Ask Lemmy

What I want from AI companies is really simple.

We have a thing called intellectual property in the United States of America. If I decided to make a Jellyfin instance that I charged access to, containing material I didn't own, somehow advertising this service on the stock market as a publicly traded company, you would bet your ass that I'd have a 1 way ticket to a defense seat in court.

AI companies, otherwise, operate entirely on data they don't own and don't pay licensing for ANY of the materials that are used to train their neural networks. So, in their eyes, any image, video (tv show/movie) or book that happens to be posted on the Internet is fair game in their eyes. This isn't how intellectual property works for individuals, so why exactly would a publicly traded company have an exception to this rule?

I work a lot in the world of FOSS and have a firm understanding that just because code is there doesn't make it yours. This is why we have the GPL for licensing. In fact, I'll take it a step further and say that the entirety of AI is one giant licensing nightmare, especially coding AI that isn't actually attributing license details with the code they're sampling from. (Sampling code being notably different than, say, learning from. Learning implies self-agency, and not corporate ownership.)

It feels to me that the AI bubble has largely been about pushing AI so hard and fast that people were investing in something with a dubious legal state in the US. Nobody stopped to ask whether or not the data that Facebook had on their website (for example, they aren't alone in this) was actually theirs to own, and what the repercussions for these types of decisions are.

You'll also note that Tech and Social Media companies are quick to take ownership of data when it benefits them (artists works, intellectual property that isn't theirs, random user posts about topics) and quick to deny ownership when it becomes legally burdensome (CSAM, illicit drug deals, etc.) to a degree that no individual would be granted. Hell, I'm not even sure a "small" tech startup would be granted this level of double-speak and hypocrisy.

With this in mind, I am simply asking that AI companies pay for the data that they're using to train AI. Additionally, laws must be in place that allows for the auditing of all materials used to train an AI with the legal intent of verifying that all parties are paid accordingly. This is how every other business works. If this were somehow granted an exception, wouldn't it be braindead easy to run every "service" through an AI layer in order to bypass any and all copyright laws?

Otherwise, if facebook and others want to claim that data hosted on their website is theirs to own and train off of -- well, great, but there should be no exceptions to this and they should not be allowed to host materials they then have no ownership over. So pictures of IP they don't own or materials they want to claim they have no ownership over must be removed from the platform. I would much prefer the first of these two options, however.

edit: I should note, that AI for educational purposes could be granted an exception for this under fair use (for university) but would still also be required to site all sources used to produce the works in question (which is normal for academics, in the first place.) and would also come with some strict stipulations on using this AI as a "product" (it would basically be moot, much like some research papers). This basically the furthest I'm willing to give these companies.

I want all of the CEOs and executives that are forcing shitty AI into everything to get pancreatic cancer and die painfully in a short period of time.

Then I want all AI that is offered commercially or in commercial products to be required to verify their training data and be severely punished for misusing private and personal data. Copyright violations need to be punished severely, and using copyrighted works being used for AI training counts.

AI needs to be limited to optional products trained with properly sourced data if it is going to be used commercially. Individual implementations and use for science is perfectly fine as long as the source data is either in the public domain or from an ethically collected data set.

Wishful thinking? Models trained on illegal data get confiscated, the companies dissolved, the ceos and board members made liable for the damages.

Then a reframing of these bs devices from ai to what they actually do: brew up statistical probability amalgamations of their training data, and then use them accordingly. They arent worthless or useless, they are just being shoved into functions they cannot perform in the name of cost cutting.

Training data needs to be 100% traceable and licensed appropriately.

Energy usage involved in training and running the model needs to be 100% traceable and some minimum % of renewable (if not 100%).

Any model whose training includes data in the public domain should itself become public domain.

And while we're at it we should look into deliberately taking more time at lower clock speeds to try to reduce or eliminate the water usage gone to cooling these facilities.

I'd like to see it used for medicine.

My biggest issue with AI is that I think it's going to allow a massive wealth transfer from laborers to capital owners.

I think AI will allow many jobs to become easier and more productive, and even eliminate some jobs. I don't think this is a bad thing - that's what technology is. It should be a good thing, in fact, because it will increase the overall productivity of society. The problem is generally when you have a situation where new technology increases worker productivity, most of the benefits of that go to capital owners rather than said workers, even when their work contributed to the technological improvements either directly or indirectly.

What's worse, in the case of AI specifically it's functionality relies on it being trained on enormous amounts of content that was not produced by the owners of the AI. AI companies are in a sense harvesting society's collective knowledge for free to sell it back to us.

IMO AI development should continue, but be owned collectively and developed in a way that genuinely benefits society. Not sure exactly what that would look like. Maybe a sort of light universal basic income where all citizens own stock in publicly run companies that provide AI and receive dividends. Or profits are used for social services. Or maybe it provides AI services for free but is publicly run and fulfills prosocial goals. But I definitely don't think it's something that should be primarily driven by private, for-profit companies.

I think two main things need to happen: increased transparency from AI companies, and limits on use of training data.

In regards to transparency, a lot of current AI companies hide information about how their models are designed, produced, weighted and use. This causes, in my opinion, many of the worst effects of current AI. Lack of transparency around training methods mean we don't know how much power AI training uses. Lack of transparency in training data makes it easier for the companies to hide their piracy. Lack of transparency in weighting and use means that many of the big AI companies can abuse their position to push agendas, such as Elon Musk's manipulation of Grok, and the CCP's use of DeepSeek. Hell, if issues like these were more visible, its entirely possible AI companies wouldn't have as much investment, and thus power as they do now.

In terms of limits on training data, I think a lot of the backlash to it is over-exaggerated. AI basically takes sources and averages them. While there is little creativity, the work is derivative and bland, not a direct copy. That said, if the works used for training were pirated, as many were, there obviously needs to be action taken. Similarly, there needs to be some way for artists to protect or sell their work. From my understanding, they technically have the legal means to do so, but as it stands, enforcement is effectively impossible and non-existant.

Firings and jail time.

In lieu of that, high fines and firings.

AI models produced from copyrighted training data should need a license from the copyright holder to train using their data. This means most of the wild west land grab that is going on will not be legal. In general I'm not a huge fan of the current state of copyright at all, but that would put it on an even business footing with everything else.

I've got no idea how to fix the screeds of slop that is polluting search of all kinds now. These sorts of problems ( along the lines of email spam ) seem to be absurdly hard to fix outside of walled gardens.

If we're talking realm of pure fantasy: destroy it.

I want you to understand this is not AI sentiment as a whole, I understand why the idea is appealing, how it could be useful, and in some ways may seem inevitable.

But a lot of sci-fi doesn't really address the run up to AI, in fact a lot of it just kind of assumes there'll be an awakening one day. What we have right now is an unholy, squawking abomination that has been marketed to nefarious ends and never should have been trusted as far as it has. Think real hard about how corporations are pushing the development and not academia.

Put it out of its misery.

I want real, legally-binding regulation, that’s completely agnostic about the size of the company. OpenAI, for example, needs to be regulated with the same intensity as a much smaller company. And OpenAI should have no say in how they are regulated.

I want transparent and regular reporting on energy consumption by any AI company, including where they get their energy and how much they pay for it.

Before any model is released to the public, I want clear evidence that the LLM will tell me if it doesn’t know something, and will never hallucinate or make something up.

Every step of any deductive process needs to be citable and traceable.

If we're going pie in the sky I would want to see any models built on work they didn't obtain permission for to be shut down.

Failing that, any models built on stolen work should be released to the public for free.

Part of what makes me so annoyed is that there's no realistic scenario I can think of that would feel like a good outcome.

Emphasis on realistic, before anyone describes some insane turn of events.

I would likely have different thoughts on it if I (and others) was able to consent my data into training it, or consent to even have it rather than it just showing up in an unwanted update.

I want people to figure out how to think for themselves and create for themselves without leaning on a glorified Markov chain. That's what I want.