Re:The first general computer action model
by AndrewKemendo on Wednesday February 25, 2026 @06:14PM
(#47159417)
This looks like a really promising approach
In particular the Forward rollout module is very important. It aligns your (effectively) world model with what it expects from the world, and keeping those in sync I think gives this the power it needs to be able to generate the state action pairs to continuously train semi supervised
Re:The first general computer action model
by dangoodmanUT on Wednesday February 25, 2026 @06:15PM
(#47159424)
11 million hours of data is a lot, did you have to synthesize it at all, or was it purely collected?
Re:The first general computer action model
by nee1r on Wednesday February 25, 2026 @08:13PM
(#47160502)
collected! no synthetic
Re:The first general computer action model
by arkmm on Wednesday February 25, 2026 @07:06PM
(#47159959)
Get ready for the acquisition offers.
Re:The first general computer action model
by ilaksh on Wednesday February 25, 2026 @07:25PM
(#47160134)
How do I access this? Any HF or API coming?
Any benchmark comparisons to Fara-7B or Sonnet 4.6, Qwen 3.5 etc.?
This is a preview of a very different type of computer use model—we train on the internet. Specifically we have 11 million hours of computer video stored on our storage cluster (previously shared https://news.ycombinator.com/item?id=45438496 !) and the model can work in 30 FPS. Since we match the fundamental form factor of computer-use, we can get our model to do CAD, browse websites, and even drive a car using arrow keys. I’m super excited to see what our model can do as we scale more, it's a fun frontier to work on (not language models :) ).
The team and I will be online responding to the comments, so drop any questions.