As we saw with GPT-5 the RL technique of training doesn't scale forever | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		poorman 5 months ago \| parent \| context \| favorite \| on: GPT-OSS vs. Qwen3 and a detailed look how things e... As we saw with GPT-5 the RL technique of training doesn't scale forever

energy123 5 months ago | [–]

Unless GPT-5 is 30% cheaper to run than o3. Then it's scaling brilliantly given the small gap between release dates. People are really drawing too many conclusions from too little information.

oezi 5 months ago | [–]

I meant scaling the base training before RL.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact