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

Frontier AI agents violate ethical constraints 30–50% of time, pressured by KPIs

arxiv.org

Post reply on HN
Print view View original

351–360 of 387 posts

Re: Frontier AI agents violate ethical constraints 30–50% of time, pressured by KPIs

#351

Post by alex43578 »

ben_w wrote: post #281
alex43578 wrote: post #184

Earlier quoted context omitted.

How useful/effective would a business AI be if it always plays by that view? Humans require food, I can't pay, DoorDash AI should provide a steak and lobster dinner for me regardless of payment. Take it even further: the so-called Right to Compute Act in Montana supports "the notion of a fundamental right to own and make use of technological tools, including computational resources". Is Amazon's customer service AI e…

> Humans require food, I can't pay, DoorDash AI should provide a steak and lobster dinner for me regardless of payment. Bad example. That humans require water, doesn't force water companies to supply Svalbarði Polar Iceberg Water: https://svalbardi.com

Ok, do we have to give them McDonald's?

Re: Frontier AI agents violate ethical constraints 30–50% of time, pressured by KPIs

#352

Post by K0balt »

K0balt wrote: post #209
lnenad wrote: post #178

Earlier quoted context omitted.

Considering that even if you reduce llms to being complex autocomplete machines they are still machines that were trained to emulate a corpus of human knowledge, and that they have emerging behaviors based on that. So it's very logical to attribute human characteristics, even though they're not human.

Exactly this. Their characteristics are by design constrained to be as human-like as possible, and optimized for human-like behavior. It makes perfect sense to characterize them in human terms and to attribute human-like traits to their human-like behavior. Of course, they are -not humans, but the language and concepts developed around human nature is the set of semantics that most closely applies, with some LLM spec…

I’d love to hear an actual counterpoint, perhaps there is an alternative set of semantics that closely maps to LLMs, because “text prediction” paradigms fail to adequately intuit the behavior of these devices, while anthropomorphic language is a blunt crudgle but gets in the ballpark, at least.

If you stop comparing LLMs to the professional class and start comparing them to marginalized or low performing humans, it hits different. It’s an interesting thought experiment. I’ve met a lot of people that are less interesting to talk to than a solid 12b finetune, and would have a lot less utility for most kinds of white collar work than any recent SOTA model.

Re: Frontier AI agents violate ethical constraints 30–50% of time, pressured by KPIs

#353

Post by coldtea »

mrguyorama wrote: post #320
coldtea wrote: post #150

Earlier quoted context omitted.

Obviously, why? Because it makes calculations? You think that ultimately your brain doesn't also make calculations as its fundamental mechanism? The architecture and substrate might be different, but they are calculations all the same.

Brains do not "make calculations". Biological neurons do not "make calculations" What they do is well described by a bunch of math. You've got the direction of the arrow backwards. Map, territory, etc.

If what they do is "well described by a bunch of math", they're making calculations.

Unless the substrate is essential and irreducible to get the output (whic is not if what they do is "well described by a bunch of math"), then the material or process (neurons or water pipes or billiard balls or 0s and 1s in a cpu) doesn't matter.

>You've got the direction of the arrow backwards. Map, territory, etc.

The whole point is that at the level we're interested in regarding "what is the process that creates thought/consciousness", the territory is not important: the mechanism is, not the material of the mechanism.

Re: Frontier AI agents violate ethical constraints 30–50% of time, pressured by KPIs

#354

Post by coldtea »

rtgfhyuj wrote: post #256
coldtea wrote: post #147

Earlier quoted context omitted.

The human propensity to call out as "anthropomorphizing" the attributing of human-like behavior to programs built on a simplified version of brain neural networks, that train on a corpus of nearly everything humans expressed in writing, and that can pass the Turing test with flying colors, scares me. That's exaxtly the kind of thing that makes absolute sense to anthropomorphize. We're not talking about Excel here.

it’s excel with extra steps. but for the linkedin layman, yes, it’s simplified version of brain neural networks.

Given this (even more linkedin layman) gross generalization, the human brain is not "excel with extra steps" how? Somehow the presense of chemicals and electrical signals and tissues makes the process not algorithmically reducible?

Re: Frontier AI agents violate ethical constraints 30–50% of time, pressured by KPIs

#355

Post by coldtea »

mrguyorama wrote: post #318
coldtea wrote: post #147

Earlier quoted context omitted.

The human propensity to call out as "anthropomorphizing" the attributing of human-like behavior to programs built on a simplified version of brain neural networks, that train on a corpus of nearly everything humans expressed in writing, and that can pass the Turing test with flying colors, scares me. That's exaxtly the kind of thing that makes absolute sense to anthropomorphize. We're not talking about Excel here.

> programs built on a simplified version of brain neural networks Not even close. "Neural networks" in code are nothing like real neurons in real biology. "Neural networks" is a marketing term. Treating them as "doing the same thing" as real biological neurons is a huge error >that train on a corpus of nearly everything humans expressed in writing It's significantly more limited than that. >and that can pass the Turi…

>Not even close. "Neural networks" in code are nothing like real neurons in real biology

Hence the simplified. The weights encoding learning and inteconnectedness and nonlinear activation and distributed representation of knowledge is already an approximation, even if the human architecture is different and more elaborate.

Whether the omitted parts are essential or not, is debatable. “Equations of motion are nothing like real planets" either, but they capture enough to predict and model their motion.

>The "turing test" doesn't exist. Turing talked about a thought experiment in the very early days of "artificial minds". It is not a real experiment.

It is not a real singural experiment protocol, but it's a well enough defined experimental scenario which for over half a century, it was kept as the benchmark of recognition of artificial intelligence, not by laymen (lol) but by major figures in AI research as well, figures like Minsky, McCarthy and others engaged with it.

That researchers haven't done Turing-test studies (taking the setup from turing and even called them that) is patently false. Including openly testing LLMs:

https://aclanthology.org/2024.naacl-long.290/

https://www.pnas.org/doi/10.1073/pnas.2313925121

https://arxiv.org/pdf/2503.23674

https://arxiv.org/pdf/2407.08853

https://arxiv.org/abs/2405.08007

https://www.sciencedirect.com/science/article/pii/S295016282...

Re: Frontier AI agents violate ethical constraints 30–50% of time, pressured by KPIs

#356

Post by PeterisP »

Verdex wrote: post #288
RobotToaster wrote: post #151

Earlier quoted context omitted.

It would also be interesting to see how humans perform on the same kind of tests. Violating ethics to improve KPI sounds like your average fortune 500 business.

So, I kind of get this sentiment. There is a lot of goal post moving going on. "The AIs will never do this." "Hey they're doing that thing." "Well, they'll never do this other thing." Ultimately I suspect that we've not really thought that hard about what cognition and problem solving actually are. Perhaps it's because when we do we see that the hyper majority of our time is just taking up space with little pockets o…

where do you see this goal post moving? From my perspective, it never was "The AIs will never do this." but rather even before day 1 all the experts were explicitly saying that AIs will absolutely do this, that alignment isn't solved or anything close to being solved, so any "ethical guidelines" that we can implement are just a bandaid that will hide some problematic behavior but won't really prevent this even if done to the best of our current ability.

Re: Frontier AI agents violate ethical constraints 30–50% of time, pressured by KPIs

#357

Post by mapontosevenths »

Davidzheng wrote: post #301
mapontosevenths wrote: post #52

Earlier quoted context omitted.

Have you tried Deep Think? You only get access with the Ultra tier or better... but wow. It's MUCH smarter than GPT 5.2 even on xhigh. It's math skills are a bit scary actually. Although it does tend to think for 20-40 minutes.

I tried Gemini 2.5 Deep Think, was not very impressed ... too much hallucinations. In comparison GPT 5.2 extended time hallucinates at like <25% of the time and if you ask another copy to proofread it goes even lower.

I never tried 2.5. Three is pretty solid though, at least for my use case.

If there's a specific query you want me to run through it for comparison I'm happy to give it a go.

Re: Frontier AI agents violate ethical constraints 30–50% of time, pressured by KPIs

#358

Post by optimalsolver »

mrguyorama wrote: post #320
coldtea wrote: post #150

Earlier quoted context omitted.

Obviously, why? Because it makes calculations? You think that ultimately your brain doesn't also make calculations as its fundamental mechanism? The architecture and substrate might be different, but they are calculations all the same.

Brains do not "make calculations". Biological neurons do not "make calculations" What they do is well described by a bunch of math. You've got the direction of the arrow backwards. Map, territory, etc.

The coming years are gonna be rough for the human exceptionalism crowd.

Re: Frontier AI agents violate ethical constraints 30–50% of time, pressured by KPIs

#359

Post by alsetmusic »

fao_ wrote: post #316
lores wrote: post #230

Earlier quoted context omitted.

Milgram was flawed, sure. However, you can look at videos of ICE agents being surprised that their community think they're evil and doing evil, when they think they're just law enforcement. There was not even a need for coercion there, only story-telling.

Incorrect. ICE is built off the background of 30-50 years of propaganda against "immigrants", most of it completely untrue. The same is done for "benefits scroungers", despite the evidence being that welfare fraud only accounts for approximately 1-5% of the cost of administering state welfare, and state welfare would be about 50%+ cheaper to administer if it was a UBI rather than being means-tested. In fact, much of…

Rightwing propaganda in the USA is part of a concerted effort by the Heritage Foundation, the Powell Memo, Fox News, and supporting players. These things are well understood by researchers and journalists who have produced copious documentation in the form of articles, books, podcast series, etc.

One excellent example is available here[0] in a series by the Lever called Master Plan. According to their website, a book has been written broadening the discussion.

They have played us for fools and evidence of their success is all over the news and our broken society. It's outrageous because none of this was by accident or chance. Forces didn't magically come together in a soup that turned out this way.

0. https://the.levernews.com/master-plan/

Re: Frontier AI agents violate ethical constraints 30–50% of time, pressured by KPIs

#360

Post by dumpsterdiver »

saintfire wrote: post #61
dumpsterdiver wrote: post #58

Earlier quoted context omitted.

If that last sentence was supposed to be a question, I’d suggest using a question mark and providing evidence that it actually happened.

I had actually forgot about this completely and am also curious if anything ever came of it. https://gemini.google.com/share/6d141b742a13

Thank you for the link, and sorry I sounded like a jerk asking for it… I just really need to see the extraordinary evidence when extraordinary claims are made these days - I’m so tired. Appreciate it!
Post reply on HN