Hacker News Reader

First rule is AI cannot make a management decision

Because it cannot be held liable

HR loves this one trick…

That will probably change sooner than you think. Now that it's commonly accepted* that AI has reached AGI status there are already movements to grant legal personhood and rights to it, which would mean legal liability for its actions. Eventually we get to the cyberpunk trope of entirely autonomous corporations that only ever hire humans for gig work and pay them in femtobitcoins or something.

* I'm not arguing what that actually means, much less, whether or not that has actually happened, just that it's what people believe, and as we are a belief-driven culture rather than a data or fact driven culture, that belief is what will drive policy.

That sounds like a horrific dream than anything. How will you even punish an AI model? I think the cyberpunk trope is a bit out of touch from reality, but a lighter version of it can definitely happen where we do get AI companies and entire board coups replacing CEOs with AI for short amounts of time, but in the end people (both consumers and investors) will want something to blame and punish properly, and they will also want control (realistically if you told any of the big people in the AI scene you're replacing them with a model they will definitely reject and claim they are needed for the company to prosper)

In addition this will probably only happen on America as most movies take place lol

Reply on news.ycombinator.com