IMO, even involving location and private data in the digital ecosystem that includes a centralized LLM is a very unwise thing to do.
We both know that LLMs can and will spit out ANYTHING in their training data regrdless of how many roadblocks are put up and protective instructions given.
While they’re not necessarily feeding outright personal info (of the general public, anyways) in to their LLMs’ models, we should also both know how slovenly greedy these cunt corpos are. It’ll only be a matter of time before they’re feeding everything they clearly already have in.
At that point, it won’t just be creep factor, but a legitimate doxxing problem.
Does it if you know, though…?
IMO, even involving location and private data in the digital ecosystem that includes a centralized LLM is a very unwise thing to do.
We both know that LLMs can and will spit out ANYTHING in their training data regrdless of how many roadblocks are put up and protective instructions given.
While they’re not necessarily feeding outright personal info (of the general public, anyways) in to their LLMs’ models, we should also both know how slovenly greedy these cunt corpos are. It’ll only be a matter of time before they’re feeding everything they clearly already have in.
At that point, it won’t just be creep factor, but a legitimate doxxing problem.