Frédéric Jutant

Le format, le fond, la forme ... Intéressante cette guerre de communication adoptée par l'Iran face aux États-Unis de Donald Trump

After analyzing 42,000 blog posts with an AI detector, content classified as fully human-written outperformed AI-generated or mixed content across the top 10 positions. The gap is most striking at position 1: 80.5% probability of being human-written vs just 10% AI-generated. This may appear to conflict with survey data showing that 72% of SEOs believe AI content ranks as well as human-written content. However, the data points to a more nuanced reality. The key difference emerges at the very top of the rankings. From position 5 onward, the performance gap between human-written and AI-generated content narrows significantly. In other words, AI content performs competitively when the benchmark is ranking on page one. However, in the highest-ranking positions, human-written content clearly leads. Full study: social.semrush.com/4skYo3T.

Google a lancé 3 updates majeures en 7 semaines. Discover. Core. Spam. Résultat : → 55% des sites impactés → Les fermes de contenu IA : -60 à 80% de trafic → HubSpot : -70% organique → Le spam update le plus rapide de l'histoire (20h) Ce n'est pas une purge. C'est Google qui hausse le niveau d'entrée. Le SEO médiocre est terminé.

Anthropic has just released a comprehensive guide to building and developing your own Skills in Claude resources.anthropic.com/hubfs/The-Comp…

Who did this? Lmao

@Divs_King Et acheter des actions, des ETF et de la crypto en vrac sans PEA depuis une appli bancaire genre Revolut c'est niveau combien ?

Niveau 0 : avoir l’intégralité de son épargne sur des livrets / comptes courants Niveau 1 : investir une partie de son épargne sur une assurance vie en fonds €. Niveau 2 : investir une partie de son épargne sur un PEA Niveau 3 : ouvrir également un CTO chez un courtier solide et international Niveau 4 : investir une partie de son épargne via des véhicules plus développés et optimisés : assurance vie luxembourgeoise, CTO personne morale, etc. Les niveaux 0 et 1 font littéralement perdre de l’argent (frais, inflations, sous-performance...) Si vous êtes au niveau 2, vous faites déjà mieux que 90% des Français 🇫🇷

🚨 Changement important Google : désormais, Googlebot ne récupère que 2MB max d'un fichier, contre 15MB avant ! Cette limite concerne les fichiers pris en charge par Google Search. Vérifiez l'impact sur votre indexation pour éviter des surprises ! 📉developers.google.com/search/docs/cr… #SEO

Introducing Claude Opus 4.6. Our smartest model got an upgrade. Opus 4.6 plans more carefully, sustains agentic tasks for longer, operates reliably in massive codebases, and catches its own mistakes. It’s also our first Opus-class model with 1M token context in beta.

I hope 2026 will be better than 2025 🙏 First week of January:

I feel this way most weeks tbh. Sometimes I start approaching a problem manually, and have to remind myself “claude can probably do this”. Recently we were debugging a memory leak in Claude Code, and I started approaching it the old fashioned way: connecting a profiler, using the app, pausing the profiler, manually looking through heap allocations. My coworker was looking at the same issue, and just asked Claude to make a heap dump, then read the dump to look for retained objects that probably shouldn’t be there; Claude 1-shotted it and put up a PR. The same thing happens most weeks. In a way, newer coworkers and even new grads that don’t make all sorts of assumptions about what the model can and can’t do — legacy memories formed when using old models — are able to use the model most effectively. It takes significant mental work to re-adjust to what the model can do every month or two, as models continue to become better and better at coding and engineering. The last month was my first month as an engineer that I didn’t open an IDE at all. Opus 4.5 wrote around 200 PRs, every single line. Software engineering is radically changing, and the hardest part even for early adopters and practitioners like us is to continue to re-adjust our expectations. And this is *still* just the beginning.

we definitely need more brilliant ideas like this

Today we released the December 2025 core update. We'll update our ranking release history page when the rollout is complete: status.search.google.com/incidents/Dsir…

Google tests merging AI Overviews with AI Mode techcrunch.com/2025/12/02/goo… via @techcrunch #SEO

🇫🇷 ALERTE INFO | L’URSSAF a été PIRATÉE : le service Pajemploi a été VICTIME d’un VOL de données concernant jusqu’à 1,2 million de personnes. Noms, prénoms, dates et lieux de naissance, adresses postales, numéros de Sécurité sociale et noms d’établissements bancaires ont été récupérés par les pirates. ✖️ Mais aucun numéro de compte bancaire, aucune adresse mail, aucun numéro de téléphone ni mot de passe de connexion n’a été compromis, souligne l’URSSAF dans un communiqué. (BFMTV)

BREAKING 🚨: Grok 4.1 Beta is rolling out on the Grok web! It is available as a standalone option, next to the existing Grok 4 modes. Testing time 👀

Introducing Scribe v2 Realtime – the most accurate real-time Speech to Text model. Built for voice agents, meeting notetakers, and live applications, it transcribes in 150ms across 90+ languages, including English, French, German, Italian, Spanish, Portuguese, Hindi, and Japanese. Available today by API and through ElevenLabs Agents.

How a Gemma model helped discover a new potential cancer therapy pathway @google blog.google/technology/ai/… #CancerResearch

"On a trouvé une molécule qui peut remplacer votre sang, je veux rester en France pour la développer mais j’étouffe à cause des normes" share.google/K0ihKznSUPS3eB…

Your smartphone is a parasite, according to evolution theconversation.com/your-smartphon… via @ConversationEDU

Incroyable numerama.com/politique/1992…