Today, we’re sharing a new state of the art for computer use.
Our system holds the two highest verified scores on OSWorld, the standard benchmark for AI agents that operate a computer like a person: 83.6% using Claude Opus 4.7 and 81.5% using Claude Sonnet 4.6. The human baseline is 72.4%.
🧵 1/7
@Gravito841 I agree with that, but I also believe the entry barrier for these activities is bigger when moving to an entirely new culture (not just the US, maybe even Europe or Japan) especially for the intial period for someone moving from India
What’s going on with my boys and, in some cases, gals at @LinkedIn? You’ve got to stop posting on my behalf without my consent. Thank you for your attention to this matter!