Rumor: GPT-6.1 Astra variant fails safety tests
A leak claims the 6.1 variant of Astra failed safety tests, but the team responded quickly and other model training was unaffected; the real bottleneck is a shortage of compute.
Yesterday we focused on Anthropic rumors, but make no mistake, OpenAI isn't just sitting around doing nothing. The GPT-6.1 Astra variant failed its safety tests, but the team is responding incredibly quickly and professionally, and this hasn't affected the training of other models. Don't mistake their current situation for a lack of progress. They're literally short on compute. That's it. And they're already working miracles with the compute they have available. You'll be seeing more from them soon.