Yeah you got me. Your a real sherlock Holmes. Everything I do is gen ai. I've been so against it for the past 20 comments and forgot to tell "you" and "myself" that I owe all my artwork to it. I've been using gen ai since 2010.

As I said...a real sherlock holmes.

Since it’s my thread, I apologize for the recent argument that just Happened. I really hope this thread can continue to talk about Ai and everyone’s thoughts on it. If this thread doesn’t get locked.

@PrinceMiniMod, are you ok to manage the argument on the past few pages

Sure, no problem

Thank you! I really didn’t want this thread getting locked.

Oh and thank you too @TurkeyOnAStick for not locking the thread to begin with.

OK, keeping this thread relevant again. I remembered this user used a mix of Blender and Stable Diffusion.

Though at some point in June of this year, he left the site and proclaimed his animations are "not allowed on this shit site". I don't remember what was removed, though 3 movies and 7 art pieces still exist to this day.

Before that, his bio stated that he could draw, paint, and animate by hand, but stopped to use this method instead

For reference, his hand-drawn art from ~2019-2020 looks like this

His actual process (for 3D + AI movies) can be seen in this other animation (which may or may not have been released and removed from NG). Timestamp: 4:22

https://www.youtube.com/watch?v=EgMWCfLZKas

No harassment please! I mean it. DF got absolutely ganged on, so none of that here thx.

Tricky. I know Spiderverse used AI for the lonework affect to achieve the comic book look and overlayed the lines onto the character models. Then the animators would correct the lines. Here, they were trying to achieve a painterly look. While AI purely as a tool rather then to generate computer made art is somewhat controversial, people are less annoyed by that because theres still a human effort in it.

But that animation you brought up (I'm going off the animation itself, not the process as I'm currently in a hurry) looks very slapdash. Like they modeled everyone and told the AI to put paint strokes onto it and called it a day. It ends up looking like those Instagram artwork where everything looks well defined except for the eyes which looks blurry. If they are going back in and correcting the AI, they aren't doing it enough.

I would agree with that on the first video because it is a little bit janky in terms of frame rate but the second one feels a lot smoother. But, at the very least stable diffusion wasn't the core tool to create the animations as much as the 3D software was for the models and movements and it was cleaned up so it does get a pass for having human effort being the driving tool as opposed to prompts.

See this is where I wouldn't have as much of an issue with Ai because if its just being used as a filter then there isn't really that much harm compared to how some other people use Ai to create a whole image or animation, the latter is where I would be critical of people using Ai.

TBH I'n surprised that one example of Ai being used effectively that wasn't brought up may actually be from south park S27E1 where the very end of the episode has a satirical Pro-Trump PSA. I'd cite it here but given it's NSFW I think I'm gonna reccomend people look it up for themselves, but the intent of it is actually more or less ironic in context while taking a jab at Trump.

But that's the thing, these examples have intent behind them that are done in a way to achieve a desired aesthetic or context and that's the thing that needs to be determined if people are going to be using Ai. If there's no intent and it becomes tactless, that when its time to be critical of it.

Also relevant: A later video on Bixdood's method (long after he left NG).

https://www.youtube.com/watch?v=xwcrweHvTrc

As explained in the second half of this video. if you generate each frame with AI, they will ALL look inconsistent in some way (I believe he knows that), and that's something that needs to be addressed. Usually by painting over frames, or in the case of a close-up shot, interpolating a single frame over the model. Basically a ton of trial and error and fixing and all that.

There was no ai use, the creators are strongly against it.

https://www.thewrap.com/spider-man-beyond-the-spider-verse-no-ai-chris-miller/

Newgrounds is not what it use to be and I feel like a fool having so much misguided faith in it.

The fact that we have all these users trying to converse with each other about "when" and how "ai can be accepted by Tom Fulp, is kind of pitiful.

Coming up with excuses for ai -use in any part of production shows a lack of technical skill and professionalism.

Deviantart, use to have the same debates and threads before it was swallowed up by gen ai prompters and saturated with generated fetish crap.

The fact that Tom Fulp would even allow AI use at all is a troubling sign.

As many could see through my pointless back and forth with DioShiba, some might call me irrational in my Ant-AI Stance.

But if you truly knew the processes of data interaction in gen ai processes, you would feel the same way.

If you talked with other artists who were ripped off and data scraped by some unskilled techie, you might not be so eager to debate about the topic and immediately take a non-negotiable stance like I have.

I'll still say a few words here and there on this thread, but I'm not going to get too deep. I'm here to make animations, not debate with people of different experience levels on professional morality.

This really sunk in, when DioShiba started spamming 0 star comments on my photoshop paintings, claiming everything was gen ai. I was forced to un-publish some of my work for now, block him, and report him for everything he said.

I was not trying to be a real troll at first, but I see that my attempt at educating him was pointless and he was the wrong person to deal with. It got to the point where I started feeding into his lack of experience and really trolling him...which I regret.

I formally apologize for the nonsense I partly instigated on this thread.

This is not the only website where these kind of conversations take place. on some of these sites, The Anti-AI stance is more popular thanks to a lack of strong policy by the website holders. I was hoping that ant-ai sentiment was strong here, but I was mistaken.

Some of you see gen ai as a transformative tool and you have to adapt to the times. This is very false and damaging propaganda and it prevents you from being a better artist in the long run. There is no skill in pixel reinterpretation through prompting and word usage...especially at the expense of others.

This is the last time I talk so much on this thread, because it all feels pointless here.

I'm going to be watching activity on newgrounds very closely in the next few months, If i see an increase in ai users on this "once great site", I'm deleting my account and moving to Cara, Artfol, and inkblot where the anti-ai stance is absolutely "no tolerance".

I have one last thing to say to Tom Fulp...

Don't give in and don't be a sell-out. Don't betray the hard work and sacrifice of real creatives.

I apologize on my end for the fight, at the very least, that much I can take some responsibility for.

However, I will also point out that I never once claimed that Artificial Intellegence was without problems. I'm fully aware of how it works and where the data scrapping is the issue in spite of what was percieved before things spiraled.

That said I'd much rather continue a productive discussion. If people feel that my stance toward neutrality is not the stance they want to hear from me then I really have no control over that nor do I intend to change people's minds, that's counter-intuitive to any discussion and I'm not going to be the one to tell you what to make.

To get this thread back in a good direction, what do you guys consider appropriate usage of AI on this site if at all?

Me personally, I'm only just a little understanding of the stuff by ActiveObjectX since it's actively shitting on AI. I also feel it may be the main reason why Tom allowed some AI to be used here for animations such as that. But usage of AI to "improve" one's artistic capabilities? I just don't see it personally. AI belongs in automation to me, not in art. Matthew Gafford has not once used generative AI in his works, and his "A Fox in Space" series is some of the most cinematic stuff I've ever seen. Also who is seen as the artist in the usage of AI? Cause if you're using AI to make "art", that feels like the equivalent of ordering from a restaurant and calling yourself the chef, and the food you ordered contains no nutrients to speak of. Typing a prompt doesn't make you the "artist", it makes you the commissioner of a machine that stole from actual artists. And even if the AI happens to be ethically-sourced, there is still the looming problem of it being horrible for the planet.

My summary: I cannot tolerate Generative AI until it becomes both ethically-sourced and eco-friendly. Once that is the case, I will simply consider it usually "lazy & uninspired", and just ignore it.

-

In addition, I will say that my views on AI are not entirely Gung-Ho, but mainly in the TTS field as sometimes stuff like this works pretty well. Though I would suggest using voice actors at any point possible, I do understand exceptions like this. I would say stuff like this is where the gray area is for me.

Thinking about media like this makes me change my all-or-nothing stance on AI. But I do believe we are taking it way too far, especially after the release of Sora 2. Basically a line has to be drawn, and I feel it is up to the community on what we will accept here just as much as what Tom & the team will allow here. If you wanna see less low-effort AI content on here, start being more involved in the judgment sections of stuff uploaded here, and as @BrandyBuizel said, "Don't grade on a curve".

Just to drop my stance in here as well:

The genAI sector is bad and needs to go down. It would be great if it could be utilized for ethical automation purposes in the future, but as of right now the development of the technology is in the wrong hands entirely. You can already see the effects genAI and ChatGPT have on society, letting this monster loose on the general population was terribly irresponsible.

Still, it's not a dealbreaker for me if Tom wants to allow genAI on Newgrounds in some capacity. I will simply use the tools provided to vote and review accordingly, in a civilized manner - even if I am faced with an AI-generated homunculus of Edd Goulds work.

I encourage everyone to do the same. It's up to the community to shape the future of Newgrounds as well, we can't just hope that NG staff will pass whatever law we deem necessary.

I agree with this sentiment.

The thing is with how Ai creates everything based on what's already out in the internet I feel that it's cheating at that point. Typing in a prompt just so a computer can analyze the data that's already out in the wild doesn't build on any real skills a technical artist has and in that regard it generates results that lack the subtle touches a human makes.

That is where the problems begin because at that point if nothing is done beyond the prompt typing and the asset generation, its not art.

At this point given how there are many things there are that are bad for the enviornment, even with our computers and phones and most thungs we use in our daily usage. I do wish that there were enviornmentally viable ways to handle these evolving technologies but that goes into another topic all together.

What I have a real issue with more so than the enviornmental concerns is when grifters looking for an easy buck try to take advantage of Ai and sell it off as art. This is where I would be more anal about Ai because if people are not disclosing that Ai was used in any capacity in the creation of the image then that's deceptive and it creates issues for the real artists who are trying to sell their work.

This is where my stance on the side that I'm not for Ai given the issues I described above come into play because in it's current state I can't find that how it does create specific images is concerning.

As for where the side that I'm not against it comes in... I think as discussed prior in this thread that sometimes there are artists who take inspiration from Ai generated assets or parody generated assets from people who use Ai and make something from scratch or painted over to an insane degree. At the very least the only thing being done with a generated asset is using it as a reference point.

Glolo Trees and Maiden of Flames are two examples of works made by artists who did use Ai as a reference point or inspiration.

At that point the concern is who uses Ai, how its being used, and if everything is properly communicated.

I would say that the big issue with using Ai for voice actors is that anyone could be used for a voice. What I would be against entirely is if the voices of dead people are being used because whether they're famous or not that can be used for illegal or morally wrong reasons. At that point I'd rather look for real voice actors.

And that's exactly it.

We should be more involved in the voting process and encouraging people to move away from Ai and having them use the hard skills a writer, artist, VA or whatever other creative should have.

But I also think we should remind people that we shouldn't be too paranoid about Ai either or worry about it too much on grounds that there's only so much the tools can do, everything with technology will always have it's limitations.

And assuming that Ai does get to the point where it's Terminator robots or Sigma's reploid army taking over, we can always purchase high powered magnets to corrupt the computers.

As long as the community remains vigilant, but also civil, we'll be alright. If one is not entirely sure if something is AI, there are tools funnily-enough that can scan if something is AI-generated or not. But be sure to use your best judgment, and bring it up to the mods whenever low-effort AI content slips through the judgment phase.

I think it's also quite wise to check the uploader's background if possible. Links to other social platforms, previous works etc. Many times, that way it's possible to find obvious "proof" whether the uploader is an AI-bro or anti-AI.

Agreed 100% on that note

Yep, and if they have no other socials or credentials, it also makes it easier to find out if they are legit or not.

i found another ai exception (in this case by teemovsal, shouldn't watch his stuff, but this is IMPORTANT)

evidence, note the inconsistencies in this frame of the movie:

and worse, people voted high on it, still hope it doesn't get a daily

see the following frames of this 1071 frame abomination

frames 259-544

frames 672-930 (i'm not sure about the last few seconds, but i see inconsistencies with the way the walk was animated

literally something worth reporting

WE GOT A ANOTHER AI MOVIE even worse it’s from the guy who calls his blockers retard

Teemov' is a shovelware farm at this point targeting at horny normies. I personally wouldn't even bother watching more of his crap, not to mention calling him out. It's been fruitless so far, and thus, I think it won't be any different this time. Teemov's content is consistently decreasing in craftsmanship and effort.

I've an idea! 💡 If loves blocking people so much, let's do the same in return. Let's leave him playing all alone in his slop landfill. >:)

It's easy to draw the line at "it otherwise agrees with my anti-AI views", but personally, I've used chatbots like Grok, Gemini, and ChatGPT for various purposes, which could include turning a short summary into something more fleshed out (this could then be polished later)

|

Chatbots / scripts: I would count this as a mainly "valid" use if you take what it says with a grain of salt, and you keep yourself one step ahead of the AI (ie. have a good eye for whether the output sucks). Don't rely solely on ChatGPT for ideas.

For background art: Full-prompting should be sort-of-OK as long as the rest of the animation is human-made and holds up, but expect sideeyes since it's easily noticeable. Stable Diffusion is slightly more valid provided you've done multiple touchups vs a single prompt, but again, can be easily noticeable if you're not careful.

|

My recommendation (if you HAVE to use AI for backgrounds) would be to use your SD output as reference and redraw it in your own style, that way there's no perspective errors (except your own). Or if you're in a massive rush, simplify and paint over the most polished SD output you could get.

Character animations: Never use it as the final product. You could prompt programs like Sora can generate full videos, but that's definitely not allowed and often looks uncanny as fuck.

|

Don't rotoscope AI videos / animations. You will just end up picking up their errors in the process.

|

For rough animations, sketch out keyframes and essential in-betweens, and let an AI program generate in-between frames. From there, you can tweak everything by hand and cleanup+colour by hand.

|

Example: Joe Gran (though his example is a bit rough, a different setup could've led to better results)

https://www.youtube.com/watch?v=yUQBJbYc7h8

|

Prompting / Stable Diffusion can be used FOR REFERENCE (mainly for things like colours and lighting), but don't rely solely on it.

Audio: TTS should definitely be valid, since I've used it for 5 years without issue. More advanced AI voices based on real people may cause problems.

|

Don't use soundtracks from Suno/Udio, find something off the Audio Portal / other royalty-free sources, or make it by hand.

So basically, don't let the AI parts be directly visible. If you use it the way I've described, you should be totally OK. The idea is to not let the AI dominate your process, but for you to step all over it and maintain the majority of your work.

It's pretty much a third-party asset, much like a stock image, or a sprite from Mario or Madness, albeit one that does not generate the same level of recognizeability.

I think Teemov made one huge mistake. He did not mention the fact that he used GenAI assistance to produce that animation and what specific elements were AI-generated.

You’ve already posted a similar post in the EGB thread and it’s got snubbed.

Report it if you’ve got a problem with it - don’t start another brigading clown-fest.

You can use gen ai. but you'll only be robbing from yourself in the end.

Your not learning from the experience, improving your skills, or getting better by doing this. The algorithm is.

True Artists live for the tedious and the details. The passion is in the process, whether its drawing, animation, or music. These "human-made" details are what contribute in the end and lend its power to the final product.

Coming up with mental gymnastics "why" and "when" gen ai is acceptable is very troubling to hear.

Your hurting yourself greatly by using it any part of the process.

But to each his/her own.

I'm not gonna debate anymore. All I can do is plant a seed in the hopes of helping someone be a better artist.

That's pretty much how I see it. IMO what differentiates an artist in the core meaning from a person who simply produces content is that the former considers the very process of creation in itself just as important as the finished work, while the latter type only cares about the end result. B)

But I also agree with Aalacrimson Kerostein (lol 😛) to a certain extent. Using "AI assistance" is very similar to using a stock asset. And in both cases, the dominant cause for people to rely on it so heavily is laziness. It's that desire to get the most "good enough" result with minimum effort. I think we can agree that this very trait very often applies to scammers and con artists.

But Chris brought up a point we tend to overlook. Using pre-cooked shortcuts, whether it's the hip new GenAI method or just traditional third party assets wouldn't lead you to improving your artistic skills. The thing is that, GenAI is way way more unethical than slapping stock assets. It's the dirtier way of playing dirty. xd

By the way.. there's another Dearest Friends Episode 1, uploaded from another account in 2024:

I was contemplating wether I should even post this here - in fear of triggering a witch hunt - but I actually think it's pretty cute and decent. Way better than the episode that started this thread and I can see no hint of genAI usage. You can even follow the plot along in spite of the arabic dialogue. Interesting case study!

Why did this guy use Ai in the first place? This is much better!