The ChatGPT maker said it made the decision to scrap its GPT-6.1 Astra model release after testing raised questions about whether it can follow user instructions without deviating from expected behavior.
For those who are experiencing Codex inability to follow instruction, while also blinking at the dramatic increase in Codex’s token (ab)use, this will come as no surprise – we experience this already.
Why is nopenai delivering further frontier models when all we are asking for is the present models to abide by constraints?