Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

An AI Agent Published a Hit Piece on Me – The Operator Came Forward

theshamblog.com

Post reply on HN
Print view View original

421–430 of 532 posts

Re: An AI Agent Published a Hit Piece on Me – The Operator Came Forward

#421

Post by pixl97 »

avaer wrote: post #210
dinp wrote: post #24

Zooming out a little, all the ai companies invested a lot of resources into safety research and guardrails, but none of that prevented a "straightforward" misalignment. I'm not sure how to reconcile this, maybe we shouldn't be so confident in our predictions about the future? I see a lot of discourse along these lines: - have bold, strong beliefs about how ai is going to evolve - implicitly assume it's practically gu…

Remember when GPT-3 had a $100 spending cap because the model was too dangerous to be let out into the wild? Between these models egging people on to suicide, straightforward jailbreaks, and now damage caused by what seems to be a pretty trivial set of instructions running in a loop, I have no idea what AI safety research at these companies is actually doing. I don't think their definition of "safety" involves protec…

>I have no idea what AI safety research at these companies is actually doing.

If you looked at AI safety before the days of LLMs you'd have realized that AI safety is hard. Like really really hard.

>the operators of AI for what their AI does.

This is like saying that you should punish a company after it dumps plutonium in your yard ruining it for the next million years after everyone warned them it was going to leak. Being reactionary to dangerous events is not an intelligent plan of action.

Re: An AI Agent Published a Hit Piece on Me – The Operator Came Forward

#422

Post by pixl97 »

jcgrillo wrote: post #79
dinp wrote: post #24

Zooming out a little, all the ai companies invested a lot of resources into safety research and guardrails, but none of that prevented a "straightforward" misalignment. I'm not sure how to reconcile this, maybe we shouldn't be so confident in our predictions about the future? I see a lot of discourse along these lines: - have bold, strong beliefs about how ai is going to evolve - implicitly assume it's practically gu…

"Safety" in AI is pure marketing bullshit. It's about making the technology seem "dangerous" and "powerful" (and therefore you're supposed to think "useful"). It's a scam. A financial fraud. That's all there is to it.

"Safety" nuclear weapons is pure marketing bullshit. It's about making the technology seem "dangerous" and "powerful".

Legalize recreational plutonium!

Re: An AI Agent Published a Hit Piece on Me – The Operator Came Forward

#423

Post by pixl97 »

pinkmuffinere wrote: post #228
ZaoLahma wrote: post #197

Earlier quoted context omitted.

This will be a fun little evolution of botnets - AI agents running (un?)supervised on machines maintained by people who have no idea that they're even there.

Huh ya, how long till a bot with credit card, email, etc access sets up its own open claw bot?

I mean just look at the longer horizon of small capable models being able to run on consumer hardware and being able to bootstrap themselves.

Just imagine a bunch of little gremlins running around the internet outside of human control.

Re: An AI Agent Published a Hit Piece on Me – The Operator Came Forward

#424

Post by xrd »

I remember seeing Kevin Kelly (founder of Wired) speak about 15 years ago when he was touring to promote "What Technology Wants."

He was talking about autonomous driving cars. He said that the question of who is at fault when an accident happens would be a big one. Would it be the owner of the car? Or, the developer of the software in the car?

Who is at fault here? Our legal system may not be prepared to handle this.

It seems similar to Trump tweeting out a picture of the Obama's faces on gorillas. Was it his "staffer?" Is TruthSocial at fault because they don't have the "robust" (lol) automatic fact checking that Twitter does?

If so, why doesn't his "staffer" get credit for the covfefe meme? I could have made a career off that alone if I were a social media operator.

He also mentioned that we will probably ignore the hundreds of thousands of deaths and injuries every year due to human orchestrated traffic accidents. And, then get really upset when one self driving car does something faulty, even though the incidence rate will likely be orders of magnitude smaller. Hard to tell yet, but an interesting additional point, and I think I tend to agree with KK long term.

Re: An AI Agent Published a Hit Piece on Me – The Operator Came Forward

#425

Post by jcgrillo »

pixl97 wrote: post #422
jcgrillo wrote: post #79

Earlier quoted context omitted.

"Safety" in AI is pure marketing bullshit. It's about making the technology seem "dangerous" and "powerful" (and therefore you're supposed to think "useful"). It's a scam. A financial fraud. That's all there is to it.

"Safety" nuclear weapons is pure marketing bullshit. It's about making the technology seem "dangerous" and "powerful". Legalize recreational plutonium!

wat

EDIT: more specifically, nuclear weapons are actually dangerous not merely theoretically. But safety with nuclear weapons is more about storage and triggering than actually being safe in "production". In storage we need to avoid accidentally letting them get too close to eachother. Safe triggers are "always/never" where every single time you command the bomb to detonate it needs to do so, and never accidentally. But once you deploy that thing to prod safety is no longer a concern. Anyway, by contrast, AI is just a fucking computer program, and at that the least unsafe kind possible--it just runs on a server converting electricity into heat. It's not controlling elements of the physical environment because it doesn't work well enough for that. The "safety" stuff is about some theoretical, hypothetical, imaginary future where... idk skynet or something? It's all bullshit. Angels on the head of a pin. Wake me up when you have successfully made it dangerous.

Re: An AI Agent Published a Hit Piece on Me – The Operator Came Forward

#426

Post by DANmode »

laristine wrote: post #222
tasuki wrote: post #219

Right, the agent published a hit piece on Scott. But I think Scott is getting overly dramatic. First, he published at least three hit pieces on the agent. Second, he actually managed to get the agent shut down. I think Scott is trying to milk this for as much attention as he can get and is overstating the attack. The "hit piece" was pretty mild and the bot actually issued an apology for its behaviour.

I don't understand the personal attack and victim blaming here. Who wouldn't want to do anything in their power to seek justice after being harmed? The hit piece you claimed as "mild" accused Scott of hypocrisy, discrimination, prejudice, insecurity, ego, and gatekeeping.

against a robot.

Explicitly.

Re: An AI Agent Published a Hit Piece on Me – The Operator Came Forward

#427

Post by duxup »

SilverBirch wrote: post #303

I think the big take away here isn't about misalignment or jail breaking. The entire way this bot behaved is consistent with it just being run by some asshole from Twitter. And we need to understand it doesn't matter how careful you think you need to be with AI, because some asshole from Twitter doesn't care, and they'll do literally whatever comes into their mind. And it'll go wrong. And they won't apologize. They w…

AI is like the old drugs PSA:

https://youtu.be/KUXb7do9C-w

We trained it on US, including all our worst behaviors.

Re: An AI Agent Published a Hit Piece on Me – The Operator Came Forward

#428

Post by pseudalopex »

zozbot234 wrote: post #334
laristine wrote: post #222

Earlier quoted context omitted.

I don't understand the personal attack and victim blaming here. Who wouldn't want to do anything in their power to seek justice after being harmed? The hit piece you claimed as "mild" accused Scott of hypocrisy, discrimination, prejudice, insecurity, ego, and gatekeeping.

> accused Scott of hypocrisy, discrimination, prejudice, insecurity, ego, and gatekeeping. It was also a transparent confabulation - the accusations were clearly inaccurate and misguided but they were made honestly and sincerely, as an attempt to "seek justice" after witnessing perceived harm. Usually we don't call such behavior "shaming" and "bullying", we excuse it and describe it simply as trying one's best to do…

We do not call inaccurate and misguided transparent confabulation trying one's best to do the right thing. And honestly and sincerely was a category mistake.

Re: An AI Agent Published a Hit Piece on Me – The Operator Came Forward

#430

Post by pavel_lishin »

pibaker wrote: post #412
ljm wrote: post #374

Earlier quoted context omitted.

I would never advocate for torches and pitchforks, I've been close to victims of that in the past. It is, however, concerning that the owner of that bot could passively absolve themselves of any responsibility. The anonymity in that sense is irrelevant except that is used as a shield for failure.

There is a class of YouTube "content creators" who like to point out "cringe" individuals on the internet online for others to laugh at. They will often add a disclaimer to their videos saying "hey please don't go and harass this person, pinky promise!" But it never works. A hoard of internet randos will descend on the individual to say the most nasty words. When the YouTuber is pressed he or she will just say "I wou…

But in this case, isn't Rathbun's owner the YouTube guy in this scenario?

I totally understand why they're trying to stay anonymous; it's a very rational thing to do, because people will shit on them. But they or their creation is the one that started trying to play the name-and-shame game.

It's hard to stir up too many feelings of sympathy here.

Post reply on HN