The #Aletheia paper is finally available on arXiv https://
arxiv.org/abs/2602.10177! Excited to share the 1st wave of papers on AI for math research! More to come very soon, stay tuned! Blog: https://
deepmind.google/blog/accelerat
ing-mathematical-and-scientific-discovery-with-gemini-deep-think/
…
MULTIMODAL AI
-
Aletheia: AI Advances Mathematical Research with Gemini Deep-Think
By
–
-
Real-time world models enabling instant actor control
By
–
What comes next is even crazier: you can do this in real time and tell the actor to kick or punch, and it will happen instantly. I just posted about real-time world models, which are in their early days but will bring these capabilities to us soon. Then, sometime in the next 18
-
Why real-time world models matter more than video AI
By
–
Why real-time world models are more important than SeeDance or Kling.
— Robert Scoble (@Scobleizer) 11 février 2026
You've seen all the crazy videos. I've posted many of them myself from SeeDance, Kling, and others.
AI is getting amazing at creating video content, but don't miss what's going on elsewhere in the real-time… https://t.co/eSLTms8skPWhy real-time world models are more important than SeeDance or Kling. You've seen all the crazy videos. I've posted many of them myself from SeeDance, Kling, and others. AI is getting amazing at creating video content, but don't miss what's going on elsewhere in the real-time
-
Visual AI Agents: Video Reasoning and Actionable Insights with NVIDIA
By
–
The next wave of AI agents won’t just read text—they’ll reason over video and extract actionable insights.💡
— NVIDIA AI (@NVIDIAAI) 11 février 2026
In this #NVIDIAGTC session, be the first to hear about the new features powering visual AI agents with NVIDIA Blueprint for Video Search and Summarization (VSS), open… pic.twitter.com/lPPzlDW3vaThe next wave of AI agents won’t just read text—they’ll reason over video and extract actionable insights. In this #NVIDIAGTC session, be the first to hear about the new features powering visual AI agents with NVIDIA Blueprint for Search and Summarization (VSS), open
-
Progress and Challenges in Multimodal AI Generation
By
–
How far things have come since 2023, and yet how much further we have to go. 2.5d/3d/4d is so much harder than other modalities, but we are just about getting to the fun part! https://t.co/E8IJLh7qod
— Bilawal Sidhu (@bilawalsidhu) 11 février 2026How far things have come since 2023, and yet how much further we have to go. 2.5d/3d/4d is so much harder than other modalities, but we are just about getting to the fun part!
-
Workflow for 3D Reality Reskinning and Spatial Reconstruction
By
–
Clean workflow to reskin reality
— Bilawal Sidhu (@bilawalsidhu) 11 février 2026
360 pano > nano banana > world labs
Big limit is the navigable volume. Once we can auto fuse multiple 360 panos this approach gets v compelling.
Y capture so much 3d data when are you need are a few nicely sampled panos?pic.twitter.com/vq2fSQnVR3Clean workflow to reskin reality 360 pano > nano banana > world labs Big limit is the navigable volume. Once we can auto fuse multiple 360 panos this approach gets v compelling. Y capture so much 3d data when are you need are a few nicely sampled panos?
-
Advancements in Computer Use Agents and VLMs for Automation
By
–
Between OpenClaw and VLMs getting better much at computer use – the loom recording to automated workflow dream is nearly ready for prime time
-
Seedance 2.0 Generates Impressive AI Video of Will Smith Fighting Spaghetti Monster
By
–
Vous vous souvenez de la vidéo de Will Smith qui mangeait des spaghettis générée par IA ?
— VISION IA (@vision_ia) 11 février 2026
C'était devenu un mème tellement c'était mauvais.
Seedance 2.0 vient de générer un Will Smith qui SE BAT contre un monstre en spaghettis. Style film d'action des années 80. Plusieurs… pic.twitter.com/SPbHiIltXPVous vous souvenez de la vidéo de Will Smith qui mangeait des spaghettis générée par IA ? C'était devenu un mème tellement c'était mauvais. Seedance 2.0 vient de générer un Will Smith qui SE BAT contre un monstre en spaghettis. Style film d'action des années 80. Plusieurs
-
Playing with Kling 3.0 for AI-generated movies
By
–
Kling claws.
— Robert Scoble (@Scobleizer) 11 février 2026
I’m playing around with Kling 3.0, which was hot last week as a great way for creating AI-generated movies. Getting ready for a new look at all the text to video tools.
Say hi to the lobster, which I created in Kling in about two minutes. You just need a little… pic.twitter.com/qcscZx8dFGKling claws. I’m playing around with Kling 3.0, which was hot last week as a great way for creating AI-generated movies. Getting ready for a new look at all the text to video tools. Say hi to the lobster, which I created in Kling in about two minutes. You just need a little
-
Multimodal AI model gains text-to-video and image-to-video
By
–
2. Upgraded audio. You can now pinpoint the exact character speaking.
— Robert Scoble (@Scobleizer) 11 février 2026
3. Much improved text. Precise lettering capabilities. Great for making visuals for slide decks and web sites. And, of course, X posts.
4. Unified model. You can do text-to-video. Image-to-video.… pic.twitter.com/tYde4uzBt42. Upgraded audio. You can now pinpoint the exact character speaking. 3. Much improved text. Precise lettering capabilities. Great for making visuals for slide decks and web sites. And, of course, X posts. 4. Unified model. You can do text-to-video. Image-to-video.