osmarks (@o@osmarks.net)

I'm so glad OpenAI uses only the most robust safety practices when training the newest and most capable models.

osmarks microblog