Some say GPT-6 Astra got stronger a week after launch
A user compares Astra at launch with a week later, saying it improved rather than declined like GPT-6.1 Sol and Luna, and discusses whether OpenAI is heading toward recursive self-improvement.
🚨 OPENAI RSI? THE MODELS SEEM TO BE GETTING BETTER OVER TIME «Over the past week, I compared Astra at launch vs. today, alongside GPT-6.1 Sol and Luna The difference is honestly INSANE These models feel significantly more capable than they did at launch We’ve seen Astra reportedly get nerfed, and Claude models face similar complaints But with OpenAI’s recent models, I’m noticing the opposite trend Instead of getting worse, they seem to be getting BETTER And I’m not the only one noticing these changes Recent mathematical results and continuous model improvements make the question even more intriguing I’m not claiming OpenAI has achieved recursive self-improvement — but the possibility is fascinating» Is OpenAI quietly taking its first steps toward the RSI era?