[nopinsight]

They are focusing on technical advances to enable people to save themselves from future AGI. For that, they need more time since AI Safety research lags quite a bit behind capability research.

https://openai.com/blog/introducing-superalignment