Je conseille à mes followers de s’abonner au compte d’Olivier Babeau @OlivierBabeau Olivier a vraiment une vision passionnante sur les liens entre économie, politique et technologie Il décrypte parfaitement les conséquences de l’IA sur l’avenir de notre société Allez-y !
GENERATIVE AI
-
Motion Sketch: Gen-4.5 Image to Video with Sketch Annotation
By
–
Ever wish you could scribble a prompt instead of needing to write it?
— Runway (@runwayml) 3 février 2026
Now you can. With Motion Sketch, for Gen-4.5 Image to Video.
Upload your first frame image. Sketch and annotate what you want to happen. Then hit generate.
Try it today at the link below. pic.twitter.com/lfikEBEfBCEver wish you could scribble a prompt instead of needing to write it? Now you can. With Motion Sketch, for Gen-4.5 Image to Video. Upload your first frame image. Sketch and annotate what you want to happen. Then hit generate. Try it today at the link below.
-
Jamba2: Open Source Model for February On-Device Applications
By
–
Model to try out in February: Jamba2. Fully open source under Apache 2.0, includes 3B (dense) model for on-device apps, intended for grounded QA workflows that don't call for the heavy “thinking token” overhead of reasoning models. Further details:
-
Jamba2: Open Source Model for On-Device QA Applications
By
–
Model to try out in February: Jamba2. Fully open source under Apache 2.0, includes 3B (dense) model for on-device apps, intended for grounded QA workflows that don't call for the heavy “thinking token” overhead of reasoning models. Further details:
-
Structured Prompting for Better AI Outputs
By
–
PATTERN 3: Specify output structure Bad: "Be organized" Good: "Structure: [Problem] → [Solution] → [Implementation]. Use code blocks with language tags. Max 3 paragraphs per section. No emojis unless I use them first." ChatGPT follows this 94% of the time now.
-
Negative instructions improve ChatGPT output
By
–
PATTERN 1: Tell ChatGPT what NOT to do Bad: "Be concise" Good: "Never use: delve, landscape, robust, utilize, leverage, it's important to note, in conclusion" Why it works: Negative instructions are specific. Positive instructions are vague. Output quality jumped 2.1x with
-
5 Models to Improve ChatGPT
By
–
ChatGPT's custom instructions feature is incredibly powerful. But 99% of people write poor instructions. I tested over 200 custom instruction sets. These 5 patterns boosted output quality by 3.4x.
-
Runway Gen-4.5 Video Model Now Available on Poe
By
–
Runway Gen‑4.5 is now on Poe!
— Poe (@poe_platform) 3 février 2026
Runway’s frontier video model generates cinematic, high‑fidelity video from natural language and images, offering fine‑grained control over motion, camera behavior, composition, and visual style.
You can try it in Poe app on all platforms and in… pic.twitter.com/vrttvv2UE1Runway Gen‑4.5 is now on Poe! Runway’s frontier video model generates cinematic, high‑fidelity video from natural language and images, offering fine‑grained control over motion, camera behavior, composition, and visual style. You can try it in Poe app on all platforms and in
-
Play-testing the Genie 3 AI model
By
–
My pleasure! Dropped a new video with a lot more genie 3 play testing:
-
In-depth analysis and interview on Google DeepMind’s Genie 3 model
By
–
I tested Google’s world model Genie 3… Then DeepMind told me everything
— Bilawal Sidhu (@bilawalsidhu) 3 février 2026
00:00 – Intro & Authoring Workflow
00:27 – Genie 3 Playtesting & Demos
05:33 – Interview w/ Google DeepMind (Genie 3 co-lead @jparkerholder and Sr. PM Diego Rivas)
06:54 – Wildest emergent behaviors… pic.twitter.com/7Aafs67kaYI tested Google’s world model Genie 3… Then DeepMind told me everything 00:00 – Intro & Authoring Workflow
00:27 – Genie 3 Playtesting & Demos
05:33 – Interview w/ Google DeepMind (Genie 3 co-lead @jparkerholder and Sr. PM Diego Rivas)
06:54 – Wildest emergent behaviors