Let's take for example, Tifa Lockhart.

In 2020, when generative AI was in its shitposting stage, you could generate a bunch of blobs that looked like a character if you squinted. Tan blobs for the skin, black blobs for hair.... and boom, it's Tifa Lockhart..... if you squinted really, really, really hard!

In Late 2022, AI generated art had a shockingly high boom. You could now generate ART that actually looked like human drawn art. However, the tech was still in its baby phase. At best, you'd only be able to get a very generic anime girl that merely resembled Tifa Lockhart (black long hair, white tanktop, big boobs, red gloves, etc.) but only through the most generic moe artstyle available.

Now... ....In this current date of 2025, you can not only generate a Tifa that looks virtually the same as she does in the Remake Trilogy, but you can generate her in many different styles. From Porn Artist A's drawing, to Porn artist G's drawing, to Tifa from the PS1 game's cutscenes, to a hypothetical looney tunes version. Oh, and let's not forget photorealistic images that look exactly like if Tifa were a real girl with an Onlyfans.

In 5 years, we went from paint blobs to basically ANYTHING that could convince newbies, normies, and boomers that it's either real art or photos of a real woman.

It is absolutely mindblowing. I think this is something we ALL need to stand by and marvel at.

A lot of people sleep on how much AI has advanced, usually due to being over exposed to outdated StableDiffusion models with the same prompts used over and over, or people relying on DallE 3, but when someone is up to date and actually curates a prompt, it's incredibly impressive. I had followed Machine Learning since ~2015 when a guy made a video on using ML to try and get a machine to learn how to play Mario, and I've been fascinated since, seeing the titanic leap up in 2020 and then the steady progress since then has been exciting, and basically every naysayer was grasping that "Well surely it can't keep improving" which is just magical thinking.

Obviously I still believe AI should be watermarked and contained to it's own sites, no matter how good, it's the sheer volume that makes it dangerous to sites and search engines.

It's amazing and people need to take it seriously, don't gotta consume it, but recognize how "good" it is or end up falling for it. Always practice the old internet etiquette, "If it's too good to be true, it probably isn't."

Just wish AI filter was an option on image searches on more websites than just rule34 of all things

Its crazy what it can do but somehow I can always still tell when art has been made by AI... it just has that look somehow (even when there arent too many fingers). And maybe it should stay that way, but if it doesnt then it better keep getting properly labeled.

Ngl tho, I do have about a dozen or so AI images saved (among my 1000s of other images, lol).

No, we ALL don't need to do such thing. As always, you are generalizing your own opinions and demand that people think the same as you.

I personally still fucking hate that we have a groundbreaking technology that could potentially push mankind decades if not centuries forward in one leap, but we are still mostly using it to make things worse to widen the economic gap, hasten environmental collapse, make war even more inhuman and industrial, and, in connection with thread, to hollow out out real communication and emotions, to make human involvement and the passion and flow state that comes with practicing real art shallow and obsolete.

We were promised a future where machines drive our cars and do menial jobs, solve resource distribution and immortality, so we have time for learning, thinking and making art. Why should we ALL be happy about a present that is going in the completely opposite direction? Great, a machine can create pictures, so now finally I can have go back to work, then die in a ditch twenty years from now when human knowledge and bodies are obsolete and the ones who own the resources decide that the water I drink now needs to go to a datacenter. What a relief!

Being happy about Tifa's tits generated in Art Nouveau or whatever style is short sighted and I don't understand how people adopt such a one-sided consumerist mindset that ignores all context. All that bullshit about "idea first", "creativity without craft", "democratizing creativity" seems absolute misery and completely alien to me.

Why am I even in this thread, then? Now, that is the good fucking question, so sorry for the schizo rant, I'll see myself out of this thread. Just remember that whatever extreme opinion you hold in the spur of the moment, either pro or contra, there are going to be people who think the opposite, so tone it down a notch. Except if you are only looking for a circlejerk of like minded people about how good you think AI pictures are, in which case ignore me and carry on.

I think most people, or really the general tone I get from humanity towards AI as a whole, is a mix of amazement/excitement and fear/apprehension/concern, and I think that dichotomy is appropriate, expected, and understandable. I said this before in another thread with the same convo about AI, but my biggest fear about it isnt the tech its self but who is making it and in charge of it... corporations are inherently monstrous by their nature, no matter what their CEOs say to the public, so you can bet no good will come from them gaining even more power, as we've been seeing happen the past few years already.

And it's true, probably right now the biggest concern about AI is the resources it eats up, and that's not a future concern it's a problem right now and probably one that most people don't think about or care about whenever they use AI for their convenience. They don't even use ocean water or recycled water, they use fresh water by the pool-loads. It's easy to type stuff out and press enter while not literally seeing all the water you just used evaporate while there's a drought going on somewhere and people are going thirsty.

Yeah, after getting the previous post out of my system and having a breath of fresh air, I do agree with you. That's why I say that the technology itself is great. We should do better with it, and as you say, have better people in charge or at least more regulation to police it.

I just fear that's not totally possible in a capitalist world. We've already failed to regulate a lot of things that should have been regulated a long time ago, in the name of the economy.

Sadsap making a positive thread? Am i dead?

Agreed, 100%.

I also wish to share my existential dread about AI, but for chat bots instead of art bots, to change the topic slightly. As a college history teacher, I feel under siege by AI from two directions.

For students, it is easier than ever to cheat by generating essays practically instantly, but it takes me so much longer to manually prove it is not original work beyond reasonable doubt to the dean. I find myself doing far more detective work to catch AI-based cheating than I do actual grading, and it's an infuriating waste of my time. The amount of student fraud with AI is also unbelievable. Ever since COVID, basically half of all classes are now online, and at least half of those online "students" are bot accounts that are just signing up to siphon financial aid money. They're usually pretty easy to spot and cull on the basis of inactivity (I can dismiss students for non-attendance), but the clever fraudsters sign up for a class multiple times under different names, and maintain their enrollment using AI to complete all of their assignments. They don't care that they get zeros and reported to the dean every week for AI-based academic dishonesty; they only care that they are still marked as virtually "attending," so I legally can't do anything to kick them from the class to stop them from collecting those financial aid checks.

From the side of college leadership, the response to AI has been completely impotent and uncoordinated, and all of us faculty basically just have to make up our own policies towards AI ourselves. I've had more than one student complain about being dinged for using AI when professor So-and-So encouraged using AI in another class. Worse still, I've attended faculty seminars about how us teachers can use AI to generate syllabi and questions, and even analyze students' essays! Well, at that point, what the fuck do they need us teachers for? At a certain point, it'll just become a circle-jerk of robot teachers grading essays written by robot students.

From what I've seen, I think the post-ChatGPT student generation is intellectually fucked. Maybe in a perfect world, AI could be used like a virtual coach who can lead students to learn in Socratic fashion via multiple stages of questions and answering. But I think we all know that's not how students are using it. They just have a bot instantly spit out something that looks plausible for the least amount of effort. There is no incentive nor consistent accountability for students to, you know, actually develop their own critical reading, research, and writing skills. I feel am fighting a losing battle to insist that students complete their own authentic work, and I fear for the future when a generation of people rely on robots to do their critical thinking.

Anyway, doomer rant over. Time to appreciate some hand-drawn tiddies!