Ask Lemmy

Lately, I just wish it didn't lie or make stuff up. And after drawing attention to false information, it often doubles-down, or apologises, and just repeats the bs.

If it doesn't know something, it should just admit it.

2 chicks at the same time.

Just Mass public hangings of tech Bros.

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.

I would love to see regulation, that any contet created by AI cannot be used commercially.

I love e.g. that parents can make their own children books, but nobody should profit from all the stolen work of artists.

Not much, just don't build it over theft.

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.

Our current 'AI' is not AI. It is not.

It is a corporate entity to shirk labor costs and lie to the public.

It is an algorithm designed to lie and the shills who made it are soulless liars, too.

It only exists for corporations and people to cut corners and think they did it right because of the lies.

And again, it is NOT artificial intelligence by the standard I hold to myself.

And it pisses me off to no fucking end.

I personally would love an AI personal assistant that wasn't tied to a corporation listening to every fkin thing I say or do. I would absolutely love it.

I'm a huge Sci-Fi fan, so sure I fear it to a degree. But, if I'm being honest, AI would be amazing if it could analyze how I learned math wrong as a kid and provide ways to fix it. It would be amazing if it could help me routinely create schedules for exercise and food and grocery lists with steps to cook and how all of those combine to effect my body. It would be fantastic if it could point me to novels and have a critical debate about the inner works with a setting of being a contrarian or not so I can seek to deeply understand the novels.

It sounds like what our current state of AI has right? No. The current state is a lying machine. It cannot have critical thought. Sure, it can give me a schedule of food/exercise, but it might tell me I need to lift 400lbs and eat a thousand turkeys to meet a goal of being 0.02grams heavy. It might tell me 5+7 equals 547,032.

It doesn't know what the fuck it's talking about!

Like, ultimately, I want a machine friend who pushes me to better myself and helps me understand my own shortcomings.

I don't want a lying brick bullshit machine that gives me all the answers but they are all wrong because it's just a guesswork framework full of 'whats the next best word?'

Edit: and don't even get me fucking started on the shady practices of stealing art. Those bastards trained it on people's hard work and are selling it as their own. And it can't even do it right, yet people are still buying it and using it at every turn. I don't want to see another shitty doodle with 8 fingers and overly contrasted bullshit in an ad or in a video game. I don't want to ever hear that fucking computer voice on YouTube again. I stopped using shortform videos because of how fucking annoying that voice is. It's low effort nonsense and infuriates the hell out of me.

I think Meta and others went open with their models as firewall protection against legal action due to their blatant stealing of people's work to train with. If the models has stayed commercial and controlled within the company, they could be (probably still wouldn't be, but could be) forced to shut down or start over properly. But it's far too late now since it's everywhere there is a GPU running, even if models don't progress past current state.

That being said, not much is getting done about the safety factors. Yes, they are only LLMs and not AGI, but there's commonality in regards to not being sure what's going on inside the box and if it's really doing what it's told to do. Now is the time boundaries and research should be done, because once something happens (LLM or AGI) it's too late. So what do I want to see happen? Heavy regulation and transparency on the leading edge of development. And stop the madness of more compute being the only solution with its environmental effects. It might be the only solution, but companies are going that way because it's the easiest way to throw money at a problem and reap profits, which is all they care about.

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'd like to see it used for medicine.

I'm not super bothered by Tue copyright issue - the copyright system is barely serving people these days anyway. Blow it up.

I'm deeply troubled by the obscene power use. It might be worth it if it was a good tool. But it's not.

I haven't gone out of my way to use AI anything, but it's been stuffed into everything. And it's truly bad at it's job. AI is like a precocious 8-year-old, butting into every conversation. And it gives the right answer at about the rate a ln 8-year-old does. When I do a web search, I then need to do another one to check the AI's answer. Or scroll down a page to get past the AI answers to real sources. When someone uses it to summarize a meeting, I then need to read through that summary to make sure the notes are accurate. And it doesn't know to ask when it doesn't understand something like a proper secretary would. When I go looking for reference images, I have to check to make sure they're real and not hallucinations.

It gets in my way and slows me down. It needed at least another decade of development before being deployed at all, never mind at the scale it has, and it needs to be opt-in, not crammed into everything. And until it can be relied on, it shouldn't be allowed to suck down as much electricity as it does.

Ideally the whole house of cards crumbles and AI goes the way of 3D TV's, for now. The world as it is now is not ready for AGI. We would quickly end up in a " I have no mouth and I must scream" scenario.

Otherwise, what everyone else has posted are good starting points. I would just add that any data centers used for AI have to be powered 100% by renewable energy.

https://mastodon.social/@jessie/114449062193857643