i reflected on my failings as an engineering leader integrating AI into my team (before i quit AI entirely); the result is beware: chronicling over 30 real risks from first-hand experience of using LLMs on a software team https://brettcodes.com/beware/