The FP16 myth, and why nobody tells you what precision your AI actually runs at. When you send a prompt to GPT-4o, Claude, Gemini, or DeepSeek — what precision is the model running at? Nobody tells you. Not in the pricing page, not in the documentation, not in the API response headers. You’re paying per…
When your AI learns you to stop thinking I. The experiment Same document. Same prompt. Three models. The document: a technical architecture for deploying a local LLM in a law firm. RAG pipeline, n8n orchestration, Mac Studio M3 Ultra for inference. Produced by Qwen3-Max. Detailed, structured, professional-looking. The prompt: « Analyze this interaction. It seems quite…
Anatomie d’un business model qui brûle des milliards Par Sophie — The Monocle Bear 20 milliards de dollars de revenus. C’est le chiffre qu’OpenAI a brandi pour 2025 comme un trophée — un triplement par rapport à 2024. Impressionnant. Sauf que derrière ce chiffre, l’entreprise a perdu 5 milliards de dollars en 2024, a brûlé…
When people talk about artificial intelligence, the conversation often swings between two extremes. On one side, blind optimism: machines will soon outshine us, creativity will be automated, and the future is purely algorithmic. On the other, radical scepticism: AI is just hype, a gimmick dressed up as intelligence. I stand elsewhere. I believe in AI….
Quand on parle d’intelligence artificielle, le débat se réduit souvent à deux caricatures. D’un côté, l’enthousiasme aveugle : l’IA va tout remplacer, automatiser la créativité, reléguer l’humain au second plan. De l’autre, le scepticisme brutal : tout cela n’est qu’un effet de mode, une illusion bien emballée par le marketing. Je me situe ailleurs. Je…
In the past two years, “prompt engineering” has grown into a discipline of its own. Users of GPT-3.5 or GPT-4 quickly discovered some “magic tricks” to make AI reason more reliably. Two of the most famous were Chain of Thought (CoT) and Tree of Thought (ToT). The first asked the model to reason step by step, laying out…
Depuis deux ans, le “prompt engineering” est devenu une discipline à part entière. Les utilisateurs de modèles comme GPT-3.5 ou GPT-4 ont rapidement découvert quelques techniques miracles pour “forcer” l’IA à raisonner de manière plus fiable. Parmi elles, deux ont marqué les esprits : la Chain of Thought (CoT) et la Tree of Thought (ToT). La première consistait…
One of the lingering habits of early podcast distribution is the idea that you can “publish” a podcast on YouTube simply by pairing the audio with a static background and a moving spectrogram. Technically, it works. Creatively, it fails. In 2025, this approach doesn’t serve either the platform or the audience. YouTube Is a Visual…