Working on: 1) less biased defaults, 2) lots of user customization within very broad bounds, 3) public input on bounds and defaults https://
openai.com/blog/how-shoul
d-ai-systems-behave/
…

→ View original post on X — @gdb,