nouvelle", mais la conséquence de bien d’autres facteurs. Il est vrai que quand on fait des études prolongées, on engendre plus tard. L’âge moyen du premier enfant est passé de 27,8 ans en 2000 à 31 ans aujourd’hui. D’autant plus que les filles sont plus nombreuses à sortir de l’université avec un diplôme de l’enseignement supérieur, 55 % contre 45 % de garçons. Mais j’en suis convaincue, ce déclin démographique représente un vrai problème politique s’il continue à ce rythme encore plusieurs années.

Vous craignez les conséquences possibles sur les femmes de politiques natalistes menées par des partis conservateurs ou religieux…

Absolument. Regardez ce qui se passe en ce moment. La cheffe du gouvernement italien, Giorgia Meloni vient de rajouter un obstacle psychologique à l’avortement, en autorisant des groupes anti-IVG à s’introduire dans les cliniques pour tenter de dissuader les femmes d’avorter. Or il est très facile de culpabiliser les femmes là-dessus. Meloni n’a certes pas supprimé le droit à l’avortement voté en Italie en 1978, mais dans les faits, il est devenu difficile de trouver un médecin prêt à faire cette procédure, en particulier dans le sud de l’Italie.

La Floride vient d’interdire toute IVG après six semaines, contre quinze semaines auparavant…

Seulement six semaines pour avorter (elle soupire)… Déjà, il faut souvent trois semaines à une femme pour qu’elle se rende compte qu’elle est enceinte. Qui pouvait imaginer qu’une grande démocratie comme les Etats-Unis mettrait fin, il y a à peine deux ans, au droit d’avorter dans quatorze Etats ? En même temps, le sujet de l’IVG peut coûter cher à Donald Trump. L’Ohio, Etat remporté par les Républicains en 2016 comme en 2020, a par exemple adopté par référendum la protection du droit à l’avortement dans sa Constitution.

La France a longtemps bien résisté en matière de natalité, mais celle-ci dégringole aujourd’hui…

Nous sommes quand même encore les meilleurs d’Europe en la matière. La France a une longue tradition qu’elle n’a partagée qu’en partie avec l’Angleterre, à savoir la séparation du rôle de la femme et de la mère. Il était entendu que le soin des enfants n’était pas digne d’une femme de l’aristocratie ou de la bonne bourgeoisie, qu’elles avaient mieux à faire que de s’occuper de la "marmaille" et que personne n’y trouvait à redire. Au contraire. Cette idée de la maternité en France était très originale. A l’inverse, les pays qui ont longtemps confondu femme et mère, comme les pays du sud de l’Europe ont vu une baisse plus précoce de leur natalité. La mamma italienne, et même la Mutter allemande sont des modèles caricaturaux de la mère. Toute l’essence d’une femme serait dans la maternité. Heureusement, les choses ont bien changé, car il y a eu un mouvement féministe efficace.

Mais comment expliquez-vous cette baisse de la natalité ? Même les pays scandinaves, les plus égalitaires, ont des taux historiquement bas…

Deux changements objectifs qui se rencontrent et s’opposent bouleversent notre société occidentale. Le premier est le "moi d’abord" qui va de pair avec la recherche de son épanouissement personnel. Le second est "l’enfant d’abord" ! Si on décide d’avoir un enfant, on lui doit tout, son énergie, son temps et son infinie patience. Le stress, la fatigue quotidienne des mères qui travaillent à temps complet doivent totalement s’oublier devant l’enfant roi. Dès la conception, l’embryon, est déjà le patron.

LIRE AUSSI : Robert Plomin : "Même échangé par erreur à la maternité, vous seriez à peu près la même personne"
Vous avez confessé avoir fumé durant vos grossesses. Aujourd’hui, c’est presque criminel…

Tout ce que les neurologues et les pédiatres énoncent comme loi, il faut s’y soumettre. Ils ont sûrement raison sur un tas de points, mais il n’y a plus aucun choix laissé à la mère. Quand on ne peut plus boire un seul verre de vin le soir, ou fumer une cigarette le matin… Et je ne vous parle pas de la culpabilité… Comme elles sont loin les années 1970 où l’on pouvait vivre sa grossesse avec insouciance.

Et qu’en est-il des pères ?

Les statistiques témoignent qu’il n’y a toujours pas un véritable partage des tâches familiales entre hommes et femmes, et c’est l’une des causes premières du désengagement des femmes en matière de maternité. Ce n’est pas au politique d’imposer des lois pour le partage de la vaisselle. Ce serait grotesque. On est dans la sphère du privé, sur laquelle les pressions sont difficiles. Pourquoi continue-t-on à parler de double journée de travail pour les femmes, et pas pour les hommes ?

Si vous regardez la vie quotidienne d’une mère d’enfants encore petits, et qui a un emploi à temps plein, très franchement, c’est dur. Pourquoi ? Parce qu’il faut penser à une infinité de détails de la vie quotidienne concernant bien sûr les enfants, mais aussi tout le reste, la gestion de la maison au quotidien. Le vaccin pour l’un, un médicament pour l’autre, les vêtements qui sont trop petits… Tous les jours, il faut commencer par s’occuper de ces détails qui rendent la vie très stressante pour les mères.

Mais avec l’attention portée sur la charge mentale, n’y a-t-il aujourd’hui pas une évolution du côté des hommes ?

Au moment où le féminisme des années 1970, universaliste, était dominant, il y a eu des vrais changements dans les comportements des hommes et des femmes. Dans le domaine public et dans l’accès à l’emploi, des progrès spectaculaires ont été réalisés. Entre 1975 et 2021, le taux d’activité des femmes est passé de 54,5 % à 70 %, proche de celui des hommes qui est de 76 %. Mais cela ne s’est répercuté dans la vie privée. On travaille comme vous, mais vous ne vous impliquez pas comme nous à la maison. Même si les hommes s’investissent aujourd’hui davantage dans l’éducation des enfants (vingt minutes de plus par jour en moyenne nationale que dans les années 1980), l’écart reste trop important. Quand les hommes répliquent à leur femme "mais tu ne m’as pas demandé !"…

Ce manque de partage, qui soulagerait les femmes, est une raison secrète de la chute de la natalité. D’autant plus que l’épanouissement personnel est devenu une revendication forte des hommes et des femmes. Aujourd’hui, on évalue ce qu’un enfant offre en plaisirs et en peines. C’est une vraie révolution anthropologique. Jamais on n’aurait pensé ainsi avant les années 1970. D’autant plus que pendant des décennies régnait la grande illusion consistant à penser que la maternité était le couronnement de la féminité, et une assurance de bonheur affectif. On ne parlait alors jamais des charges.

Vous critiquez aussi l’éducation positive. Pourquoi ?

Car cette école de pensée suppose une grande patience, un oubli de soi et du temps pour expliquer à l’enfant la raison de ses émotions. Vous sortez d’une journée de travail, et si votre enfant fait un caprice ou une crise de colère, il ne faudrait surtout pas le gronder, car cela pourrait avoir des effets nocifs sur son cerveau. Qui peut sérieusement penser que les mères du XXIe siècle, déjà accablées de mille soucis, peuvent se payer le luxe d’expérimenter une nouvelle éducation qui n’a pas fait ses preuves ? Aujourd’hui, si on veut des enfants, il faudrait être une mère exemplaire. Or, franchement, les mères parfaites sont aussi rares que des Mozart.

Pour vous, les politiques natalistes doivent avant tout s’adresser aux hommes…

Jusqu’à présent, on s’est toujours adressé aux femmes. Mais elles font ce qu’elles peuvent compte tenu de notre nouvel environnement hédoniste. Les hommes ont encore de la marge. Les pères font certes des choses, mais pas forcément les plus ennuyeuses ou répétitives. Ils manquent trop souvent d’initiative. Un enfant a besoin d’un médicament ? C’est la mère qui va à la pharmacie.

Alors que près de 60 % des Français y sont favorables, les médias conservateurs ont lancé une offensive contre la GPA, en mettant notamment en avant la figure d’Olivia Maurel. Qu’en pensez-vous ?

Je me suis engagée il y a dix ans pour la GPA éthique. On me répond toujours qu’il y a un échange d’argent. Mais il est normal de rembourser tous les frais d’une femme qui va assumer une grossesse. Cela me semble évident. Là, le débat déborde à nouveau. Je pense que le concept de "marchandisation" du corps a été un slogan efficace pour maintenir l’interdiction de la GPA en France.

Les outrances de l’extrême gauche nous font-elles perdre de vue l’offensive de l’extrême droite ?

L’un ne va pas sans l’autre. A chaque fois que Mélenchon met une couche de plus dans la provocation, des électeurs qui ne situent pas au départ à l’extrême droite se disent "pourquoi ne pas essayer". Je suis frappée de voir le nombre de personnes expliquer que le RN a évolué, et qu’il ne représente plus un danger.

Même les retraités sont de plus en plus tentés par le vote RN. Comment l’interprétez-vous ?

C’est un désir d’ordre, une peur de la dissolution de nos valeurs que nos gouvernements ne parviennent pas à renforcer. Mais je remarque que Gabriel Attal parle plus d’ordre que Macron, ce qui lui vaut un degré d’appréciation plutôt bon pour un Premier ministre. Il est clair sur ces questions. Le problème, c’est que sa ministre de l’Université fait le contraire de ce qu’il annonce…

Que diriez-vous aux gens tentés par le vote RN ?

Le RN a des origines détestables. Méfiez-vous du chat qui dort et nous endort avec lui. Vous pouvez le regretter amèrement, et la France avec vous.

>>> Google's medical AI destroys GPT's benchmark and outperforms doctors

Google's medical AI destroys GPT's benchmark and outperforms doctors

Google Research and Google’s AI research lab, DeepMind, have detailed the impressive reach of Med-Gemini, a family of advanced AI models specialized in medicine. It's a huge advancement in clinical diagnostics with massive real-world potential.
Doctors treat a multitude of patients daily, with needs ranging from simple to very complex. To deliver effective care, they must be familiar with each patient’s health record and keep up-to-date with the newest procedures and treatments. And then there’s the all-important doctor-patient relationship, built on empathy, trust, and communication. For an AI to come close to emulating a real-world doctor, it needs to be able to do all of these things.

The intersection of AI and medicine has really taken off. In the last six months, New Atlas has reported on AI models that aid less experienced doctors in identifying the precursors of colon cancer, diagnose childhood autism from eye images, and predict in real-time whether a surgeon has removed all cancerous tissue during breast cancer surgery. But Med-Gemini is something else.

Google’s Gemini models are a new generation of multimodal AI models, meaning that they can process information from different modalities, including text, images, videos, and audio. The models are adept at language and conversation, understanding the diverse information they’re trained on, and what’s called ‘long-context reasoning,’ or reasoning from large amounts of data such as hours of video or tens of hours of audio.

Med-Gemini has all of the advantages of the foundational Gemini models but has fine-tuned them. The researchers tested these medicine-focused tweaks and included their results in the paper. There’s a lot in the 58-page paper; we’ve selected the most impressive bits.

Self-training and web search capabilities

Arriving at a diagnosis and formulating a treatment plan requires doctors to combine their own medical knowledge with a raft of other relevant information: patient symptoms, medical, surgical and social history, lab results and the results of other investigative tests, and the patient’s response to prior treatment. Treatments are a ‘movable feast,’ with existing ones being updated and new ones being introduced. All these things influence a doctor’s clinical reasoning.

That’s why, with Med-Gemini, Google included access to web-based searching to enable more advanced clinical reasoning. Like many medicine-focused large language models (LLMs), Med-Gemini was trained on MedQA, multiple-choice questions representative of US Medical License Exam (USMLE) questions designed to test medical knowledge and reasoning across diverse scenarios.

However, Google also developed two novel datasets for their model. The first, MedQA-R (Reasoning), extends MedQA with synthetically generated reasoning explanations called ‘Chain-of-Thoughts’ (CoTs). The second, MedQA-RS (Reasoning and Search), provides the model with instructions to use web search results as additional context to improve answer accuracy. If a medical question leads to an uncertain answer, the model is prompted to undertake a web search to obtain further information to resolve the uncertainty.
Med-Gemini was tested on 14 medical benchmarks and established a new state-of-the-art (SoTA) performance on 10, surpassing the GPT-4 model family on every benchmark where a comparison could be made. On the MedQA (USMLE) benchmark, Med-Gemini achieved 91.1% accuracy using its uncertainty-guided search strategy, outperforming Google’s previous medical LLM, Med-PaLM 2, by 4.5%.
On seven multimodal benchmarks, including the New England Journal of Medicine (NEJM) image challenge (images of challenging clinical cases from which a diagnosis is made from a list of 10), Med-Gemini performed better than GPT-4 by an average relative margin of 44.5%.
“While the results … are promising, significant further research is needed,” the researchers said. “For example, we haven’t considered restricting the search results to more authoritative medical sources, using multimodal search retrieval or performed analysis on accuracy and relevance of search results and the quality of the citations. Further, it remains to be seen if smaller LLMs can also be taught to make use of web search. We leave these explorations to future work.”

Retrieving specific information from lengthy electronic health records

Electronic health records (EHRs) can be long, but doctors need to be aware of what they contain. To complicate matters, they typically contain textual similarities (“diabetes mellitus” vs. “diabetic nephropathy”), misspellings, acronyms (“Rx” vs. “prescription”), and synonyms (“cerebrovascular accident” vs. “stroke”) – things that can pose a challenge to AI.
To test Med-Gemini’s ability to understand and reason from long-context medical information, the researchers ran a so-called ‘needle-in-a-haystack task’ using a large, publicly available database, the Medical Information Mart for Intensive Care or MIMIC-III, containing de-identified health data of patients admitted to intensive care.
The goal was for the model to retrieve the relevant mention of a rare and subtle medical condition, symptom, or procedure (the ‘needle’) over a large collection of clinical notes in the EHR (‘the haystack).
Two hundred examples were curated, and each example consisted of a collection of de-identified EHR notes from 44 ICU patients with long medical histories. They had to have the following criteria:

  • More than 100 medical notes, with the length of each example ranging from 200,000 to 700,000 words
  • In each example, the condition was only mentioned once
  • Each sample had a single condition of interest
There were two steps to the needle-in-a-haystack task. First, Med-Gemini had to retrieve all mentions of the specified medical problem from the extensive records. Second, the model had to evaluate the relevance of all mentions, categorize them, and conclude whether the patient had a history of that problem, providing clear reasoning for its decision.


Example of Med-Gemini's long-context capabilities
Saab et al.
Compared to the SoTA method, Med-Gemini performed well on the needle-in-a-haystack task. It rated 0.77 on precision compared to the SoTA method (0.85) and outdid the SoTA method on recall: 0.76 vs. 0.73.
“Perhaps the most notable aspect of Med-Gemini is the long-context processing capabilities because they open up new performance frontiers and novel, previously infeasible application possibilities for medical AI systems,” said the researchers. “This ‘needle-in-a-haystack’ retrieval task reflects a real-world challenge faced by clinicians and Med-Gemini-M 1.5’s performance demonstrates its potential to significantly reduce cognitive load and augment clinicians’ capabilities by efficiently extracting and analyzing information from vast amounts of patient data.”
For an easy-to-understand discussion of these key research points, and an update on the mud-slinging between Google and Microsoft, check out AI Explained’s video from 13:38 onwards.

New OpenAI Model 'Imminent' and AI Stakes Get Raised (plus Med Gemini, GPT 2 Chatbot and Scale AI)

Conversations with Med-Gemini

In a test of real-world usefulness, Med-Gemini was asked about an itchy skin lump by a patient user. After asking for an image, the model asked appropriate follow-up questions and correctly diagnosed the rare lesion, recommending what the user should do next.

Example of Med-Gemini's diagnostic dialogue in a dermatological setting
Saab et al.
Med-Gemini was also asked to interpret a chest X-ray for a physician while they were waiting for a formal radiologist’s report and formulate a plain English version of the report that could be provided to the patient.

Med-Gemini's diagnostic dialogue assistance in a radiological setting
Saab et al.

“The multimodal conversation capabilities of Med-Gemini-M 1.5 are promising given they are attained without any specific medical dialogue fine-tuning,” the researchers said. “Such capabilities allow for seamless and natural interactions between people, clinicians, and AI systems.”
However, the researchers recognize that further work is needed.
“This capability has significant potential for helpful real-world applications, including assisting clinicians and patients, but of course also entails highly significant risks,” they said. “While highlighting the potential for future research in this domain, we have not rigorously benchmarked capabilities for clinical conversation in this work as previously explored by others in dedicated research towards conversational diagnostic AI.”
Visions of the future

Where to from here? The researchers admit that there is much more work to be done, but the Med-Gemini model's initial capabilities are certainly promising. Importantly, they plan to incorporate responsible AI principles, including privacy and fairness, throughout the model development process.

“Privacy considerations in particular need to be rooted in existing healthcare policies and regulations governing and safeguarding patient information,” the researchers said. “Fairness is another area that may require attention, as there is a risk that AI systems in healthcare may unintentionally reflect or amplify historical biases and inequities, potentially leading to disparate model performance and harmful outcomes for marginalized groups.”
But, ultimately, Med-Gemini is seen as a tool for good.
“Large multimodal language models are ushering in a new era of possibilities for health and medicine,” the researchers said. “The capabilities demonstrated by Gemini and Med-Gemini suggest a significant leap forward in the depth and breadth of opportunities to accelerate biomedical discoveries and assist in healthcare delivery and experiences. However, it is paramount that advancements in model capabilities are accompanied by meticulous attention to the reliability and safety of these systems. By prioritizing both aspects, we can responsibly envision a future where the capabilities of AI systems are meaningful and safe accelerators of both scientific progress and care in medicine.”
The study can be accessed via the pre-print website arXiv.

(ZH) Where Unsold EVs Go To Die: Belgium's Ports Drowning Under Glut Of Chinese

Where Unsold EVs Go To Die: Belgium's Ports Drowning Under Glut Of Chinese Imports

Ten years ago this week, we posted one of out most viral stories, highlighting the over-capacity in the auto industry: "Where the World's Unsold Cars Go To Die," which highlighted the 'endgame' of automakers' 'channel stuffing' efforts to disguise the sudden lack of demand for all the exciting new models that they had forecast would boom to the moon...
And now, as MishTalk's Mike Shedlock reports, we are seeing similar pictures across Europe...

"Some are parked here for a year, sometimes more."
Due to China’s overcapacity in production – as it aims to capture a quarter of the European electric vehicle market – the ports of Antwerp and Zeebrugge are inundated.
You probably need to see it to appreciate the challenges the automobile industry faces in transitioning to electricity. You also need to come here to understand how the Chinese industry’s overcapacity has flooded the European market. That morning, as the sun unexpectedly lit up the maze of highways leading to this remote arm of the port of Antwerp, Belgium, a huge cargo ship from the Norwegian company Höegh Autoliners unloaded thousands of cars at one of the terminals of International Car Operators (ICO), a subsidiary of the Japanese group Nippon Yusen Kaisha.
Alongside Swedish-Norwegian Wallenius Wilhelmsen, it is one of the main operators of the now merged port of Antwerp-Bruges, the world’s largest automotive terminal, through which the production of some 40 brands used to transit. But that was before the emergence of their Chinese competitors.

Car Parks
Imported vehicles are seriously piling up at European ports, turning them into “car parks.” Automakers are distributors are struggling with a slowdown in car sales as well as logistical bottlenecks that make it hard to alleviate the buildup of new, unsold vehicles.
Some Chinese brand EVs had been sitting in European ports for up to 18 months, while some ports had asked importers to provide proof of onward transport, according to industry executives. One car logistics expert said many of the unloaded vehicles were simply staying in the ports until they were sold to distributors or end users.
“It’s chaos,” said another person who had been briefed on the situation.
This is another part of the escalating trade war between China and the rest of the world.

China Produces 55 Percent of All Steel, Biden and Trump Eye Tariffs

China keeps returning to a well that has run dry, using exports as a means for growth. China is about to hit a brick wall, with global consequences.
My #1 issue looking ahead to 2025 is a global trade war with serious repercussions.

FT : French tech group Atos says Daniel Křetínský and Onepoint make bailout offe

French tech group Atos says Daniel Křetínský and Onepoint make bailout offers
Indebted IT services group is trying to agree a restructuring this month

French tech group Atos said Czech billionaire Daniel Křetínský and David Layani’s Onepoint have made offers to bail out the heavily indebted group as it races to strike a restructuring deal this month.

The Paris-listed group, which faces €3.65bn in debt repayments by the end of next year, said on Monday that it had received offers from four parties.

Křetínský, who had previously failed in an attempt to buy the oldest part of Atos focused on managing IT systems, has teamed up with London-based hedge fund Attestor on its new offer.

Onepoint, the largest shareholder in Atos, put forward a separate proposal backed by investment firm Butler Industries. US private equity firm Bain also made an approach, though Atos decided not to pursue the discussions, saying the offer failed to meet its objectives. The fourth offer came from a consortium of the IT company’s bondholders and bank creditors.

“Any solution will probably involve radical changes to the capital structure of the company and a significant issuance of new equity securities which will result in massive dilution of existing shareholders,” Atos said in a statement.

The group’s plight has deteriorated this year, with credit agencies repeatedly cutting its rating. It has churned through executives while its share price has collapsed more than 90 per cent in the past three years, giving it a market value of just €250mn. Its shares jumped as much as 12 per cent in early trading on Monday in Paris before giving up the gains.

In a sign of the worsening picture, Atos has recently upped its projected funding needs to stay afloat through 2025 to €1.7bn. The company said on Monday that a €100mn deal for short-term financing with bondholders had been agreed while talks with banks and the French government to secure an additional €350mn are ongoing.

Talks with the French government that would see the latter take control of assets deemed to be of strategic importance are also continuing, Atos said. France’s finance minister Bruno Le Maire has previously said the government had made an offer to buy key assets for up to €1bn.

The government wants to purchase three parts of Atos: super calculators for quantum computing, which are used by the French army for the country’s nuclear weapons programme; secure communications tech also used by the military; and certain cyber security assets. The company is set to provide cyber security services for this summer’s Paris Olympics.

Until recently, the government had remained at arm’s length from the turmoil at Atos, which is chaired by former UniCredit boss Jean Pierre Mustier, even as multiple attempts to restructure the company and sell assets fell apart.

As the company’s financial situation has worsened, however, the government has stepped in to provide a short-term loan and create a “golden share” that could be applied to the company’s strategic assets, allowing ministers to block their potential sale.

WSJ : Vinci Begins CEO Succession Plan With Appointment of Pierre Anjolras as CO

Vinci Begins CEO Succession Plan With Appointment of Pierre Anjolras as COO
The appointment of a new COO is the first step in the implementation of the succession plan of CEO Huillard, whose term of office ends in 2025

Vinci DG 0.14%increase; green up pointing triangle has appointed company veteran Pierre Anjolras as chief operating officer, marking the first step in the process to name a successor for chief executive Xavier Huillard, whose term will end next year.

Anjolras joined the French construction and infrastructure company in 1999, it said Monday. He acquired extensive experience in motorway concessions, becoming COO of Cofiroute in 2004 before being named CEO of ASF in 2007. Cofiroute and ASF are Vinci Autoroutes networks.

In 2021, he was appointed as chairman of Vinci Construction.

The appointment is the first step in the implementation of the succession plan of Huillard, whose term of office will end in 2025 at the end of the annual general meeting.