The solution I’m talking about should already be the standard by most devs (especially small studios), even before LLM was a thing. See, small teams can’t afford QA, at least not to the same extent as big studis, so they need to add checks to stuff in a way that catches large problems, and a placeholder making it into the final game is a big problem. Even before generated images were a thing devs would just use any random image they had that more or less worked, and those images could have copyright or be problematic in any other way, so ensuring none of that made it into the final release has always been important.

Dude, naming the textures placeholder_<name> doesn’t take any more time and ensures you won’t ship a game with a placeholder. This is, or at least should be, common practice even without using LLMs, and only takes a couple of seconds, not enough to cause any inconvenience.

That’s not what a concept artist does, concept artists (if they had one) did the work before, game artists are still doing the work while the generated placeholders are in place, no person’s job was compromised by using generated placeholders. That being said, if any placeholder made it into the final game then fuck them.

I agree with almost everything here, I think using LLMs to generate placeholders is fair game and allows studios to nail down the feeling of the game sooner. That being said there’s one thing I disagree:

However, it is obvious to see that occasionally you’ll forget to replace items with this technique

There are ways to ensure you don’t forget, things like naming your placeholders placeholder_<name> or whatever so you ensure there are no more placeholders when you make the final build. That is the best way to approach this because even extremely obvious placeholders might be missed otherwise, since even if you have a full QA team they won’t be playing every little scene from the game daily looking for that, and a few blank/pink/checkered textures on small or weird areas might be missed.

I think it’s okay for studios to use generative AI for placeholders, but if one of them makes it to the release you screwed up big time. And like I said there are ways to ensure you don’t, it’s trivial to make a plugin for any of the major engines (and should be even easier if you’re building the engine yourself) where it would alert you of placeholders in use at compile time.

Then there’s nothing I can tell you. You haven’t even finished the game tutorial. I just found a video on YouTube of someone playing, the tutorial is around 20 minutes, after that you get into the game proper, once you get to the ship that’s when the game starts.

It does, it’s not a very large spoiler, so if you want to I can let you know. But without spoiling anything, I can tell you it’s very close to what you’re describing.

Obvious spoilers for people who’re still playing, but: Push through what exactly? Have you gotten to the main game loop? Do you know what’s the mechanic I’m hinting at on my original answer? Does 22minutes mean anything to you?

I’m going to assume yes (if not, push through until this makes sense). Now you have a whole solar system to explore. One thing that I didn’t noticed at the beginning and made me frustrated, there’s a computer on your ship, you have log entries there, sometimes reading those logs might give you a push onto what to explore next. That being said, this game is all about exploration, every planet has interesting things to explore, I’m not going to spoil too much, but it’s all about exploring and figuring out stuff, if you’re getting bored you don’t know what to do, if you’re getting frustrated you might be trying to solve a problem without all of the puzzle pieces, go out, explotó other places and you might find something that helps you.

There’s a game where the best way to experience it is to go blind and discover everything, I think you might enjoy it, it’s called Outer Wilds, and it’s one of the best games I’ve ever played.

For years my grub entry read rWindows, it doesn’t make sense in English, but “Ruim” is portuguese for bad, so it sounds similar to badows.

No, you are changing the rules: it is not my credit card number, it is a mathematical representation. It isnt my art, it is a conglomeration of all art.

Again, the credit card number was to show you that both the math is math and the information is free are terrible arguments, which you clearly understood despite trying to avoid accepting it.

I am 100 percent behind the idea that if you make a drawing, it is yours. Not mine. But if I sketch a similar one based on yours because I saw it, this new result is mine.

Let me introduce you to a new word plagiarism, copying someone else’s art is not acceptable.

The information is free, the original is also intact.

Again, give me your credit card number, if information is free you should have no issue with this. But you do, because you clearly recognize that information can have value.

Also, a AI made art is not going to stop someone from buying from some one else.

Except it does, go chat with any artist out there and you will be told how their commission work dropped, how they were fired, and even if they themselves were not affected they certainly know someone who has.

I buy from artists I know that have something so say. A different artist AI or not has no influence on my purchase from the ones I like.

You are you, but someone writing a book and in need of a cover, someone wanting concept art for a game, someone with a YouTube channel in need of a logo/avatar, or many other examples don’t care about what the artist have to say, they just need a piece of art to be a part of the thing they’re trying to sell. Artists can’t live by selling art to the random person who might like what they have to say, the vast majority of artists make a living doing commission work or hired at a company to produce art for them. Now the art that these people have produced has been misappropriated and used in ways which they gave no consent for, and that way is making them lose on jobs and affecting their livelihood.

You’re either a troll or just plain stupid, regardless it’s clear to me that it’s pointless keeping this up. We’ve gone in circles around everything, you claim information should be free but refuse to give away your information, that math is math but refuse to tell me numbers that have meaning to you, that digital artists are not artists but refuse to elaborate. You clearly don’t know any artist in real life, but I hope whatever it is you do for a living is the next thing LLMs starts to imitate, maybe then you will understand that the society we live in doesn’t care about the quality of the thing, but rather the cost and speed.

The credit card was to make you understand that even though no one owns math that doesn’t mean that all information should be free. You’re refusing to tell me your credit card numbers or even what you do for a living, this proving that you’re either a hypocrite or don’t truly believe that ALL information should be free.

Someone paying for art from someone who’s not an artist is definitely a loss for the artist who would have gotten hired instead. The fact that you’re even refusing to acknowledge this simple fact proves just how up your own ass you are.

Then don’t commission digital art, completely your prerogative, but the artist that does digital art is an artist, the person prompting an LLM is not. If you use digital art to make a profit you should have the rights for it, and a person who prompted an LLM can’t because they don’t own the training data, nor any derivation from it.

For the music, go and record chunks from several different music to see if you won’t get the same result, which is essentially what an LLM is doing with other people’s art.

Was it? Is it missing?

If someone grabbed the numbers of your credit card and used it to buy stuff, would you report it stolen? Or you would think that since you still have your card nothing got stolen? Same thing here.

And when did we mention anything about person B? Where is this person? How does this affect them in any way?

Pay attention, person B is the artist who’s not getting hired/commissioned and/or whose data was used to generate the image. Which is why I said I have no issue with personal use, the artist wouldn’t have gotten hired to draw an artifact that I will show for 5 seconds to my RPG players, no harm no foul. However if I was running the game on YouTube, or otherwise earning money from it then I should pay for it or not use it.

When I learned to play guitar did I steal the chords? I certainly learned to play other peoples songs, did I steal those too? I am now influenced by those songs, did I suddenly take away someones livelyhood?

If you recorded and sold those songs you would have quickly found out that there are copyrighted. This is the same case, we’re talking about someone profiting from it, not using for personal use.

In any case, so if I understand your argument: if the data was trained on publicly available data, you wouldn’t care.

If it was trained using data that the creators gave explicit permission for it to be used in that way then no, I wouldn’t have any issue with it. But publicly available data to view does not equate publicly available data to train a model, same as it wouldn’t allow you to print it and sell copies. Displaying something publicly doesn’t give you ownership of it.

Again, what exactly is the immoral act?

Like I said many times, the training data is stolen, if someone stole your personal data and impersonates you they’re committing an immoral act, this is exactly the same thing. Person A does the immoral act of using stolen data from person B to generate images directly harming person’s B livelihood, the fact that a person C is willing to pay for it is completely pointless.

You really are bad at making arguments.

No, you’re really bad at understanding them. You yourself made the argument that math is math and can’t be owned nor stolen, by that standard numbers are numbers and can’t be owned or stolen either. Stop to think things through before making blanket statements.

Regardless of the immoral act, your argument is wrong. A third party willing to pay has no bearing on the morality of an act, doesn’t matter how much you try to escape this.

You didn’t replied what you do for a living, I’m sure you didn’t because you know that there’s a very high chance I can show you you don’t truly believe that all knowledge must be free. Let me ask you other question then, what’s your credit card numbers, expiration date and code, it’s just numbers, by your own logic you shouldn’t have any claim to own them, therefore you should be okay to share them. The fact that you won’t is proof you understand that even if numbers can’t be owned, the information numbers convey is a different story.

And no, the companies are not claiming to own math, but to own the algorithm, the math on which those are based is (in general) public knowledge, and even in the cases where it’s not, like you said, math is math, others might have discovered it individually. Multiple of those companies might be using the same math independently without realizing it.

I don’t consider jaywalking immoral, so no, not the same.

Regardless of the seriousness of the immoral act, my point is the same, person A’s immoral act that affects person B doesn’t become OK because person C is willing to pay for it. Which is your argument, I’m pointing out how ridiculous an argument it is by using something you should easily consider immoral, and not in any way suggesting that generating images for profit should be penalized in the same manner or that is equally immoral, just that your logic does not apply to immoral acts.

I strongly suspect you do believe that in the world we live in ideas can be owned, let me ask you, what do you do for a living? Because if ideas can’t be owned, intellectual work shouldn’t be remunerated, as you can simply grab whatever is produced without paying the person and it wouldn’t be theft.

Yes, math is math, no one is claiming to own the math behind LLMs, but that math is applied to training data that does have an owner. You might as well claim you didn’t kill the person you shot, physics and biology did. The immoral act is the stealing of the training data, and any byproduct of that is fruit of the poisoned tree.

It’s the same argument you’re making, that morality doesn’t matter as long as there are paying customers, if morality doesn’t matter because there are paying customers for AI stolen art, why does it matter for child pornography? Either morality matters or it doesn’t.

It’s still free in Ireland, at least for home use.

I mean, calories in/out is real, you can’t get fat if you’re eating less than what you’re spending. On the other hand you definitely can thin up eating more calories than you spend by for example going into ketosis where calories don’t matter all that much.

All of that being said, calories in/out is not the whole picture, like you mentioned there are plenty of other stuff that might make it so that two people eating the same and exercising the same amount get drastically opposite results. At the end of the day our bodies have a calorie budget they’re trying to stick to, eating less (or actually eating better) is the solution, exercising helps but not in increasing your calorie budget, only in directing your budget to be more healthy.

Don’t mix up optimistic nihilism with moral wrongness, you can be “nothing matters let’s have a good time” while at the same time wanting everyone else to also have a good time.

MMA also has rules and it’s also a sport, but it has the less amount of rules so it’s the sport that more closely resembles real fights. But still self-defense (think Krav-Maga or similar) has lots of stuff that is not allowed in MMA. That being said, MMA is the safest way to train against a fully uncooperative attacker, so it’s the best way to train self-defense on the long run, but some classes and training on how to properly kick balls, bite, and gauge eyes are a great addition for real life-or-death situations.