/furi/ - Yaff

AI "art" looks so bland and generic it functions as the opposite of art or porn.

can we get this trash quarantined into it's own containment thread?

>>3707571
I consider the excessive posting of it to be spam, like over in the otter thread at the moment. Gonz disagrees though.

>>3707579

AI is eventually going to be harder to distinguish from art made by a person. Any decision made will be subjective at that point. I don't want to create unrealistic expectations. Plus, lulz has always been lax on type of content is allowed to be posted.

That being said, if you think you can enforce it, be my guest. I'm sitting this one out.

>>3707592
It's blatantly obvious all the images in the OP are AI abortions. Even with a cursory glance it's obvious that despite mimicking the same aesthetic as though they were made by the same artist depicting the same character the variation in limbs and fingers while generically posed makes it obvious they've been shat out by a prompt.

>>3707610

The whole point of my entire post went over your head. AI generated image rendering is a rapidly booming tech, and will be better quality in the future, daresay indistinguishable.

If Dan really wants to do it, let him. I really don't care.

>>3707616

Ai art is cool. The hater are "On the wrong side of history"

>>3707610
Don't moths, like all insects, normally have SIX LIMBS though?
Isn't that normal for them??

>>3707571
Have an AI moth, asshole.

>>3707592
the problem with ai porn right now, is that its too perfect. It sounds odd, but I can immediately tell if something was ai generated by the butthole.

Buttholes that perfect should not exist, and all ai that i've seen always makes them exactly the same.

>>3707568
Some AI Bees?

>>3707774
moth bee orgy

>>3707766
>too perfect
Apples to oranges, but they said the exact same thing about drum machines. This is why a "Swing" parameter was created, which adds the random imperfections in the groove. If you want to make music for a bunch of people high as balls at a party to dance to, keeping it tight to the clock is important.

The only reason I bring this up is because AI image gen could definitely use something like that "swing". But there will be people that embrace the "perfection".

Personally, I'm with that one guy that said its nice to look at even if you don't save it. I have yet to find AI porn I can fap to. 99% of the time, the lighting is just so unnatural to me.

Also this is cute, post more like this please. >>3707737

>>3707766
> Buttholes that perfect should not exist
No, every butthole should be unique, just like real ones. In fact, the FBI should create a database of buttholes in porn images for possible later identification.

The FBI can thank me for that.

>>3707837
> This is why a "Swing" parameter was created, which adds the random imperfections in the groove.

And it still sounds artificial, because human "swing" isn't randomly distributed. The "error" has a repeating self-similarity over time, which creates a secondary rhythm or a kind of melody over the main beat.

>>3707616
AI art has the same problem as ChatGPT has - it doesn't have any fundamental understanding of what it is drawing, it's merely "predicting" a likely outcome given some starting point. For both algorithms, the starting point is a "prompt" that defines what sort of special solutions the algorithm should use, and then it just throws random dice and mixes and matches the solution patterns using the random numbers.

Because of this, ChatGPT has a fundamental by-design tendency to hallucinate and ramble nonsense or simply make stuff up, and Stable Diffusion will always make stupid errors in anatomy or produce nonsensical distortions and errors that have to be corrected and fixed by "prompt engineering" to guide the algorithm to give you the answer or result that you want. Even then, you're merely improving your odds of getting the answer you want, not actually eliminating the chance of error. These faults cannot be fixed or improved upon because they're caused by the very principles by which these algorithms operate (Markov chains). It is not given that anyone will "eventually" come up with anything better either: insisting that AI will inevitably get better is just begging the question.

As you try to get the algorithm towards more and more detailed and fine output, you will reach a point where it's just faster and simpler to fix the mistakes yourself rather than attempting to trick the AI to behave. It's a neat trick, but very limited in the end, because the only way to "improve" is to set up templates and automated tests for the outcome and simply run the random number generator longer, discarding all the attempts that went wrong.

Creating those tests then becomes a chore in itself, because you have to figure out a way to define things like, when are the fingers of a character appropriate to the intent of the picture… YOU have to start giving the algorithm exact details to match and spending enormous amounts of processing power to meet all your requirements at the same time by random chance.

Coincidentally, the simplest way to define what you want out of the algorithm is to draw it yourself. Which is exactly what the Japanese are doing:

https://techxplore.com/news/2023-08-anifacedrawing-generative-ai-powered-high-quality-anime.html

Draw a crude sketch, the AI completes it into a generic Manga character.

Though of course, if you want any control over the fine details, you have to learn how to draw finer details. But, even if you do, that doesn't actually help you.

Imagine you're throwing ten dice. If you're happy with any outcome, your probability of getting what you want is 100% every time - but - if you demand that the third dice must come up 6, you'll only get the desired result every sixth throw of the dice. If you demand particular numbers to come up for for any five dice, you need 7776 tries on average to get it. If you demand all the dice to come up in a particular way, your need over 60 million attempts to have a good chance of getting it.

Those ten dice represent ten particular details that you want out of your picture. You need exponentially more processing time, more discarded attempts, with each demand on detail you add. The system only works if you don't mind that the algorithm outputs "whatever" - without you ever attempting to control it beyond a very superficial level.