alt Hacker NewsBut how does a true AI ban actually look in practice? The open weight models are out there, the hardware is out there. Spamming is already illegal but so commonplace that it's pretty laughably enforced at this point. Even really heinous stuff in the dark corners of the web still finds its way into the hands of users, I don't see how AI is any different in this regard.
Replies
A true [AI as-we-know-it-today] ban would require author consent on all materials added to a model, resulting in royalty payments (btw, I believe that’s how it should have been done from the start).
>Spamming is already illegal but so commonplace
here in Germany commercial cold calling carries fines of tens of thousands per call and I have received one in 20+ years.
The overwhelming majority of spam is advertisement. Someone by definition cannot advertise themselves without making themselves liable, so just up the penalty beyond the expected value of the spam and it works perfectly well.
> Spamming is already illegal but so commonplace that it's pretty laughably enforced at this point
I see comments similar to this about tech issues, and it always reads as if the person is assuming that these illegal tech activities are being done under the jurisdiction of the laws making it illegal. All you need is one jurisdiction that is not under the influence of the country with laws, and you make it impossible to stop these "illegal" activities