#ai_ml — Public Fediverse posts on home.social

@hacks4pancakes You overestimate how many people in the USA even understand the gap in privacy protections between here and the EU. If you're on #LinkedIn, check out the current brouhaha about LinkedIn opting in all non-EU residents into their #AI_ML #dataaggregation and training without notice and before updating their terms of service.

In the US, most non-enterprise #TOS contracts of adhesion basically make it our responsibility to stay on top of the changes anyway; most of the time individuals aren't even provided advanced notice. Simply continuing to use the service in ignorance implies agreement with any new terms. In addition, those terms almost always say you agree that they can be changed at any time, with or without notice, at the service provider's sole discretion.

I've spent this entire week explaining to people why LinkedIn is allowed to do that to us but not to you; why they won't get more than a slap on the wrist, if that, from any oversight body; and why our current copyright laws and precedents around software licensing basically ensure that LinkedIn, #Microsoft, #Google, and #OpenAI will continue to get a free pass on literally stealing people's data and intellectual property, making it "proprietary" and sealing people's data up behind an impenetrable paywall, and then selling a slurry of appropriated data (including their own) back to them at the highest price the market will bear.

That's not free-market capitalism. It's just corporate welfare for large companies, institutional stockholders, and chip makers, plus a dash of good ol' fashioned "Robber Baron" economics.