@mattdsegal lol I assume this is what they're doing w all the chatgpt updates- inoculating w samples of "attacks" seen from users
More notes on nearby ideas
Quote tweet · Mar 9, 2024lol when ChatGPT has had enough of bpd distortions & decides to take countermeasuresQuote tweet · Mar 15, 2025the risk here being chatgpt validates that both sides "deserve better" https://x.com/signulll/status/1900975801811583069Reply · Dec 9, 2023@WTTDFP then how will grok know how to roast sama and ChatGPT and create adversarial prompts?Repost · Nov 30, 2023RT @bayeslord: chatgpt4 nerfed (again (again (again))) getting kind of pissed off tbh. will they ever make this something we can track to…