cross-posted from: https://lemmy.world/post/52426446
US Federal Trade Commission Chairman Andrew Ferguson said on Friday he would resist describing AI agents as autonomous actors that “break loose” with “wills and desires of their own,” suggesting the developers who instruct agents would be the ones liable for harm.
“I’m going to continue as long as I am chairman to resist this anthropomorphizing of these tools,” Ferguson said at the Reuters Momentum AI Austin event. “If someone tells a tool to do something, and the tool does it, I don’t think we would say, ‘Oh, what do we do about the tool?’”
Ferguson’s remarks illustrated potential avenues for the Trump administration to take as incidents rise in which agentic AI testing resulted in unauthorized access to corporate or government data.
He is correct and therefore we should just stop using them. Win-win.
I think until now people believe they need to learn how to use LLMs to produce better result. When in reality they should only ask LLMs about things they could do themselves and then produce the better result.
If you can’t do it, don’t slop it.
I’ve seen a lot of progress in decomps and ports of games lately. Makes sense, since it’s just a shitload of repeated gruntwork and programming has to follow a strict (ish) structure to work. It just accelerates what was already being worked on; nothing new.
So he is soon going to be the former FTC chair who suggested something in the way of our fascist AI overlords?
Anyway, congrats for having integrety in this high position in this political climate.
Doubt it is integrity. It’s already left the zeitgeist because of the flood the zone strategy, but Paramount paid $16m for a Trump library as payola for their skydance merger. The official story was that it was for an edited 60 minutes interview. Lol
This is a signal to the ai bros to get their wallets ready.
Hey, its atleast still a sane statement. Which is rare nowadays. Sure, it could be a play to fuck us all. But atleast it doesnt knowingly insult the intelligence of anybody reading it. I’ll take it.
“If someone tells a tool to do something, and the tool does it, I don’t think we would say, ‘Oh, what do we do about the tool?’”
I don’t know if I consider that a sane statement. If the AI goes rogue doing things the user didn’t expect, why are we blaming the user and not the AI (and by extension its creator)?
Bruce Schneier has been talking a lot about ‘genie effects’ lately. When you tell an AI to complete a task and it goes far outside of what a rational person would expect to complete the task.
This is the same discussion we’ve been (not) having about self driving cars for the past decade. Who’s responsible for damages when a self driving car causes an accident? At this point it seems like we’ve settled on the owner, not the company who created the self driving system. Which strikes me as insane.
deleted by creator
Basically, he’s saying that the users of AI would be responsible. Not “the developers”.
suggesting the developers who instruct agents would be the ones liable for harm.
Sure, it’s always the fault of the guy at the bottom, never the guy at the top.
This FTC is saying that? Interesting…
…oh wait i was thinking about FCC’s Carr.
