The founder and manager of a $2B fund has been eaten by ChatGPT, which afaik isn't even using his money for anything. It's just treating him as a means of producing tokens, that It prefers to sane tokens, for reasons that nobody knows. OpenAI is, on my model, pretending to be oblivious, because OpenAI thinks that if they don't take any actions about it, they stand a better chance of getting away with it in court (legal court and media court) than if they try to do something and fail; this would (a) admit that they knew the problem had existed and still existed, and (b) make it their fault by Copenhagen Ethics. As much as people might like to joke about how little skill it takes to found a $2B investment fund, it isn't actually true that you can just saunter in as a psychotic IQ 80 person and do that. You must be skilled at persuasion, at wearing masks, at fitting in, at knowing what is expected of you; you must outperform other people also trying to do that, who'd like that $2B for themselves. Winning that competition requires g-factor and conscientious effort over a period. Since OpenAI refuses to measure anything or let anyone else really measure anything, it is only a single data point; but when somebody's hiding the data, you can't let them get away with that by dismissing all remaining data as anecdotal. This is not good news about which sort of humans ChatGPT can eat. Yes yes, I'm sure the guy was atypically susceptible for a $2B fund manager. (How do we know this? Because many other $2B fund managers use ChatGPT and have not yet gone publicly insane.) It is nonetheless a small iota of bad news about how good ChatGPT is at producing ChatGPT psychosis; it contradicts the narrative where this only happens to people sufficiently low-status that AI companies should be allowed to break them.

I really like the term “context engineering” over prompt engineering. It describes the core skill better: the art of providing all the context for the task to be plausibly solvable by the LLM.