AI PBs
-
-
@Pavel There’s not even internal consistency, forget about citations.
-
Reality_is_often_disappointing.jpg
There’s not even internal consistency, forget about citations.
You’re just not patient enough to see the big picture. We are all trapped on this train. It’s going to crash wether we like it or not. Just sit back and enjoy the tea while you still have service.
It’s going to suck for a while but long after you and I are dead that bright future will happen. Just not for us.
-
-
@Pavel There’s only so much you can do with a string of disjointed statements.
-
-
I’m confused. Is genAI a fun, harmless tool to be more efficient at creating art, or is it the catalyst for the race to the bottom, a “painful and violent” future for humanity? I feel like the original argument deviated a bit.
-
I’m confused. Is genAI a fun, harmless tool to be more efficient at creating art, or is it the catalyst for the race to the bottom, a “painful and violent” future for humanity? I feel like the original argument deviated a bit.
Because it has. Red rocket is just gish galloping and moving goalposts. Very little reason to engage them seriously, as they don’t seem interested in anything besides bad faith contrarianism (see: their other posts). Walls of weird, miserable nonsense.
-
@Yam Earnestly? Right now, it has elements of both. Though “creating art” might be a bit of a stretch, but “creating media” will do as a substitute: Philosophical arguments can and will continue on that point.
It can be a plaything used to make an image for a character you don’t care about enough to pay an artist to render for you, or construct an essay scaffold. It also can be a service used by malicious, incompetent, stupid (or all three), companies and institutions to replace human labour as a cost-cutting measure without care or understanding about the dangers such a replacement will have.
And this isn’t even mentioning the ethical qualms of using unknown, potentially IP-law violating information in training data, the intense energy costs that lead to environmental damage, or other as-yet unknown unknowns.
ETA: It is a complex and nuanced topic, many aspects of which will need to be decided in courts of law and other such places. Anyone speaking with absolute certainty on the topic (beyond certainty in their own opinions) should be looked upon with askance.
-
I’m here trying to figure out how AI image generation is gonna reduce menial labour for me. What is it gonna do? Make an image of someone doing my laundry???
-
@Juniper That’s part of the disconnect I think, at least if I take arguments at face value. At least one party to this argument views the actual “creating” part of “creating art” to be menial labour.
-
I’m confused. Is genAI a fun, harmless tool to be more efficient at creating art, or is it the catalyst for the race to the bottom, a “painful and violent” future for humanity? I feel like the original argument deviated a bit.
It can be both.
I think it is both. Especially since AI continues to grow and evolve.
I find AI exceptionally helpful and useful. It has saved me time, effort, and energy and has allowed me to finish menial task faster (work and fun related) so that I have more free time to enjoy other things. and it has allowed other enjoyable things to be even more enjoyable than they were before. To me, AI has been amazing and exciting.
I laugh at the people who complain about AI’s faults and errors. It is like criticizing a toddler for making errors on the bar exam. AI is still in its infancy and will continue to grow better at everything, for good or ill. Just like a child.
Remember dial up? How is internet now? Remember Pong? How’s that compare to Balder’s Gate 3? Knock it and get your jokes in now while you can. It’s just going to get better at everything by orders of magnitude. (Meanwhile, companies are also improving quantum computing at an alarming rate. Don’t even consider if these two paths meet…)
That being said…
I also think that it will bring more harm than good the more it evolves.
I think this, not because of what AI can do or will do, but because of what people will do with it. I don’t think the problem is AI. As with most things, the problem is people.
We don’t need AI to solve world hunger. We could do that, easily, already if the right people wanted to. They just don’t. The same issues of what people will do with power will only worsen with AI, but that isn’t the fault of the technology.
People are the worst. Plain and simple.
You can argue about the legality of what AI does all day long, but the law is a joke. If anything, we’ve learned over the last decade that the law is whatever the people in power decide it is. Laundering cartel money is a crime. The punishment is paying a fine that doesn’t even scratch the surface of the profits made from it. OpenAI is being sued by god knows how many people. Even if they all win (they won’t), is the punishment fine gonna shut down OpenAI? Even if it did (it won’t), would that even put a dent in the AI community considering how many other AI companies are out there and how many continue to be formed every month/week/day?
Again, the problem isn’t AI. The problem is people.
If AI can solve that problem, the world will be better off.
(And yes, I know that this is not a new conclusion. This is the plot of many a sci-fi movie. And for good reason. I’m just not sure that they got the hero/villain roles correct.)
-
-
Cue every AI movie/series compilation to the tune of We Didn’t Start the Fire.
Grand. Now I have to make a watch list starting with 2001: A Space Oddesey and all the way through to Zardoz or Zoe.
-
@Warma-Sheen said in AI PBs:
Again, the problem isn’t AI. The problem is people.
Of course the problem is people, but imagine what the world would be like if we’d taken that stance with other technological advancements. We can’t un-invent LLMs any more than we can un-invent the steam engine, but that doesn’t mean we have to let tech companies run rampant either.
The industrial revolution caused a whole lot of chaos before we had reform and regulations to make it better. And for all the faults of the modern world, things are better in countless ways than they were in the 1870s.
Copyright law came about after the invention of the printing press, to recognize that just because you had a copy of a book, that didn’t give you the right to re-print it and sell your own copies. Those laws have evolved over time, but the core idea has remained the same: it’s not right to make money off of someone else’s creative work, and it’s not good for the world if artists have no incentive to share their art.
-
-
but that doesn’t mean we have to let tech companies run rampant either.
That’s a tricky phrase. I’m not sure the context you mean it in.
At this point it is highly unlikely, in any practical sense, that we can do anything about tech companies running rampant.
When has that ever happened before in any meaningful way?
and it’s not good for the world if artists have no incentive to share their art.
The artist issue is one that I have an unpopular opinion about, at least on this forum, but that statement is definitely a problematic opinion.
It suggests that the only, or main, incentive for creating art is financial gain. But there are many other reasons that art is created. The trope of the starving artist is a trope for a reason. Many artists do art just for art’s sake, even if there is no money in it, which, in most cases, there is not. Art for art’s sake is the motto of MGM Studios, despite the mountains of cash they make in entertainment. If you’re only creating art for financial incentive, you’re not being displaced by AI, you’re just competing with it in a market that’s historically been brutal for artists. It just means that your art has to be subjectively better than what AI can produce - by any possible metric. You have a soul. AI doesn’t. You can truly create. AI can’t. These are all the arguments made, but at the end of the day, does that make your art more marketable than AI art? If the reason you do art it is financial, you better hope so.
Is it easier to do art when you’re getting paid for it? Sure. But art is hardly the only medium that AI is taking over and I don’t know why it gets romanticized as a protected class that needs saving, exempt from the same pressures that affect every other job in a capitalist system. I’d love to game, or dance, or write all day instead of working a job, but there isn’t any money in it for 99.9% of people who can do it. So I don’t do it. The world I live in sets the reality I have to operate within. And AI is now part of the world that we live in.
But on the flip side, there are plenty of venues where computers are better but haven’t taken over. Computers can play chess and beat grand masters, but they haven’t stopped chess tournaments. AI can outplay most people in competitive video games, but e-sports hasn’t crumbled.
Again, the problem isn’t AI. The problem is people. When people value human work/creation/skill, there will be money to follow. If they don’t, there won’t be. There are tons of concern for artists put out of work by AI, but not much for the customer service reps that are being cut at a far higher rate. Hasbro’s Dungeon’s and Dragons tribulations are a battle ground for that right now. They first said they wouldn’t use AI art, then they got caught using AI art, then they pretty much abandoned their pledge to not use AI. There was a big ado and calls for boycotts. But do people still buy their product? Yes. Absolutely.
If people value art made by other people, they’ll find it and pay for it. If they don’t, no amount of regulation or gatekeeping will save it. AI doesn’t kill art. People choosing convenience, price, or novelty does. And because people are the worst, artists suffer.
And copyright law… well, that’s also a joke, like most laws. The punchline here is that the “law” is heavily favored to big corporations and companies with expensive legal teams that make it increasingly easy to steal from individuals who can’t afford to fight back in court. It definitely does not favor the majority of artists, most of whom are forced to sell the rights to their art to some soulless company or corporation “in perpetuity and throughout the universe” in order to make scraps of money from it.
The industrial revolution caused a whole lot of chaos before we had reform and regulations to make it better. And for all the faults of the modern world, things are better in countless ways than they were in the 1870s.
I agree.
The Industrial Revolution happened over many decades. But it took just as many decades for laws and regulations to catch up to what has happening - for many of the same reasons it won’t work today. Too much money and influence on the side of the people with the new toys. In the meantime people were brutally and inhumanly pushed past their limits for generations with what we’d today consider abuses in order to feed the industrial machine. But here, today, AI isn’t playing out over decades, it’s evolving over months, across dozens of industries.
Governments are barely able to define what an LLM is, let alone agree on how to regulate it. Meanwhile, companies are training new models, with trillions of parameters, trained on questionable data sources, and putting them out before the public even understands the risks.
So yes, regulation is an option. But trying to regulate AI is like putting a rookie traffic cop on a Formula 1 track. They can try to slow things down, but no one’s gonna listen. In the absolute worst case scenario, they move their servers to another country with less scruples without missing more than a beat.
-
@Warma-Sheen said in AI PBs:
The Industrial Revolution happened over many decades. But it took just as many decades for laws and regulations to catch up to what has happening - for many of the same reasons it won’t work today. Too much money and influence on the side of the people with the new toys.
This is a misrepresentation of the historical facts. Laws and regulations that reformed the worst excesses of the Industrial Revolution happened because ordinary people demanded it in spite of overwhelming monetary opposition from the incredibly wealthy and influential capitalists of that time. John D. Rockefeller’s net worth is estimated at around $253 billion in 2013 dollars, Cornelius Vanderbilt at $203 billion, and Richard Mellon (of Carnegie-Mellon) at $103 billion. These and other “captains of industry” of that era commanded money and influence on the same scale as any modern tech billionaire.
The government did not decide to regulate because there was no money telling them not to. “Strikes organized by labor unions became routine events by the 1880s. There were 37,000 strikes between 1881 and 1905.” We owe the eight-hour workday to continual agitation by hundreds of thousands of workers over more than fifty years. The President of the United States sent federal troops to end strikes (these are all different Presidents), and Rockefeller was widely blamed for organizing the murder of 21 people, striking miners and their families. Between 1850 and 1937 almost 900 people were killed by the authorities in labor disputes. Regulations did not “catch up to what was happening,” they were dragged kicking and screaming by we the people.
This kind of mealy-mouthed defeatism serves no one but the ruling class. Ordinary people have stood up for themselves and demanded better treatment in the past, and we can do it again.
-
The fact that our government is currently useless is not really an argument against regulation. Like I too am depressed by living through a goddamn authoritarian takeover as a goddamn career public servant but it has nothing to do with my opinions on what should be. It’s almost a red herring - what the law currently is has no real bearing on what is right and that is something that has been true practically forever.
Eventually the law will catch up to reality, or eventually the heat-death of the universe will render this all moot, but that doesn’t change whether those materials used to train AI were stolen in fact (they were) or whether AI is having a negative impact on art that outweighs its positive applications as a tool (it is).
@Trashcan ty for those receipts.
let it be a challenge to you
-
@Trashcan It isn’t a misrepresentation. Its just a retelling. You’re saying the same thing I did, but with a different, antagonistic slant for no other reason than to be mean and nasty, presumably because I don’t agree with you.
You quoted the first part of what I wrote, then left out the rest.
@Warma-Sheen said in AI PBs:
In the meantime people were brutally and inhumanly pushed past their limits for generations with what we’d today consider abuses in order to feed the industrial machine. But here, today, AI isn’t playing out over decades, it’s evolving over months, across dozens of industries.
But if helps you to give a more detailed picture restating and supporting my points, then insult me anyway, you do you. That is definitely a pretty common response these days.
If your argument is that we should start the march against AI now so that in 70 years we might have a chance of winning the fight like we did during the Industrial Revolution, feel free. As I agreed, regulation is an option.
But the rest of my point remains that this isn’t a fight that will last 70 years and the same tactics won’t work because the circumstances are drastically different and the thing we’re trying to stop is moving much, much faster.
