User claims Astra improved after launch, suspects OpenAI is entering the RSI era
The author compares launch-day and current videos of Astra, argues it has gotten better rather than worse, cites GPT-6.1 Sol and Luna as supporting evidence, and speculates that OpenAI is doing recursive self-improvement.
🚨 OpenAI RSI Openai might be doing recursive self-improvement And I have evidence that makes me question everything Over the last week, I made videos comparing Astra at launch vs today, alongside GPT-6.1 Sol and Luna The difference is INSANE.l These models feel significantly better than they did at launch We've seen Astra get nerfed We've seen Claude models get nerfed, too But with OpenAI's recent models, I'm noticing something different Instead of getting worse, they seem to be getting BETTER And I'm not the only one noticing this I'm not saying OpenAI has achieved it, but their recent math results and model improvements make me wonder Is OpenAI quietly entering the RSI era?