15 ms·
That's dangerous, because mitigations after the fact can be very far reaching. Not only dangerous for users, but also companies. So lets imagine it for a moment
by zelphirkalt 2y ago
That's dangerous, because mitigations after the fact can be very far reaching. Not only dangerous for users, but also companies. So lets imagine it for a moment, that something like this needs to be mitigated:
Who used any resulting model to create anything at all? In which cases did something too similar to a real person result from the usage? How do we track all that down and delete it? Then we will need some IT guys coming over to the conpany in question going through their stuff, to ensure that all remnants are deleted, which only happens after looking at possibly millions of logs, to know who got those too similar results in the first place.
Making right after the fact always has a potential to be magnitudes more work than not allowing someone to use some personal data.
- scarab92 2y agoIt’s always going to be easier to ban every new invention. But public policy shouldn’t be based on what’s easiest, it should be based on what’s best for society, and that requires considering the trade offs. Right now there’s no evidence that anything Facebook are doing causes real world harm, so the trade off favours allowing it. If that changes, then policy can change at the same time.
- zelphirkalt 2y ago> But public policy shouldn’t be based on what’s easiest, it should be based on what’s best for society, and that requires considering the trade offs. Surely that is not simply grabbing people's personal data and using it for new purposes without consent. > Right now there’s no evidence that anything Facebook are doing causes real world harm, so the trade off favours allowing it. If I had been living under a rock, I might have claimed this. However, FB is complicit in genocide, so naah, not so sure about that claim. I would rather we have a very close eye on them.