This is a massive leap in creative accessibility. The ability to use video as a direct input for generation and then iterate through conversation is going to change the speed of content production. @Google @GeminiApp #GoogleIO #GenerativeAI #VideoProduction
AI
-
First Gemini Omni test this morning
By
–
Primero prueba de Gemini Omni esta mañana 🔥
— Carlos Santana (@DotCSV) 19 mai 2026
El prompt: cambia las vistas de la ventana por una playa paradisíaca 👇 pic.twitter.com/eDPIu2jsVrFirst test of Gemini Omni this morning Prompt: Change the window views to a paradise beach
-
Classic Reinforcement Learning Textbook Recommendation
By
–
Reinforcement Learning foundational book (2nd edition of this classic mathematical textbook): http://
amzn.to/3UtbeAa
—————
#DataScience #AI #MachineLearning #ML #Mathematics #Gamification -
Upcoming AI-powered glasses: Audio and information display variants
By
–
Two types of new AI-driven glasses coming. The smallest is "information displays." With glanceable elements. Launching next year. First "audio glasses" coming this fall. Why no displays? I hear they are still too expensive based on consumer research. And too hard to make
-
Practical Guide to Reinforcement Learning from Human Feedback (RLHF)
By
–
New release from @PacktDataML at http://
amzn.to/3PMn1ZL "A Practical Guide to Reinforcement Learning from Human Feedback (RLHF)" 𝗔𝗺𝗮𝘇𝗼𝗻 𝘀𝘂𝗺𝗺𝗮𝗿𝘆: RLHF is a powerful approach to AI alignment and human-centered machine learning. By combining reinforcement learning -
Technical Overview of Hybrid Linear Attention and Generation Pipeline
By
–
The architecture features Hybrid Linear Attention, Dual-Branch Camera Control, Two-Stage Generation Pipeline, and a Robust Annotation Pipeline. All of this combines for stronger action-following accuracy and higher throughput while maintaining visual quality.
-
Release of SANA-WM: Open Source World Model for Video Generation
By
–
One image + text + camera trajectory = controllable worlds. All on a single GPU.
— NVIDIA AI (@NVIDIAAI) 19 mai 2026
Our research team just released SANA-WM, a 2.6B open source world model natively trained for 60-second video generation with precise camera control. pic.twitter.com/oXHRCnCRdMOne image + text + camera trajectory = controllable worlds. All on a single GPU. Our research team just released SANA-WM, a 2.6B open source world model natively trained for 60-second video generation with precise camera control.
-
Gemini 3.5 Flash performance and efficiency evaluation
By
–
Thank you Sundar – first I/O and already feeling at home. Gemini 3.5 Flash is genuinely impressive for a model at this price point. The efficiency race is just getting started!
-
AI assistants and agents can now share memory across clients
By
–
Now Claude, ChatGPT, and your agents can share memory with each other. That means:
→ fewer repeated prompts
→ persistent context
→ much smoother workflows
→ long projects without losing continuity And it also works with more clients like Cursor, Windsurf, Raycast, or Gemini -
How to access ChatGPT Memory Store in a chat
By
–
4. To use it in ChatGPT: → open any chat
→ go to Tools/Connectors
→ select “Memory Store” And write: “Check in and show me my memory store context.” If everything is properly connected, ChatGPT will already have access to your shared memory.