papetoast comments on papetoast’s Shortforms

[papetoast]

Reinforcement learning towards broadly and persistently beneficial models (linkpost)