gpt-6POST
Commentary: OpenAI pulling GPT-6.1 Astra counts as real accountability
A short comment arguing that OpenAI dropping its own GPT-6.1 Astra model over a failed safety bar is a rare case of responsible behavior.
Ed Elson
@edels0nArchived 2026.09.29Originally posted 2026.09.29
Ed Elson @edels0nOriginally posted 2026.09.29
Some good news: OpenAI killed its own model, GPT-6.1 Astra, because it failed to meet its safety bar, the company says. Finally, some real accountability.
x.com/edels0n/status/2104716216166428693Archived 2026.09.29 10:12