Mathias Humbert (@MathiasOZ) header
Mathias Humbert (@MathiasOZ) profile picture

Associate professor @unil. @EPFL and @UCBerkeley alumnus. Views are my own. He/him

Lausanne, Suisse
Joined April 2011
Photo by Mathias Humbert (@MathiasOZ)Photo by Mathias Humbert (@MathiasOZ)Photo by Mathias Humbert (@MathiasOZ)Photo by Mathias Humbert (@MathiasOZ)Photo by Mathias Humbert (@MathiasOZ)Photo by Mathias Humbert (@MathiasOZ)Photo by Mathias Humbert (@MathiasOZ)Photo by Mathias Humbert (@MathiasOZ)Photo by Mathias Humbert (@MathiasOZ)Photo by Mathias Humbert (@MathiasOZ)
Mathias Humbert retweeted
La minute nécessaire : Le délit de faciès d’@EPhilippe_LH
2,182
489
7,489
1,248,051
Mathias Humbert retweeted
🔴️️️️️️ Discussion sans filtre avec François Lenglet 👉️️️️️️ youtube.com/watch?v=5qS3PZkp… Bon visionnage et n'hésitez pas à partager !
181
1,631
112
4,528
314,628
Mathias Humbert retweeted
Voici le choix qu'ont fait 3 chaînes dimanche. Les programmateurs se sont donc battus pour avoir le raciste en chef. Toutes savent la haine qu'il va cracher. Il est invité pour ça. Ne me dites plus jamais "c'est la faute à Cnews". Tout le paysage audiovisuel français a basculé.
462
2,021
103
5,412
100,609
Mathias Humbert retweeted
Violences du 17 octobre: les profils des mis en cause loin de l'ultragauche ou des pilleurs d'opportunité
473
726
190
2,320
448,087
Mathias Humbert retweeted
Le mouvement lycéen et étudiant signe le crépuscule du macronisme: les jeunes veulent se former et accéder aux études, et un pouvoir élitiste leur répond en réduisant de 20% le budget par étudiant entre 2017 et 2026 et en gérant la pénurie avec un algorithme opaque (Parcoursup)
241
1,940
128
3,737
111,802
Mathias Humbert retweeted
Mouvement des lycéens : au JT de France 2, de lundi à jeudi : 8% sur les revendications… et 75% sur les violences et les dégâts.
118
3,386
226
7,700
350,582
Mathias Humbert retweeted
This photo says everything about the explosion of extreme wealth and its influence on society. Such a concentration of wealth and power in a few hands is unprecedented and dangerous. 🧵
150
1,987
189
3,997
289,994
Mathias Humbert retweeted
Je ne sais rien des révélations de @Mediapart sur des écrits antisémites et orduriers prêtés à @J_Bardella. Je sais, pour l'avoir écrit dans ma thèse et dans mes travaux universitaires, depuis plus de trente ans (1992...), que je n'ai jamais tenu pour sérieux le storytelling autour de la prétendue "dédiabolisation" du @RNational_off. Il y a toujours de vrais fachos, au sein du parti, autour de ses dirigeants, parmi ses candidats, chaque élection le démontre. Pour en revenir à monsieur Bardella, si le président du parti a tenu ces propos, il y a plus de dix ans, l'excuse de minorité que j'entends déjà poindre, ici ou là, ne tiendra pas la route. J'ajoute que, généralement, lorsque @Mediapart commence un feuilleton, les séquences qui suivent le premier épisode sont rarement décevantes. À suivre...
154
678
16
2,607
122,729
Mathias Humbert retweeted
Cette photo mérite absolument tous les prix. Elle a même sa place au Louvre. Elle relate précisément l’inertie de l’individu face à l’effondrement. Regardez la composition. Un chef-d’œuvre involontaire. Au premier plan, le sable doré, les tongs abandonnées, la glacière. Le décor ordinaire du bonheur estival. Deux personnages assis, de dos, comme dans un tableau de Caspar David Friedrich. Sauf que Friedrich peignait des hommes contemplant le sublime. Ici, ils contemplent la catastrophe. Et ils ne bougent pas. Au centre, le parasol. Publicitaire, criard, dérisoire. Un pastis face à l’apocalypse. C’est presque trop parfait : la France qui prend l’apéro pendant que le monde brûle. Aucun scénariste n’aurait osé. Et puis le ciel. Les deux tiers de l’image. Cette muraille de fumée qui avale la lumière, ce dégradé d’ocre et de noir digne d’un Turner. La nature qui reprend la palette des maîtres pour peindre notre propre fin. La femme photographie l’incendie avec son téléphone. Mise en abyme totale : nous documentons notre disparition au lieu de l’empêcher. Cette image dit tout. Le déni, le confort, l’habitude. Nous sommes tous sur cette plage. Nous voyons tous le nuage. Et nous restons assis, parce que l’eau est bonne et qu’on a posé nos congés. Les historiens de l’art parleront un jour de cette photo comme on parle du Radeau de la Méduse. Sauf que cette fois, le naufrage, c’est nous qui le regardons depuis la plage. 💔🌍
901
8,412
559
31,880
3,578,506
Mathias Humbert retweeted
🔴 « Le numéro 1, c’est Jean-Luc Mélenchon ! De tous les candidats, c'est celui qui a la vision la plus complète en matière d’intelligence artificielle. Il a un projet de société et depuis longtemps ! » Sur BFM Business, l’éditorialiste Raphaël Legendre souligne la force et le haut niveau de préparation de Jean-Luc Mélenchon, notamment sur l’intelligence artificielle. « Le leader insoumis est aujourd’hui celui qui propose la vision la plus complète : un cloud public, des centres de données publics, une agence publique de l’intelligence artificielle et un domaine public des données. Depuis 2016, Jean-Luc Mélenchon parle du numérique, comme nouvelle frontière de l’humanité. »
117
788
20
2,201
56,366
Mathias Humbert retweeted
📢 NEW essay: the narrative that AI is replacing software engineers seems to be based on AI-washing of layoffs. Among the many lines of evidence: New York State requires firms to disclose which layoffs were due to AI. When there are legal consequences to lying, almost no company blames AI. The best data suggests that software engineer employment in the U.S. is still growing, though slightly more slowly compared to a no-AI counterfactual world. But even this doesn’t account for increased entry into entrepreneurship. Why haven’t AI agents replaced software engineers? Many kinds of knowledge work, including software development, can be seen as a “decide-execute-deliver sandwich”. AI compresses the middle but the other two layers resist automation in a way that will not be overcome by capability improvements alone. In 2026, the middle has already been squeezed and there isn’t much room for further compression, so there isn’t a future capability leap that will cause discontinuous impacts, either. The essay also has a detailed breakdown of the difference between vibe coding and agentic engineering. Conflating the two has unfortunately led to a lot of confusion. We acknowledge that there’s a lot of uncertainty about the future. But we think the unknowns are more about how we will define and carve up roles and who can adapt to the changing needs, and less about whether software engineering skills and judgment will remain in demand (they will). Given that the AI-capabilities-cause-mass-layoffs narrative seems to be false even in a sector with very few regulatory barriers, we think most other professions are likely to be even more cushioned. (The claim isn’t that automation never displaces jobs, but that there are many downstream links in the causal chain, and that we have collective agency to act on those downstream decision points to ensure better outcomes for workers.) This essay by me and @sayashk is the first in a series. Feedback is welcome! We hope that our unique vantage point (leading an AI agent evals research group + fluency with the labor economics literature + the "normal technology" framework for understanding AI impacts + being coders ourselves and well connected to the software engineering community) gives us a way to bring complementary perspectives together for a deep dive into how AI is transforming the profession, in a way that hasn't been done before. Read the essay here for lots of other details: normaltech.ai/p/why-ai-hasnt…
26
51
13
180
41,307
Mathias Humbert retweeted
El país que más invierte en inteligencia artificial del planeta acaba de sancionar una ley nacional para obligar a sus ciudades a abrir bibliotecas y volver a los libros de papel. China era el candidato perfecto para enterrar el papel: pone más de cien mil millones de dólares al año en inteligencia artificial y es de los que más la incluyen en los currículos educativos. Si algún país iba a declarar que el libro ya no hace falta, que para eso está la máquina, era este. Sin embargo, hizo exactamente lo contrario: en vez de mandar todo a la pantalla, está promoviendo el libro por ley. Desde el primero de febrero, rige una norma que obliga a cada gobierno local a poner dinero en bibliotecas, abrir espacios de lectura hasta en las zonas rurales y sostener una Semana Nacional de la Lectura. Desde ahora, es obligación del Estado. ¿Por qué un país que ya tiene la mejor tecnología se molesta en legislar la lectura? Porque separaron para qué sirve cada cosa. La inteligencia artificial te sirve para producir, competir, ir rápido: es la herramienta. El libro te entrena en lo que ninguna máquina te da: atención sostenida y criterio propio. Es la cabeza la que después decide qué hacer con esa herramienta. Como lo resume uno de sus investigadores: solo a través de la lectura se llega a un pensamiento profundo e independiente. El premier Li Qiang lo decretó dentro del mismo plan quinquenal donde está su apuesta de inteligencia artificial. Las dos cosas son estratégicas y van de la mano. Mientras tanto, Estados Unidos hace el camino inverso: batió su récord de libros sacados de las escuelas (casi veintitrés mil desde 2021) y sus chicos sacaron las peores notas de lectura en más de veinte años, con cuatro de cada diez de cuarto grado que no llegan ni al nivel básico. La inteligencia artificial la va a tener todo el mundo. La cabeza para saber qué hacer con ella, no. Usala para pensar CON vos y no POR vos.
141
1,784
176
4,627
323,235
Mathias Humbert retweeted
MIT's Nobel Prize-winning economist proved that AI is mathematically guaranteed to destroy human knowledge. They published a massive NBER paper modeling the long-term impact of AI on human cognition. And they found the most alarming conclusion in the AI literature so far. It’s called "Knowledge Collapse." Here is how human progress actually works. When you struggle to solve a complex problem, you generate two things: General knowledge about how the world works, and context-specific knowledge about your exact problem. Normally, humans acquire both at the same time. You do the hard work to solve your specific problem, and in the process, you learn a general principle. You share that principle. That is how human knowledge grows. Then comes Agentic AI. AI is incredibly good at giving you the exact, context-specific answer you need right now. It hands the solution to you on a silver platter. So you stop doing the hard work. And because you stop doing the work, you stop generating the "general knowledge" that society relies on. Acemoglu calls it the "knowledge-collapse equilibrium." When AI reaches a certain accuracy threshold, the incentive for humans to learn drops to zero. Nobody verifies. Nobody explores. Nobody discovers new fundamental truths. Society gets increasingly sophisticated automated outputs, while our actual capacity to generate new knowledge quietly erodes. But here is the most terrifying finding in the paper. Welfare is "non-monotone" to AI accuracy. That means as AI gets more accurate, society actually gets worse off.
128
529
88
1,349
105,386
Mathias Humbert retweeted
Denmark’s PM Mette Frederiksen: There is a total established link between power, political power, capital, tech giants and AI. And the purpose of this is undermining democracy. That's the goal because the people who are developing this stuff are against democracy. I met Sam Altman a couple of years ago. I spent hours with him and at the end of the discussion I said to him, "You have created a monster." And he said, "Yes, and now it's up to you to find out how to deal with it."
205
3,347
226
6,932
395,839
Mathias Humbert retweeted
Generative AI doesn’t run on magic. It runs on massive data pipelines built on privacy violations by design. Our new @Amnesty report exposes how big tech’s AI systems are powered by surveillance, data extraction, and abuse of people’s rights, at scale. We researched the models powering some of the most popular publicly available standalone generative AI tools, including GPT 3 by Open AI, Google’s Gemini, Meta’s Llama, DeepSeek and tools by Midjourney and Stable Diffusion. This is not innovation at any cost. It comes at a high price: our human rights. Read the report: amnesty.org/en/latest/news/2…
53
1,818
54
2,956
103,374
Mathias Humbert retweeted
Enfin, les JT s'emparent du scandale des ingérences israéliennes contre les candidats de La France insoumise aux élections municipales. L'enquête est ouverte. Elle doit aboutir : la démocratie doit être protégée et les puissances hostiles qui tentent de truquer nos élections doivent être durement sanctionnées.
203
3,431
54
12,517
173,888
Mathias Humbert retweeted
The study, which was used by EY consultants in Canada to market their cyber security business, used made-up data, mis-attributed citations and referenced a McKinsey report that doesn't exist. ft.trib.al/Ab4x6JN
51
460
105
1,558
218,061
Mathias Humbert retweeted
🚨BREAKING: Anthropic just published a study mapping exactly which jobs its own AI is replacing right now. The workers most at risk are not who anyone expected. They are older. They are more educated. They earn 47% more than average. And they are nearly four times more likely to hold a graduate degree than the workers AI is not touching. The argument is straightforward. Anthropic built a new metric called "observed exposure." Not what AI could theoretically do. What it is actually doing right now in professional settings, measured against millions of real Claude conversations from enterprise users. For computer and math workers, AI is theoretically capable of handling 94% of their tasks. It is currently handling 33% of them. For office and administrative roles, theoretical capability is 90%. Current observed usage is 40%. The gap between what AI can do and what it is already doing is enormous. The researchers are explicit about what comes next. As capabilities improve and adoption deepens, the red area grows to fill the blue. The demographic finding is what makes the paper uncomfortable. The most AI-exposed workers earn 47% more on average than the least exposed group. They are more likely to be female. They are more likely to be college educated. This is not a story about warehouse workers or truck drivers. It is a story about lawyers, financial analysts, market researchers, and software developers. The exact group whose education was supposed to insulate them. Computer programmers showed the highest observed AI exposure at 74.5%. Customer service representatives at 70.1%. Data entry keyers at 67.1%. Medical record specialists at 66.7%. Market research analysts and marketing specialists at 64.8%. These are not predictions. These are measurements of work that is already happening on AI platforms right now. Then there is the pipeline finding nobody is talking about loudly enough. Anthropic's researchers found a 14% decline in the job-finding rate for workers aged 22 to 25 in highly exposed occupations since ChatGPT launched. No comparable effect for workers over 25. Entry-level roles were never just jobs. They were the training ground where junior analysts became senior analysts, where junior lawyers learned how arguments hold together. If that layer disappears, nobody has answered the question of where the next generation of senior professionals comes from. The detail buried in the paper that most coverage missed: 30% of American workers have zero AI exposure at all. Cooks. Mechanics. Bartenders. Dishwashers. The technology reshaping professional careers is completely irrelevant to roughly a third of the workforce. The divide is no longer between high skill and low skill. It is between presence and absence. The company publishing this study is the same company selling the AI doing the replacing. Anthropic had every commercial incentive to soften these findings. They published them anyway. If you spent four years and $200,000 on a degree to land a white collar career, the company that builds Claude just confirmed your job is more exposed than the bartender pouring drinks at your graduation party. Source: Anthropic, "Labor market impacts of AI: A new measure and early evidence" PDF: anthropic.com/research/labor…
255
1,476
198
4,295
840,791
Mathias Humbert retweeted
🇫🇷 FRANCE JUST FIRED MICROSOFT. 2.5 million government devices. Off Windows. Onto Linux. This isn't a pilot. Every ministry has been ordered to submit a migration plan by autumn 2026. The OS, cloud infrastructure, AI platforms, collaboration tools. All of it. The minister's words: "We can no longer accept that our data, our infrastructure, and our strategic decisions depend on solutions whose rules, pricing, and risks we do not control." They already replaced Teams and Zoom with a homegrown video tool. Health data is moving off American cloud by end of year. And this isn't even new ground for France. Their national police force already moved 100,000 devices to a custom Linux build. Saves €2M a year in licensing. If this works, every EU government is watching. Microsoft has not commented.
51
64
12
333
25,629
Mélenchon et le journaliste qui le lance sur la valeur de son appartement. Pépite.
Y’a des interactions qu’on oubliera pas
227
1,135
42
9,613
456,890