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

OpenAI backs Illinois bill that would limit when AI labs can be held liable

wired.com

Post reply on HN
Print view View original

311–320 of 343 posts

Re: OpenAI backs Illinois bill that would limit when AI labs can be held liable

#311

Post by _doctor_love »

lepset wrote: post #303
_doctor_love wrote: post #280

Earlier quoted context omitted.

> Like if you have a product, and the government says the product is ok, and it's labeled per regulation and later that product turns out to be deleterious to people's health should the company be liable? But like, what if you like, totally bribed the shit out government people and like totally fabricated scientific evidence to make it seem like it was safe but then you sold it anyway? Aren't you then like a total pi…

It's going to be okay. Stop watching movies and thinking they correlate to real life.

Damn. Wish you'd told me that before I finished Schindler's List and Angela's Ashes.

Re: OpenAI backs Illinois bill that would limit when AI labs can be held liable

#312

Post by repeekad »

thegrimmest wrote: post #113
repeekad wrote: post #80

Earlier quoted context omitted.

Because if you didn’t already know that, like an immature deprived and desperate kid, being able to easily find out is really really bad.. Plenty of lazy AI apps just throw messages into history despite the known risks of context rot and lack of compaction for long chat threads. Should a company not be held liable when something goes wrong due to lazy engineering around known concerns?

No, because that would indicate there should be some sort of regulatory standard for what does/does not constitute "lazy engineering". Creating this standard in turn creates regulatory/compliance overhead for every software engineering organization. This in turn slows everything right down and destroys the startup ethos. "Move fast and break things" is a thing for a reason. The whole point of the free market is to av…

If airplanes used this logic, likely at least hundreds more would have died over the last decades. Accident rates are even going up, because of logic like yours. Yeah planes are fine most of the time, when the long tail involves safety concerns (that wouldn’t have otherwise happened) making money on people using your product becomes unethical without mutually agreed upon safety regulation, ideally motivated by voters instead of special interest groups

Re: OpenAI backs Illinois bill that would limit when AI labs can be held liable

#313

Post by repeekad »

morpheuskafka wrote: post #134
repeekad wrote: post #80

Earlier quoted context omitted.

Because if you didn’t already know that, like an immature deprived and desperate kid, being able to easily find out is really really bad.. Plenty of lazy AI apps just throw messages into history despite the known risks of context rot and lack of compaction for long chat threads. Should a company not be held liable when something goes wrong due to lazy engineering around known concerns?

> to lazy engineering around known concerns? That implies that it is already illegal to provide this information. But is it? If a human did so with intent to further a crime, it would be conspiracy. But if you were discussing it without such intent (e.x. red teaming/creating scenarios with someone working in chemistry or law enforcement), it isn't. An AI has no intent when it answers questions, so it is not clear how…

Very simply, if you provide a service for money you have a duty to ensure that service is safe. There’s a reason you have to sign a waiver when you jump on a trampoline, but companies are so rich the court cases have become parking tickets..

Re: OpenAI backs Illinois bill that would limit when AI labs can be held liable

#314

Post by RevEng »

lepset wrote: post #303
_doctor_love wrote: post #280

Earlier quoted context omitted.

> Like if you have a product, and the government says the product is ok, and it's labeled per regulation and later that product turns out to be deleterious to people's health should the company be liable? But like, what if you like, totally bribed the shit out government people and like totally fabricated scientific evidence to make it seem like it was safe but then you sold it anyway? Aren't you then like a total pi…

It's going to be okay. Stop watching movies and thinking they correlate to real life.

Cigarettes and asbestos are just two examples of where this absolutely did happen with plenty of public evidence, including successful legal cases.

Re: OpenAI backs Illinois bill that would limit when AI labs can be held liable

#315

Post by RevEng »

jamesbfb wrote: post #102

No different to preventing game studios being liable for mass shootings. Reminds me of the post-Columbine hysteria where media was super critical of Doom and Nine Inch Nails.

Except that there was no link between violent video games and mass shootings. It has been shown many times. The families blamed the video games despite there being no direct evidence of it. The games did not tell them they should go on a mass murdering spree any more than porn told you you should sleep with your cousin.

In this case there are several documented cases where the AI did over a long period of time progressively gaslight the individual and persuade them to commit suicide.

Re: OpenAI backs Illinois bill that would limit when AI labs can be held liable

#316

Post by infamouscow »

kusokurae wrote: post #63

Without getting even more eyes on me, these company boards are inadequately scared for their personal safety.

If sufficient numbers of the population perceive to have lost their livelihoods due to AI, then I'd expect to see data centers burned to the ground and a lot of people swinging from lamp posts. Jury nullification solves the rest, but even that that assumes you can even find an impartial jury.

Re: OpenAI backs Illinois bill that would limit when AI labs can be held liable

#317

Post by autoexec »

recursive wrote: post #297
imnotlost wrote: post #278

Am I alone in thinking this is easy? The human making the decision is always liable. What if the human couldn't reasonably know better? Doesn't matter - If they made the same decision without AI or with old files it is still on them. What if there's no single human decision? Someone is in charge and is responsible. The "I was ordered to" isn't a defense. Does liability without power make sense? People executing have…

> Someone is in charge and is responsible. This doesn't seem to be a given.

There's always a CEO or a president. The buck always stops somewhere. Somebody is always making the big bucks because they are in charge.

Re: OpenAI backs Illinois bill that would limit when AI labs can be held liable

#318

Post by recursive »

autoexec wrote: post #317
recursive wrote: post #297

Earlier quoted context omitted.

> Someone is in charge and is responsible. This doesn't seem to be a given.

There's always a CEO or a president. The buck always stops somewhere. Somebody is always making the big bucks because they are in charge.

For legally incorporated companies. What if it's just a crypto wallet hooked up to a network of prompts? Are you sure a human must have created it?

Re: OpenAI backs Illinois bill that would limit when AI labs can be held liable

#319

Post by estimator7292 »

mystraline wrote: post #196
WarmWash wrote: post #124

Earlier quoted context omitted.

A Michelin chef can give you their recipe, and give you their ingredients, but you still will fail miserably trying to match their dish. It's the same with drugs, whose instructions and ingredient lists have been a google search away for decades now. Yet you still need a master chemist to produce anything. By the time AI can hand hold an idiot through the synthesis of VX agents (which would require an array of sensor…

That is completely wrong. Food preparation, like pharmaceutical drug fabrication, is inherently scientific and methodologically controllable. Look no further than the Four Thieves Vinegar Collective. Original synthesis line construction is hard. But the exact formula "add this", "turn on stir bar", "do you see particulate? Yes for +10m at stir", etc. And if their results are replicated, theyre seeing 99.9% yields, co…

Spoken like someone who has never had to actually do these things in real life.

Recipes and formulae do not encode all the minutiae and expertise required to reproduce them. You can tell someone to sear a steak at whatever temperature for however long, but you can't encode the skill and experience required to reproduce in arbitrary conditions. One must learn what a correctly seared steak looks, feels, and tastes like and how to achieve that on uncalibrated cooking equipment.

Your assertion only holds true in a vacuum. If 100% of inputs, materials, environmental conditions are completely standardized and under control then sure, you can follow step by step instructions. The real world does not work that way. No stove on the market is calibrated. Reagents come with impurities. Your skillet may not conduct heat as well as expected or your mains electricity might be low causing your mantle to heat slower and your stir rod to stir slower.

These are things that one has to learn and experience in order to compensate for.

Re: OpenAI backs Illinois bill that would limit when AI labs can be held liable

#320

Post by janalsncm »

jtbayly wrote: post #309
janalsncm wrote: post #292

Earlier quoted context omitted.

Stop no, reduce yes. Same with how seatbelts didn’t end car crash fatalities. Eliminating drunk driving also wouldn’t end road fatalities, but outside of Wisconsin no one believes drunk drivers shouldn’t be in jail.

Those comparisons don’t make sense. There are millions of guns in the country, and cutting US manufacturing wouldn’t change that any time soon.

> any time soon

Mandating seatbelts on new cars also did not immediately stop people from flying through their windshields.

Post reply on HN