Was watching this latest Arch video where he compares the 'April O'Neil' that was in the latest TMNT property compared to a black April O'Neil he put together with AI. TLDR; the AI despite a few faults still made her hot and look capable in each compared to the frumpy troll the film got.
We know some of this is ideological that they want to lower the standards of beauty, but with AI becoming easier and easier to use that boomers are starting to use it, why pay for over compensated, lazy, unskilled vurtue signalling artists currently striking than just use a machine and then edit it a bit with 1 artist or AI again if it makes a third arm?
I honestly think AI is going to not really hit the blue collar jobs and heavy industry since if something goes wrong you need a human that can adapt to either fix it or turn it off to limit further damage. The service industry will be case by case like if there's been too much of an issue getting staff or with the staff there if it's more beneficial to turn it into a giant vending machine. But the one that I think will be obliterated in the west will be the Media industry, when all they do is repost each others works like a giant human centipede, why pay them when you can just get an AI to write it for a few dollars and it takes minutes not hours and doesn't require giant office's just a computer.
Entertainment is worse as I fundamentally believe we can easily have a lot of films that have the same CGI as the Final Fantasy film for cheaper than it would take to use actors nowadays. Just make sure you get a seperate AI to make the script good. We'll still get your Top Gun Mavericks, your John Wick or even your Sound of Freedoms as they appeal to customers but by large, I can see the bulk of western media being replaced with CGI in the next 2-3 years than a decade.
I don’t see this affecting Asian, Indian or even Eastern European media as much since they are still appealing to a customer base than preaching. I do think at the end of the current strikes, we are going to see a massive reduction in physical studio spaces and 'celebrity' actors and a rise in machine made Entertainment.
If they don't learn to shut their fucking mouths and give the people what they actually want, then yes.
Even if Hollywood gave what the people want, I want AI to compete with them. AI has a centuries worth of entertainment (radio and TV) with the potential to creating truly amazing stuff. Just imagine an AI generated tribute of Frank Sentra, a recreation of Chris Farley making stand up comedy again, or a new project that takes hints of all of the 1980s and 90s science fiction genre to make something to compete against classic Star Wars.
The funniest thing about AI to me on the image side is how many Tumblr-type leftist artists immediately fixated on "AI art is theft", while Japanese artists were overjoyed that they don't have to draw backgrounds anymore. It's an incredible force multiplier for any artist with a tiny bit of technical acumen and no ideology hangups.
Imagine how much manga we'll see now that Manga-ka don't have to work 100 hour weeks due to AI.
I understand that this is only one point but several things happened with the Japanese side.
Of the artists I follow, there were quite a few that embraced the technology immediately when it dropped, either having made their work easier/better or just playing with it for fun. However, for those that were vocal, many, if not most, were angry or frustrated. Their complaints ranged from "sameness"/"unoriginality" to "not putting in the effort", or "theft" like the Tumblr-types. It definitely is doing a lot of damage to repetitive-type artists or ones who price their commissions too high. Pixiv and DLsite et al went through many problems regarding their sites getting flooded or the monetization of AI works.
They are sort of going through similar issues, even though studios can now get magnificent backgrounds for dirt cheap or no expense. For individual artists, acceptance is slow depending on the quality of their style or their ego.
And again, like the video, the best rise to the top. The town of Hinamizawa, from the VN Higurashi, literally exists. The artist went to a small podunk village, and just drew what he saw in straight-up artwork, and then converted it to anime style. You can visit where every character lived. And thus, the village looks incredibly realistic, because it WAS real. An AI would piece random things together, that look good on their own. And would work just fine. But the works made by truly good artists would still be noticed and appreciated.
You could use current AI image tools to teach your computer to draw in your style, feed it photos of a city or park or whatever, and it'll redraw the photos in your style. It's capable of so much more than the people just raw prompting on Reddit are doing with it.
For sure, but that then still presupposes an artist: A photography artist.
Everyone has professional grade cameras cira 2010s in their pockets at all times. It's easier than ever to take serviceable photos.
A new fulcrum threatens people standing on a flat board that's never been tested for weight bearing.
You forget that AI is controlled by Big Tech and they're curating the algorithms and data sets. ChatGPT was for an example in part based on Reddit data. And they're already complaining about ChatGPT becoming dumber after they started rolling out censorship controls to police 'problematic' output in order to not offend the offendetrons.
Western AI is getting lobotomised, I've heard that a lot of Asian stuff isn't because they don't care if it's racist if it points out one demographic has been committing crimes more than others, that's a plus as they always believed they were inferior to them.
You don't know what you're talking about. Open Source AI is moving so quickly and routs around censorship so habitually that many of the big corps have given up on even attempting to control it.
AI will not be controlled by any individual big tech company in the immediate future. If you lock down your product it stops moving, and in the current AI boom that means that it becomes hopelessly obsolete in a matter of months if not weeks.
Nevertheless, they did nerf ChatGPT and the free version has stood still for months. Or at least, it has the appearance thereof. You can never tell with this lack of transparency. The Bing AI is low-tier garbage on certain subjects.
And there are indeed open source AIs with jailbreak elements, which have been trained for 'dark net' type activities. A universal jailbreak flaw of LLMs has also been found that takes away the limits.
But, aren't these open source AIs clones based on what has been released by Big Tech?
It might seem like that if you don't follow the space. ChatGPT is already on it's last legs. All it has is compute power and ease of use for normies left. Models half the size are already closing in on it.
And here's the thing, it doesn't matter that the open source space is all based on Facebook's model. Tweaking it can take advantage of the existing work without being retrained from scratch, it's an iterative process. The genie is out of the bottle so to speak.
A common theory is that Llama was leaked intentionally because it was kind of shit, and Facebook couldn't figure out how to fix it. Throw it out to the public, maybe some nerds will fix it for free. As a bonus all the open source techniques that people come up with will be compatible with their infrastructure.
Just like that, Facebook is a player in the AI space again.
Clone isn't the right word, they are derivatives. But there is no secret sauce. It's just a matter of compute time. You could very easily train your own model if you had 100k to blow on GPUs
This.
Not to mention the cost issue; an AI that can generate a thousand images an hour is extremely cost effective. An AI that requires a million dollar robot body, that also needs constant maintenance, and can only do one job at a time doesn't bring the same savings.
Lawyers and doctors will be replaced by AI before plumbers are (if ever).
That I don't disagree with, the point i say "see this affecting Asian, Indian and Eastern European" as much is they aren't going to pull a tantrum threatening to wreak the place because they don't have as much power anymore, like they have in the west.
Like many others said, it'll probably be good for the anime industry as it might lift the burden of doing all their backgrounds. But more than that, because the west is resisting AI and trying to restrict it while Asia especially is trying to slowly integrate it, there will be not as sharp a shock as what will happen in the West and leave room to adapt and manoeuvre.
blue collar work is harder to replace because you also need the robot body to go with it.
Not that far off IMO but just another hurdle.
Here's Why the AI Takeover is Actually a Good Thing for Japanese Animators!
I don't see a lot of downsides tbh. They're infamously overworked and shows with standout animation like OPM s1 and AOT are labors of love that take up to three years to crank out. Then OPM s2 comes along after two years and it's a piece of shit, animation-wise, because they couldn't call in all the favors used to make s1. And it still took years!!
We're already at the point where cels aren't hand-drawn anymore so if AI is used to speed up the more repetitive posing, keyframes, whatever, I don't know if anyone will notice.
Maybe a bit of a hot take / shit opinion, but I wonder if famously stressed schedules of Japanese entertainment is kind of what makes them great.
You hear about stories with the composer of Chrono Trigger pushed to the limits of his sanity, code changes hours before compiling and shipping to the mask ROM house, people sleeping under their desks.
Sacrificing for your game means you're not going to piss away that sacrifice on something substandard. If they didn't believe in it they'd be working on something else, after all.
Meanwhile in the West the communist gaming media have gotten the ear of the right people and are pushing big on better conditions. And, sure, better conditions are good, but I think you see what the lack of pressure and devotion looks like. Pozzed messaging aside, modern games have more bugs than ever despite better and more mature engines and middlewear. Even Japanese companies buying in to the Western way of doing things like SquareEnix and The Pokemon Company show of what the 9-5 "it's just a job" crowd can pump out
I think it's even less complicated, more simple. Normie corporate drones have diluted the arts and media. Smaller, less accepted/glamorous industry meant greater proportion of fans/passionate employees. When games were more haram, you had more nerds and outcast geniuses working on them.
I think working conditions are kind of a red herring. Yes, schedules change when a company reaches a certain size, and it generally causes bloat and waste. But by far the biggest problem is when big money gets involved in a studio and MBAs, rather than true believers and visionaries, get to run things. Great games and revenue generation are related concepts, but sadly it's not a direct relationship.
Homeworld is one of the greatest RTS games ever and it was made by like seven guys who quit their jobs at Microsoft or w/e. I'm sure those guys had their sleepless nights but as far as I can tell from the making-of, they had a pretty good time creating the game. On the other hand, the development of Cyperpunk 2077 was supposedly under terrible working conditions with a lot of crunch and we know how that turned out. Why? In large part because CDPR was trying to keep up appearances for shareholders.
If the directors and the teams take pride in what they're working on, the result will generally be worth playing.
This is an excellent point tbh. Like truly dedicated and competent development teams, even when small, are almost certainly going to have a better time of creating a game than an over-bloated larger studio with massive organizational and structural problems.
There's less pressure and influence from above, creative decisions and general collaboration flow more easily, and there's sometimes less pressure to aim for the big bucks and more interest in just making something good and satisfying.
I only wish we hadn't seen so many indie devs diving straight into 2D retro styled games for so long. 3D, while a little more work in the art department, is still pretty doable for a small team without straight up asset flipping in an obvious way.
The biggest losers in all of this have to be online artists, especially freelance. I paid decent money a few years ago to get a semi-realistic anime portrait that is clearly inferior to what I can mock up Anything v4.5 in under 5 minutes. The AI is less repetitious and more vivid than the human illustrator's portfolio, actually.
High-level concept artists have been dining off mobile game art for years now - I don't know any of them personally, but a look at Artstation should tell you as much - and AI can effectively ape their style. Where does this leave them? They're going to design more original character props and features than the AI, but we're talking mobile games here. The studios literally don't care about originality because their customer base might as well be AI themselves.
Maybe a silver lining is that artists will stop acting like Marie Antoinette when you try to commission them. I get that some of their customers are fire-breathing retards but online artists have to be one of the most entitled and passive aggressive groups on the planet.
I think artists will mostly just start using AI to make drawings.
The biggest issue with AI art is that it's often messed up (looking at you, fingers) and you need someone to make it look good.
We'll always need people who understand composition and perspective to layout good pieces rather than mediocre.
That's true - but most freelance artists barely understand composition. A lot of chaff will fall away from the wheat.
I'm sure those who don't have a stick up their ass will utilize it to great effect.
And of course there'll be a few con-artist types who are just spitting stuff out constantly for the money, but they'll probably be outbid pretty fast by art-gen sites and free Stable Diffusion self-setups.
The content may be generated algorithmically but humans will still be needed to "guide" it and curate the results.
Until it gets much better that is.
Even then, I think similar to examples with the holodeck in Star Trek, people will usually want to cater the content to their own preferences. Which would still integrate a bit of human touch into the creative process.
I expect that eventually this'll become something of a common element when we reach a certain stage with more complex artificially semi-automated content generation, where consumers, individually, have some control over what kind of product they get to experience.
We already see this all the time with moddable games.
Here's a question in return. Will you consume AI product?
AI is the next big technological "leap forward." Problem is these advances have generally been bad for society. We didn't see the evils of social media coming, but the writing is on the wall for AI content. I'm not going to bat for the current generation of creatives, but what replaces them could be worse. The truth of the matter is that the Luddites were right, and it's time to consider if rejecting certain new things is a good idea.
I think there's going to be lots of effort to disguise AI content. The same way you used to screw in one bolt on your widget at the local distribution center and call it Made in America
Depends on who the "creative director" is and what kind of quality their results tend to yield.
I'm less inclined to bother with any AI products being spat out by large media corporations. They've already proven how shitty they are with actual people doing the work, and they're liable to only put more effort into propagandized products when they don't have workers balking at their idiotic suggestions.
There's some creative products I frankly don't know how confidently I'd enjoy if it was produced by AI. Namely, novels and possibly music. Good novels are too condensed with so many concepts and ideas with such intricate detail that are developed over time that AI will certainly have trouble truly duplicating. Yes, AI can emulate style, technique, tone, and themes very well, but there's a certain personal touch from a good author that is difficult to simulate.
IE, how the various experiences in an author's personal life have influenced and changed how they look at the world. Their understanding of the human condition, human interaction, trial and error, challenges, fears, hopes, and dreams. Etc etc.
we're going to see it split into AAA-tier big budget AI productions which are the same old tired bullshit, and smaller more passionate projects that either die in obscurity or attain cult followings due to how well-made they are.
hey, wait a minute...
Hey! That's unfair to AI!
Sure they get fingers wrong, accidentally add a third arm and the like but they KNOW what a woman is compared to the current drones..
Three breasts and five legs? Still closer to human than what the media's been trying to push.
Sure, it'll be shit, of that there's no doubt.
Still better than what's being produced now by the "mainstream."
One thing that I love about generative AI is that it's going to enable the democratization of content creation like we've never seen before. There are so many creative types out there right now with interesting ideas who could never get them produced for either lack of funding, 'wrongthink', or because the idea is just too niche.
We're going to see a tidal wave of new content incoming once the generative video creation pipeline matures. You can do short clip text to video with a lot of different tools out there now and they're improving rapidly. I think we're perhaps a year out from a workflow capable of producing reasonably good full length video content on consumer grade hardware by dedicated enthusiasts. Once that happens the flood gates are going to open and the deluge of new content content will be incredible.
I don't see it replacing mainstream media in the near term, but I do see it taking a good share of viewership going forward. Something similar to how podcasts and amateur video producers on youtube have already taken a chunk out of MSM revenue, but on steroids. I'm looking forward to that.
What I have found with AI is that you still have to know what you are doing. I did the work of an entire team in a week or two, but none of my coworkers could do the same because they lacked the skill set.
I see a quiet office connected to the house where white collar jobs get done. The giant office in the skyscraper is now one guy with the right skills.
For fast food, most of it is very automated anyway, the teenager is just cheaper. But, machinery that can divert and do multiple jobs will replace them. It will be a single guy in an office in the back who maintains everything. The area managers won't be needed because the restaurants keep in contact.
I see blue collar work like car mechanic getting a tech upgrade with XR teaching them how to fix things. I want a 3D printer for parts. More average joes will rent out a space to fix the car, and the idiot techs will either step up or not be around.
Some of it is slow going. Some of it is so fast my coworkers are actually scared of me.
it's going to be used for sure, but will it replace western media? that remains to be seen. AI had a moor's law kinda of advance last year, but the rate of advancement already seems to be curbing.
I think people are still trying to mentally catch up to where they can take things further. Not so much due to technical complexities (though there are hurdles there too) so much as due to needing to properly conceptualize just what can be done with it. As in, actively using it in new and previously unfathomable ways, rather than just theorizing.
I'll be shocked if Hollywood doesn't debut an entirely AI built series in the next couple months. Consider:
*Multiple big projects over the last couple years have clearly shown that a lot of what the writers are churning out is not meshing with what people actually want
*The strike is now over 100 days, and shows no signs of stopping. The execs don't care, and the writers and actors are doubling down even though they're admitting that financial pain has begun
*Streaming services were already looking for other sources of content before the strike even began
If even one studio exec has a pair of brain cells to rub together then that studio has started work on one or more AI built series and they'll drop it shortly to both keep the money flowing for them and to utterly break the strike
We can only hope so
Ultimately AI content is Content. Film and Literature won't be affected, but the goyslop supplies are about to hit critical mass.