80 karma · joined March 25, 2023
> It may look like ordinary text, but when it is placed into an LLM context window, the model may interpret it as an instruction rather than as data.
I feel like as long as this is the case, we'll never have secure LLMs. It concisely summarises the alarm bell I hear every time someone talks about adding AI features to their product. I plan on using this as a sort of benchmark for future AI discussions: "how do you plan on separating data from instructions?"
The UK was amazing - no questions asked care for my health was a massive improvement over what I had in the US. Free access to a doctor at any time is more life changing that you can think - every health issue I had I could talk to a medical professional and get guidance. Even free mental health care.
Moved to Ireland last year, with a mixed public/private system, and my health is now slipping a bit. Going to a GP, trying to file for reimbursements, and knowing that every time I have a health concern, it will cost me money, is a powerful suppressing force. I miss the NHS.