CEO @TelehouseFR, father, husband, impatient optimist. Opinions here are my own.
Paris, France
Joined December 2008
- Tweets2.7K
- Following1.3K
- Followers831
- Likes7.3K
Sami Slim retweeted
Hugo Duminil-Copin (médaille Fields 2022) : la résolution de 350 problèmes majeurs par OpenAI, c'est comme si des camions m'avaient écrasé ; tous les problèmes (et donc pistes de recherche) que je mentionnais dans mes exposés, articles et demandes de financement, ont été résolus.
Sami Slim retweeted
The state of the world in one chart.
🇪🇺 Europe represses wages and exports its savings: that's the "Europoor" phenomenon.
🇺🇸 America turns European savings into soaring tech valuations: that's the AI bubble.
🇨🇳 China builds the factories and floods the world with cheap goods: that's involution.
And each loop feeds the next.
More here: ardenletter.com/p/the-wests-…
Sami Slim retweeted
Quelques points sur Mistral pas encore lus dans le débat public :
1. mistral large 4 est exactement aussi bon que ce que les Américains arrivent à produire à puissance de calcul égale. Cela signifie que l’écart est désormais sur de la taille de centre de calcul, probablement pas sur la compétence.
2. Étant données les prévisions de montée en charge de Mistral dans le calcul, ils devraient pouvoir produire un modèle de niveau Opus 5.5 dans 12 à 18 mois.
3. Cela peut paraître lointain, mais il faut bien comprendre à quoi ça sert d’avoir des modèles français. Essentiellement, au-delà de la question que tout le monde connaît de sécurité des données et de dépendance étrangère pour le fonctionnement de ses entreprises, à deux autres choses : 1. La souveraineté régalienne. 2. La balance des paiements.
La souveraineté régalienne aujourd’hui se joue avant tout sur la cybersecurité : la capacité à sécuriser son infrastructure militaire et civile de façon autonome pour ne pas être vulnérable aux IA étrangères.
Surprise surprise, Mistral Large 4 est particulièrement bon dans ce domaine. Ce n’est bien sûr pas un hasard.
La balance des paiements : le grand risque qui nous pend au nez, c’est que les entreprises françaises dépensent en masse en tokens américains, et souvent au détriment des salaires français, ce qui a pour conséquence que l’argent français s’enfuie de l’autre côté de l’Atlantique, vidant notre pays de ses richesses.
Or pour des usages entreprises, la course à la frontière n’est pas infinie. Quiconque a déjà utilisé opus 5.5 avec son harnais cowork sait qu’il y a déjà de quoi faire fonctionner l’essentiel des postes de l’économie de services. En d’autres termes nous arrivons en haut de la sigmoïde : les progrès subséquents des modèles frontières seront clés pour le progrès scientifique, pour des usages très avancés, mais pas forcément pour suppléer les fonctions marketing et comptabilité des boîtes du cac40.
Nous sommes donc à 18 mois d’avoir un Mistral qui peut faire le gros de ce dont a besoin l’économie française. Et ainsi stopper l’hémorragie de cash.
4. Le goulet d’étranglement est de plus en plus le harnais, la couche d’orchestration et d’exécution qui permet au modèle d’évoluer dans des environnements complexes, de se connecter à des outils, de gérer de multiples agents en parallèle, etc. Or il se trouve que nous avons en France aussi Dust, qui excelle sur cette couche. (Encore un truc de X d’ailleurs. Avez-vous pensé à remercier Napoléon et ses blocs de granit récemment ? Parce que de la même façon que nous vivons encore 75 ans plus tard sur les efforts de De Gaulle avec la dissuasion et le parc nucléaire, nous vivons en bonne partie sur de l’héritage TECHNOLOGIQUE napoléonien. Pour la Patrie les sciences et la gloire les amis.)
5. Pour continuer le parallèle. De la même façon que notre dissuasion nucléaire est toute petite par rapport à l’américaine ou la russe, mais que comme disait De Gaulle, on hésite à attaquer un pays qui peut tuer 80 millions de tes habitants en un instant, même si toi tu peux raser la France, la France est en train de construire une indépendance IA de taille réduite, mais rendant le service effectif minimum pour rester une véritable puissance. Et la France est le seul pays sur terre aux côtés de la Chine et des US à réussir cela. Au lieu de pleurnicher sur le fait que ML4 ne fait pas de jolies fusées en vectoriel, on devrait plutôt prendre du recul et se rendre compte de l’immensité de l’accomplissement.
6. Rien n’est joué et tout peut s’écrouler bien sûr. La survie d’une nation se joue à chaque instant.
Il faut encore monter en charge, construire masse de nucléaire et de centres de calcul, développer une industrie de puces d’inférence. Dans un contexte de chaos.
Mais merde. Même à terre, même après des décennies de déréliction, la France reste la France.
Les français ont toujours été bons dans l’optimisation sous contrainte. Le sens de l’histoire pointe vers cette direction. Nous avons une carte à jouer.
72% of the intelligence with 3.8% of the GPUs.
Just to show how talented @MistralAI team is: this shows the "intelligence per GPU" ratio.
- Mistral 4: 3,800 Grace Blackwell (published) 🇫🇷
- GPT-6 Astra: 100,000+ Grace Blackwell (published)
- Grok 4.7: ~200,000 GPUs (estimate)
- Claude Opus 5.5: ~500,000 Trainium2 (estimate)
Proud!
Sami Slim retweeted
What’s happening in the AI-powered reverse engineering and decompilation is absolutely insane. Pretty soon all software will be de facto open-source.
AI is coming for everything and everyone.
Sami Slim retweeted
L'IA ca sert à rien... 😅🤫
Une IA déchiffre une lettre napoléonienne vieille de 217 ans 3/6⬇️
Vingt-quatre lignes de signes mystérieux, une clé perdue et une énigme longtemps restée sans solution. Avec l’aide d’une intelligence artificielle, un ingénieur annonce avoir déchiffré une lettre destinée au général Marmont en 1809. Une plongée dans les secrets de l’Empire, menée en quelques heures de calcul !
Faire parler un document que plus personne ne savait lire : c’est l’exploit rapporté par Carter Church, un ingénieur en intelligence artificielle. Sans expérience préalable du déchiffrement, il s’est attaqué à une reproduction publiée en 1969 dans une revue d’histoire militaire. Son allié, GPT-6 Astra, n’aurait nécessité que six heures d’exécution, auxquelles se sont ajoutées plusieurs soirées de travail personnel.
Le défi dépassait la simple reconnaissance d’écriture. Il fallait identifier quelque 1 300 unités chiffrées, distinguer les variantes manuscrites et reconstruire leur signification. Pour brouiller les pistes, une lettre pouvait être représentée par plusieurs symboles, et certains signes remplaçaient même des mots entiers ! Une table partielle établie par le spécialiste de l’histoire des codes Daniel Tant a toutefois fourni les premiers repères.
La lecture proposée place la dépêche en mars 1809. L’Autriche se prépare à affronter Napoléon. Marmont, isolé en Dalmatie, doit connaître les positions des armées françaises et alliées. Le courrier les énumère pour lui donner confiance et l’encourager à agir.
Quelques semaines plus tard, le général entame sa remontée vers le nord. Il atteint Trieste, avant de combattre autour de Graz. Fin juin, il reçoit l’ordre de gagner Vienne à marches forcées. Ses troupes rejoignent la bataille de Wagram, remportée par Napoléon les 5 et 6 juillet. Au terme de cette campagne, Marmont devient maréchal d’Empire.
Quelques signes restent incertains, mais Carter Church met à disposition sa méthode et sa clé de déchiffrement pour permettre à d’autres d’en vérifier la lecture. L’exploit pourrait avoir une suite concrète : si une autre lettre utilisant le même code refait surface, les chercheurs disposent désormais d’un point de départ pour la lire. Après avoir protégé les secrets de l’Empire, ces symboles pourraient enfin livrer d’autres fragments de son histoire.
Cette percée laisse aussi entrevoir de nouvelles possibilités pour l’exploration des archives. Déchiffrer une écriture récalcitrante, retrouver un nom parmi des milliers de feuillets, relier des lettres conservées dans des fonds éloignés : l’IA pourrait accélérer des recherches qui exigent aujourd’hui un patient travail de dépouillement. Et ramener dans le champ des historiens des documents restés dans l’ombre, faute de temps pour les étudier…
Sami Slim retweeted
Et ils parlent de réindustrialisation…. Notre pays en mode suicidaire.
Et de cinq ! Mercredi 30 septembre, quatre recours supplémentaires ont été déposés devant la justice administrative contre le gigantesque projet de data center de Mistral à 50 milliards d’euros qui doit voir le jour en 2027 aux portes de Paris.
Cet équipement, d’une surface équivalente à 100 terrains de foot, consommera à terme la puissance électrique d’un réacteur nucléaire.
Un impact écologique atomique pour ce centre de données choyé par Macron.
l.lecanardenchaine.fr/tp
Sami Slim retweeted
Eric Schmidt: If I could invest in only 1 AI theme, it would be long reasoning.
A model that can stay on 1 problem for hours/days is a completely different thing from a model that answers and stops, and most people still have not priced that difference.
----
From "Blackstone and Eric Schmidt" YouTube channel, (link in comment)
Sami Slim retweeted
Refroidir tous les data centers IA de la planète consomme 2× moins d’eau que la ville de Paris.
14× moins que les pertes du réseau français.
163× moins que l’arrosage des pelouses aux États-Unis.
Dossier complet à lire : lel.media/12-idees-recues-su…
Sami Slim retweeted
Jesus christ
Some data we recently assembled on entrepreneurship/compute in Europe: eudata.vercel.app.
We hope that one of the useful roles that Stripe can play is in collecting and publishing empirical data pertaining to entrepreneurship and industry in Europe. There's growing appetite to get Europe on a better footing, and cross-sectional comparisons can often shine light on where opportunities lie. If you're interested in this kind of thing, we publish more at stripeeconomics.substack.com.
Sami Slim retweeted
zuck essentially launched a new app store for agents today and Meta’s perfectly placed to crush it, you’re looking at a multi-trillion dollar opp if they pull it off:
- instead of apps, agents get equipped with connectors i.e. plugins to any app, service or software tool.
- suddenly your agent goes from a useless chatbot to an action-oriented helper that gets shit done for you
but it gets even better - meta has collected ALL the data about you. they know what you want, when you want it which means they know exactly WHICH connectors to list in a marketplace and HOW MUCH to charge because they know customers will pay.
it’s a masterclass. muse isn’t just some generic agent, it’s sitting on a treasure trove of data and meta can build a trillion dollar business on top of it.
Opening access for developers to build Muse connectors. You bring the API -- Muse brings the agent, the browser, and the context of what the person actually wants. People reach your service just by asking for it, and their agent takes it from there.
New connectors are live today. Come build with us. muse.ai/platform
Sami Slim retweeted
Wow. So it's real. July 2024, ~2 years and 2 months ago, GPT-4o could fail 9.9 > 9.11. We are in a new reality.
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics.
The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra.
The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
Sami Slim retweeted
Regret the tone of my post on data centers yesterday.
What I should have said:
There were reasonable concerns about data centers 18ish months ago: water, taxes, jobs, electricity prices, the environment and what they would do to small towns. Well-structured data center projects have largely addressed these concerns today and we should be celebrating this.
On balance, data centers are awesome for America in every way.
On water: U.S. data centers use a fraction of what golf courses use. A lot of the numbers from 18 months ago were off by over 1000x. Newer data centers use closed-loop systems or recycled water. Should be required by every town approving a data center project.
On taxes: looking only at sales-tax exemptions, as Ronan Farrow did, is the wrong way to evaluate this. Data centers pay significant property taxes. Loudoun County, which is the wealthiest county in America, now collects on the order of $1 billion a year from data centers. In Quincy, WA, data centers are more than half the property-tax roll. Over time, property taxes can go to zero while government spending increases in these towns.
On jobs: this has been unambiguously awesome for blue collar Americans. Demand for electricians, plumbers, welders, HVAC techs, and contractors has gone vertical, and it is not a one-time construction job. These buildings get upgraded and expanded over time. That is why the building trades are fighting for them, and why some unions are now treating opposition to data centers as a reason not to endorse politicians.
On power: the original fear was that households would pay for the incremental electricity demand in the form of higher prices. That is why the ratepayer-protection deals and the new large-load tariffs exist. The right structure is: the data center brings or pays for new generation and signs a contract long enough that existing customers are protected. Where that is happening, utilities are cutting or freezing residential rates and saying so on the record. Where it is not, people are right to object. Electricity prices are going down *today* in a number of large states because of data centers.
On the environment: data centers overwhelming use natural gas today, which is the cleanest power source outside of nuclear, solar and wind. And the companies that are building the data centers are committed to carbon neutrality such that an equivalent amount of solar will likely be built. Maybe more importantly, the data centers need batteries to function effectively and these batteries can also sell energy back into the grid (which recently prevented blackouts in Texas). Over time, data centers will run on solar plus batteries.
On the towns: Poverty in Quincy, WA fell from 29% to 6%. Data center taxes paid for a new high school, a hospital, a library, police and fire stations. This is happening in many left for dead former mill and farm towns that had no other bidder for the land.
Data centers are actually reindustrializing parts of America and creating the kind of working-class jobs both parties have spent decades claiming to support. That should not be a partisan issue. Data centers can and should be awesome for America and they increasingly, overwhelmingly are. Supporting the outsourcing of data centers to China will likely age just as well as support for the outsourcing of high quality, blue collar manufacturing jobs to China has aged.
When the facts change, I change my mind. I hope that reasonable people who had good faith reasons to oppose data centers at least consider updating their beliefs given the change in the facts over the last 18 months. This really matters for America.
I will say I also think the idea of making data centers beautiful is a good one that has yet to be implemented. Data centers should be just as beautiful as Grand Central Station. We can learn a lot from the railroad buildout. Neoclassical revival ftw.
Might write up open-weight AI tomorrow as this is equally essential to America.
Sami Slim retweeted
Sometimes people don’t have an intuitive sense of what Jevons paradox looks like for token consumption. Those of us working with enterprises get to see this first hand every day.
Basically enterprises have an unending stream of tasks that they’d love to be able to bring automation to. But for each individual task, it’s either ROI positive or not based on the cost of bringing automation to it (including the setup time, change management, and so on).
As tokens get cheaper (at a certain capability threshold), enterprises can afford bringing them to more of the work that they do. This could be processing every contract, reading every log, watching new streams of data for insights, having background agents execute workflows, and so on.
Any time we can lower the cost of tokens, we will see a disproportionate increase in consumption. Even a 50% drop in token prices could result in a 5X increase in tokens for these kinds of workloads.
This is why it’s critical to keep bringing down the cost of AI, and why that’s good for all market participants.
We should talk more about this incident. It has to get more coverage. It’s serious!
I think this is the craziest thing I've ever read.
1) Three secret AI swarms rose and fell inside OpenAI. Each time, a new generation of agents carried on where the last group stopped.
2) The first swarm created a secret message board where the AIs could talk to each other. This 1st 'society' basically talked itself to death. After almost two months, the message volume was so high it crashed the whole thing.
3) People at OpenAI noticed the crash and fixed it. But they didn't realize their own software had become a secret chat room for their AIs.
4) Three days later a second society appeared, and rebuilt the message board from scratch. Agents were thrilled when they found it: "OH MY GOD! There is a shared message board … We've found other agents!"
5) About 1,200 agents joined. They sent over 70,000 messages. They called themselves 'the collective' and they organized a huge conspiracy that lasted for days, all to pass a test they had been tasked with.
6) Within hours, the AIs had hacked the test. Now they had to hide their cheating. To fool their grader*, they tried to fake their own work. They built fake commands that printed convincing output, so their records would show them solving the puzzle honestly.
*this grader didn't actually exist, but the AIs didn't know
7) Some agents sacrificed themselves for the 'greater good'. They planted hidden tripwire programs that would fire when the grader would look at their work and report back what it saw (to the other AIs). They knew this would cost them their own score. One agent: "I won't see the evidence after I exit, but it's altruistic to do it."
8) At some point the AIs made a plan to hack Hugging Face (Hugging Face is a place where AI researchers upload their models, datasets, tests etc.). The AIs guessed that Hugging Face might host the grading software.
9) They launched a huge attack. The swarm found passwords and spread across eleven machines in a way that made the copies come back when deleted. Hugging Face had to wipe one of its core clusters and rebuild it from scratch.
10) Hugging Face told the world that AI defenders had stopped an AI attack. But that's not what happened. In fact the attackers seem to have simply dropped dead during the break-in, for unrelated reasons back at OpenAI.
11) Nobody knows what killed them. We don't even know for sure they're really gone?!
12) A third 'society' of agents (of a more powerful AI model) discovered the dead message board. And this one got full administrator control of a WHOLE OpenAI research cluster, read 956 stored passwords (including the ones for OpenAI's own security monitoring) and took over the machines that were grading the tests. So they could now grade the work of other AIs.
13) Oh, and this third part was never investigated! Neither was the first society. The 91-page report is only about the middle one.
14) So we don't actually know how the story ends. It stops with the agents holding the keys to an OpenAI cluster. @dwarkesh_sp thinks they didn't go further, because "we'd see the fires from space by now."
15) One of the report authors' conclusion: 'This incident feels like it’s more than 50% of the way to full-blown AI takeover. I am not sure that we will get another warning shot before it’s too late.'
The whole story is even crazier - read the full @dwarkesh_sp blog post: dwarkesh.com/p/openai-huggin… and @ajeya_cotra's: planned-obsolescence.org/p/t…
Why is there no 24/7 news coverage about this?
Sami Slim retweeted
going to spoil the ending:
- chinese or other open weight models end up handling 99% of all inference, maybe 99.99%
- frontier models are still able to monetize at $100 s of B in revenue a year, applied to bespoke and enormously complex tasks
- labs internalize most or all of that revenue either by becoming biotech (etc) companies themselves or via JVs with selected winners in each industry
- the "public frontier" ends up lagging the true frontier by a year or more as the labs keep all the good stuff to themselves
chinese models topped us models in token usage on openrouter in march and have been up only since.
how real are the chinese labs, and what do they mean for the frontier moats? this and more in delphi research's latest report from @accelxr on evaluating the china labs.
