I agree with Dario that we need to pace the frontier. This has been a primary topic of discussions we've had at OpenAI in recent weeks.
Committing to having independent evaluators with employee-like access is a great idea, and we will do the same. We'll have more to share soon.
Gemini 3.8 flash is live
it seems to be benchmaxxed still its not as bad as i had hoped it was able to create this amusement park game within a minute in one shot
its not phenomenal but better than i expected
- i will test it more and see about how much aligned it is