This sounds to me like a step in the right direction. I only wish that they applied to all current datasets and not just any new model in going forward.
If you can't show the work for either what went in or the result, right to jail.
Black box shit is the bane of our modern social landscape, if we had killed these algorithms in their infancy, we'd have never suffered even a tenth of the insanity we've had to endure in the social media age.
Algorithmization is just a polite innuendo for Enshittification.
What would be the line between which algorithm fall under the category and which do not.
You can't enforce this for all the algorithms that would be impossible but declaring a criteria also would make all companies finding ways to bypass it.
I'm not sure an outright ban is justified, nor necessary. Maybe disallow copyright on AI generated material? That'd be a step in the right direction at first glance. I'd like to hear any counters to this stance.
That sounds like a weird argument, these billion-dollar companies know exactly what goes into their models. Give me a reason they wouldn't be able to disclose this if they were compelled.
Interesting that so much of the creative industry is supportive of the bill, but that the MPA isn't (or at least hasn't explicitly come out in support). I'm guessing that is reflective of the production companies' interest in using AI trained on creative content to eliminate the reliance on talent.
