“astra is a powerful model and we are working to make it generally available. we do not think it is a good strategy to keep powerful models to a chosen few. given its cyber capabilities, we need a little big longer to do do this safely. but hopefully not too long!”
Altman urges people to use OpenAI models to defend their systems in response to GPT-5.6-Cyber release
“please consider using our models to help defend your systems https://t.co/oqs3qfWmtN”
“We have paused some frontier RL training to ensure that we can meet the appropriate alignment, security and monitoring standards for the new level of capabilities in front of us. Model progress is now extremely rapid, and we always said we would take action if we felt that model”
Altman clarifies the training pause impacts further-out releases but great new models will ship soon
“(We still expect to ship great new models soon; this impacts further-out releases.)”
Altman says he was accurate on capability timelines but society and economy are integrating AI slower than expected
“@KeyTryer i don't think i was wrong on capability timelines; actually quite good on those predictions. society and the economy are taking longer to integrate the capabilities than i had guessed, but i don't think that's all bad. (also might change in the future.)”
Altman predicts an upcoming model will achieve meaningful invention capabilities for the first time.
“I expect this will be the first model where the model actually invents new things in a way that matters,”