It’s been an unbelie-veo-ble week of launches! Here’s everything that happened: — We launched Veo 3.1 with a suite of new features to give you more creative control of your video generations. — We brought Nano Banana into new surfaces like @Google Search, @NotebookLM
, Slides in
INNOVATION
-
Veo 3.1 Launch: New Creative Control Features and Nano Banana Integration
By
–
-
Open Source AI Week in San Francisco Live Coverage
By
–
Follow our live blog as we head to Open Source AI Week in San Francisco:
-

Two AI Research Leaders Launch Podcast Series October
By
–
Two leaders in applied research. One podcast. Episode drops October 21st. https://
nvda.ws/3KWlUV1 -
Previously Solved Problems Updated in AI Research List
By
–
I think there's been a misunderstanding. These were in fact previously *solved* problems. The only breakthough was updating a list that erroneously listed some as unsolved.
-
Designing Perfectly Secure Computers with OpenAI Sora
By
–
Designing the perfectly secure computer, thanks to OpenAI and Sora: pic.twitter.com/uWqIMhSSj4
— Bob Gourley – e/acc (@bobgourley) 17 octobre 2025Designing the perfectly secure computer, thanks to OpenAI and Sora:
-
ML Improves Computer Systems Through VM Lifetime Prediction
By
–
Another great example of how ML can be used to improve computer systems, by learning how the system is actually being used (in this case by predicting VM lifetimes for optimizing their placement). Nice work, Pratik Worah and @martin_maas and coauthors!
-
Codex Cloud Integration: Enhanced IDE Workflow and Task Management
By
–
Great question! It’s a different style of interaction. You can hand off tasks to Codex in the cloud, browse your cloud tasks, apply changes locally with a visual diff (excellent with Best-of-N), smart auto context… It’s a more integrated experience if you love working in an IDE!
-
NVIDIA’s Open Source Progress Finally Getting Recognition
By
–
Great to see @NVIDIAAI
's recent great progress on open source starting to get noticed. I feel like they been really under the radar. Perhaps because their earlier models were buried under crappy licenses. But they've really come around in recent months. -

SWE-grep: ML-Powered Code Search and Explanation Tool
By
–
SWE-grep is truly some inspired ML from @cognition
. Take a 1M line codebase like React. With multiple fast inference calls on Cerebras, it can fetch & explain relevant code in seconds. Here are a few measurements taken on React, Vercel, PyTorch repos. -
Open Source Power: Community-Driven Development and Innovation
By
–
not sure yet. we'll see how people use it or build on top of it (the power of open-source)