Hacker News new | ask | show | jobs
by bjourne 53 days ago
Probably? Reinforcement learning creates bots with specific styles. For example, ChatGPT is very fond of "typically", "unpack this", and "if you want".