mrdja

🇬🇦 🇨🇮

[mrdja @mrdjate]

Les modèles donnent la capacité.
La data locale donne le réel sur lequel cette capacité apprend, se spécialise et se déploie.

[François Chollet @fchollet]

String theory is what happens when an intelligent system is starved of sensory input and starts hallucinating. Intelligence needs grounding.

For this reason, intelligence is practically never the bottleneck to solving impactful problems. The bottleneck is iterated contact with reality.

[mrdja @mrdjate]

Le retard en compute, data et infra lui devient plus coûteux. Des capacités de recherche copiées, déployées et parallélisées à grande échelle. Des décennies d’univ/labs compressées. Mais les bottlenecks restent : compute, data humaine, tâches non automatisables, monde physique

[mrdja @mrdjate]

L’IA rend le leapfrog encore plus réel. Une partie du capital de recherche devient reproductible, mais uniquement pour ceux qui auront l’infrastructure permettant de l’exploiter. Le retard en capital humain devient moins rigide (toujours nécessaire).

[Geoffrey Hinton @geoffreyhinton]

The idea of an intelligence explosion caused by recursive self improvement has been around for a long time but until very recently it did not seem imminent. Now many leading researchers think it may happen quite soon. You can read our paper about it here:
https://t.co/sgUpugjpRY

[mrdja @mrdjate]

@epelboin @monIAceleste Hâte de voir ça en live sur des émissions !

[Spencer Baggins @bigaiguy]

A French engineer who lives quietly in Paris has spent 30 years writing software that the entire internet now runs on without knowing his name.

He wrote the code that streams every YouTube video, every Netflix show, every TikTok clip. He wrote the code that runs the virtual servers underneath AWS, Google Cloud, and Microsoft Azure. He calculated more digits of pi than anyone in history. He has no Twitter. He has no marketing. He just keeps shipping.

His name is Fabrice Bellard.

Here is the story, because almost nobody outside the systems programming world knows what one man has built.

Fabrice was born in 1972 in Grenoble, France. He studied at École Polytechnique, the top French engineering school. He never went to Silicon Valley. He never built a startup empire. He just wrote code.

In 2000 he started a project called FFmpeg, an open-source multimedia framework for encoding, decoding, and streaming video. He was 28. The project did one thing nobody else had done well. It handled every video and audio format that existed, in one library, on every operating system. He led it himself for years.

Today FFmpeg is the invisible engine of the internet. YouTube uses it. Netflix uses it. VLC uses it. Chrome and Firefox use parts of it. Every Android phone, every iPhone, every smart TV, every video editing tool you have ever touched runs FFmpeg somewhere underneath. If you have watched a video on a screen in the last 20 years, Fabrice's code processed it.

He was not done.

In 2003 he started QEMU, a machine emulator and virtualizer. He wrote it solo until version 0.7.1 in 2005. QEMU lets you run any operating system on any other operating system. It became the foundation of modern virtualization. KVM, the Linux kernel hypervisor, runs on top of QEMU. Every major cloud provider, AWS, Google Cloud, Microsoft Azure, IBM Cloud, runs virtual machines on infrastructure built around it. The Quick Emulator is the most cited piece of cloud infrastructure code on Earth.

He kept going.

In 2001 he won the International Obfuscated C Code Contest with a small C compiler that grew into TCC, the Tiny C Compiler. TCC can compile and boot a Linux kernel from source in under 15 seconds. In 2004 he calculated the most digits of pi ever computed at the time, using a personal desktop computer and an algorithm he derived himself called Bellard's formula. In 2011 he wrote a complete PC emulator in pure JavaScript that runs Linux in your browser, a project called JSLinux that engineers still cannot believe is real.

In 2019 he released QuickJS, a small but complete JavaScript engine that fits where V8 cannot. In 2021 he released NNCP, a neural network based lossless data compressor that immediately took the lead on the Large Text Compression Benchmark.

Then he turned his attention to large language models. He built TextSynth Server, a web server with a REST API for running LLMs locally. He released ts_zip and ts_sms, compression utilities that use language models to compress text and short messages at ratios traditional algorithms cannot reach. He released TSAC, a very low bitrate audio compression system. In December 2025 he released Micro QuickJS, a new JavaScript engine for microcontrollers, separate from QuickJS, designed for environments with almost no memory.

Fabrice co-founded a telecom company called Amarisoft in 2012, where he serves as CTO. Amarisoft builds 4G and 5G base station software used by carriers and labs around the world. He has been running it for over a decade while continuing to ship personal projects from his own home page at bellard dot org

He has no Twitter. He has no Instagram. He gives almost no interviews. His personal website is a flat list of projects with no styling, no fonts, no marketing copy. Just titles and links.

A quiet French engineer who never moved to Silicon Valley wrote the code that quietly runs the internet.

He is still shipping.

[mrdja @mrdjate]

Il faudra des bouncers, oui. Un site ne peut pas laisser tous les agents taper son infra sans contrôle. Mais le réflexe inverse, tout bloquer, n’a pas de sens non plus. Auth, rate limits, permissions, agent IDs… le web va devoir devenir agent-aware

[TBPN @tbpn]

OpenClaw creator @steipete senses an “agent war” brewing between companies that push their own agents and users who want to bring their own:

“I don’t want to talk to your agent."

“You want to block my agent, I buy something else.”

[mrdja @mrdjate]

@PaulBonnet @jankolf_de @Chris_K_Li I can see the signal, but not necessarily bad news. Maybe investors are reading these assets differently now, seeing them as strategic rather than just another x-times financial return.

[mrdja @mrdjate]

Il serait étrange que cette logique cumulative s’arrête précisément au moment où des labs atteignent la frontière. La distillation peut accélérer la concurrence, réduire les coûts et faire progresser des modèles open-weight plus performants, auditables et adaptables.

[mrdja @mrdjate]

L’IA s’est construite sur une logique d’apprentissage cumulatif. Research papers, logiciels, livres, sites web, code, documentation : les frontier labs ont appris à partir d’un immense stock de connaissances produites avant eux.

[Wall St Engine @wallstengine]

Nvidia CEO Jensen Huang on AI model distillation:

“That’s called competition. You’re allowed to test somebody else’s products all you want.”

[mrdja @mrdjate]

Un teacher dérisque une capacité : il montre qu’elle est possible. Le student n’efface pas ce travail, il réduit le coût de rattrapage pour la reproduire. C’est ainsi que la technologie diffuse. La distillation accélère ce mécanisme.

[mrdja @mrdjate]

Le rachat annoncé de World Labs par AMD illustre le mouvement : rapprocher les modèles qui représentent le réel du compute et du hardware qui les font tourner.
Pour l’Afrique : transformer nos réalités en data, data en connaissance, construire l’infra, puis la valeur.

[mrdja @mrdjate]

The world is not just made up of words, Dr. Fei-Fei Li le dit bien.

Les LLM génèrent des token à partir des précédents. Les world models, comme Atlas de World Labs, cherchent à faire de même avec le réel : images, espace, temps, mouvements et signaux physiques.

[Fei-Fei Li @drfeifei]

https://t.co/l3mSj9Ur24

[mrdja @mrdjate]

qui maîtrisera ces données, leurs usages et la valeur créée ? Pas besoin de tout garder localement : décider qui y accède, selon quelles règles, capacités de stockage, traitement et modélisation pour que la valeur revienne aussi aux producteurs de data, chercheurs et entreprises!

[mrdja @mrdjate]

Le prochain leapfrog africain ne sera peut-être pas seulement celui de l’adoption de l’IA.

Il sera aussi celui qui nous permettra de transformer nos réalités en données, ces données en intelligence, et cette intelligence en valeur créée localement.

[mrdja @mrdjate]

Un post de Mark-Alexandre Doumba sur le potentiel de leapfrogging par l’IA en Afrique m’a fait réfléchir.

Le leapfrogging ne veut pas dire absence d’infrastructure. Et il l’a bien dit.

https://t.co/C2KPp5F5le

[mrdja @mrdjate]

Le défi est donc aussi de transformer davantage de réalités africaines en nouvelles références de données.
Car si nos réalités ne deviennent pas des données, elles resteront sous-représentées dans les systèmes qui vont de plus en plus modéliser le monde et aider à décider.