[ronsor]

I don't.

Software developers and researchers should not be liable for distributing information or code, even if it's used for something illegal, as long as they aren't explicitly promoting the illegal activity and don't have any involvement with it outside of creating the software.

Not only is that consistent with previous decisions, such as those regarding copyright (i.e. torrents are fine, but making a client to torrent movies specifically isn't), but also any other decision would be a violation of the social contract with regard to open-source development.

[autoexec]

If a bridge collapses and people are hurt the engineer is at fault and should be held accountable. If software fails and people are hurt the software engineer is at fault and should be held accountable.

[ronsor]

This is a poor analogy. A better one would be if a murderer used your bridge to escape. Should you be held liable? What if the bridge were designed to handle highway speeds so he could escape faster?

[autoexec]

I'd agree that they shouldn't be liable in that case, since the bridge works the same for everybody (no matter why they're driving over it) and is working as designed/intended. It's really only the idea that developers shouldn't be liable for their code as long as they aren't explicitly promoting illegal activity and don't have any involvement with it outside of creating the software that I take issue with.

[ronsor]

General AI tools also work the same for all users. It's not as if the average AI company is optimizing for celebrity deepfake nudes or spambots.

I mean, there are really two categories of software:

* Free and/or open source software. In this case, I think there is no good reason to make the developer liable, unless they're promoting illegal use. No person wants to be attacked for giving away something for free. That's why the LICENSE.

* Commercial/paid software. In this case, it is reasonable to argue that companies should be liable if end users are harmed by the software. For paid software especially, disclaimers cannot be absolute.

But I do not think it is acceptable to hold developers liable for second-order effects - i.e., a user doing something illegal with the software and harming a third party - unless it was obvious to them that the user was going to do something illegal.

[asadotzler]

If they are knowingly including large numbers of celebrity photos in their training data, slurping it into their models, and doing nothing to block users from abusing what is a clearly foreseeable harm? That's on the companies making the product, not on the users.

If Honda put a big spike on the front of their vehicles because they thought it looked good and would sell more cars, but the spike was good at skewering pedestrians, they'd be at fault too. It wouldn't matter that their designers thought the spike was sexy and would sell more cars. You can't make something you know to be dangerous and expect to sell it to the public without being regulated.

Want to avoid the regulation, don't steal a bunch of celebrity photos and an provide your users with a tool that that creates celebrity porn deepfakes on demand.

This isn't controversial. Go to Microsoft's AI chatbot today and try to get it to create a naked image of Taylor Swift. Microsoft has spent non-trivial engineering resources making that fail. Not doing that work is irresponsible and likely to lead to lawsuit that may or may not be winnable but that Microsoft and others clearly want to avoid.

[ronsor]

Counterpoint: tons of tools are dangerous yet are still sold without much if any regulation. Knives are dangerous, but you don't need an ID to buy one from the store. We sell dangerous products all the time! We just put warnings and disclaimers on them (which AI models tend to come with).

That said, I dispute the idea that these models are "dangerous" in the first place. A box that generates texts and images is not even remotely as dangerous as a sharp spike strapped to a car. Such a comparison is hyperbolic.

People act like these models are going to be the end of US when they're literally just "instant photoshop." A dangerous model would be one designed to run a military drone or automatic weapons, not a random text and image machine.

All that aside, the deepfake issue has nothing to do with the model datasets including celebrity photos (in fact, it would work fine without any of them). And no, downloading public photos is not stealing either.

[theamk]

the original comment explicitly said "holding developing firms ..." - so this is not about software developers, it is about corporations. The moment you start to sell stuff is the moment you become liable.

[asadotzler]

HN doesn't know how to make that distinction any more. It's so overrun with corporate bootlickers who think the software engineers ARE the company and the company IS the software engineers. I presume it's just a bunch of temporarily embarrassed billionaires planning for the future, but it's a shame that a once hacker-friendly forum is now mostly focused on compensation maximization and defending trillion dollar corporations.

[bluGill]

Torrents are fine because there are legal purposes. Developers of torrent software are very careful to emphasize the legal uses.

[ronsor]

There are legal purposes for generative AI tools and even deepfakes, so there should be no issues with the tools themselves.

Obviously if a site promotes "download this tool to generate infinite nude pictures of celebrities", then that is illegal, since that particular tool was only developed for illegal uses.

[asadotzler]

There are legal purposes for cars and guns and we regulate the hell out of them because there are also plenty of not-legal purposes and even just the potential for accidents. When A.I. is as heavily regulated as cars, we can revisit the "it's just a tool" argument.

[ronsor]

We can revisit the argument when an image generation AI causes a fatal accident with 5 other vehicles.