Mes confrères médecins ne réalisent pas la transformation de notre métier 1 ChatGPT o3 est 4 fois meilleur que nous médecins pour faire un diagnostic 2 ChatGPT plus médecin c’est nettement moins bon que ChatGPT tout seul 3 déjà 40 % des patients font confiance à ChatGPT
GENERATIVE AI
-
Training Data Quality Impact on AI Model Outputs
By
–
If anyone is unlucky enough to let their model train on my collection of SVGs of pelicans riding bicycles they're going to get some VERY weird looking pelicans riding bicycles
-
Multimodal AI Risks: Vision Models and Safety Concerns
By
–
"The marginal risks of open models have been shown to not be as extreme as many people thought (at least for text only — multimodal is far riskier)." What makes multimodal riskier – assuming you mean vision is it concerns over surveillance, facial recognition etc?
-

OpenAI’s Open Source Strategy Signals Confidence in GPT-5
By
–
Very good point from “The Information”: The fact that OpenAI is releasing an OpenSource model, and they are carrying a large cost with it, shows impressively how confident they are in the strength of GPT-5. OpenAI believes that GPT-5 will be a breakthrough – and will easily
-
Model Communication Protocol: From Text-Based Design to Dominance
By
–
Model Communication is going to happen slowly, and then all at once:
— dr. jack morris (@jxmnop) 6 août 2025
Level 0: We design a text-based protocol for models & programs to communicate (read: MCP)
Level 1: Text-based model communication grows to exceed human communication in worldwide internet bitstream volume… pic.twitter.com/3vP6yPBmpTModel Communication is going to happen slowly, and then all at once: Level 0: We design a text-based protocol for models & programs to communicate (read: MCP)
Level 1: Text-based model communication grows to exceed human communication in worldwide internet bitstream volume -

GPT-5 Models Confirmed: Release Expected Today or Tomorrow
By
–
GPT-5, -mini, -nano confirmed, images already found. Either today or tomorrow probably the release. Creds to @haider1 for finding these
-
Trace Claude Code Sessions to LangSmith Integration
By
–
You can now trace your claude code sessions to LangSmith! See how to set up tracing from claude code to LangSmith in just a few minutes: https://
youtu.be/w_xp06VT6aU Check out the docs for setup instructions: -

Google’s Genie 3: Analyzing World Model Capabilities
By
–
¡NUEVO VÍDEO en el LAB! El nuevo sistema Genie 3 de Google es todo un hito en lo que a Modelos del Mundo se refiere. Capaz de crear instantáneamente cualquier mundo que imaginemos, listo para explorar! Hoy analizamos sus resultados e implicaciones, link a continuación
-

RunwayML Aleph: AI Video Editing with Text Prompts
By
–
introducing Aleph.
— KREA AI (@krea_ai) 6 août 2025
this new video model from RunwayML allows you to edit videos using text prompts.
try it now in Krea Restyle. pic.twitter.com/pLlsWWttgwintroducing Aleph. this new video model from RunwayML allows you to edit videos using text prompts. try it now in Krea Restyle.
-
OpenAI OS Models: Fine-tuning and Weight Download Capabilities
By
–
Are you guys thinking of doing anything with the OS models on the OpenAI Platform? e.g. fine tuning and downloading the weights?
