Great video, he says the gaming industry has experienced a year of studio layoffs, consolidation, and closures which has begged the natural question of... why? In this episode, Danny dives into how the pricing models and development costs of games might be the culprit.
what AI can/cannot do according to FZD, if you want to skip just to the part in the title, go to 32:00 https://youtu.be/QTj1Y4JW-KI
That's quite a good summary. It does go to show some of the faults I find with AI bro "it learns/is just like a human". The critical thing missing is actual understanding. Thinking about it that way, it feels a bit less of a threat than it has done, but even if it can't do those, there's still a tonne of work it can and will replace, ugh.
It's nice to be able to show that actual design work, anything coherent and needing consistency between multiple things, can be saved for humans, but that's not a lot of jobs compared to the amount who'll be put out of the rest of it.
Great video, he says the gaming industry has experienced a year of studio layoffs, consolidation, and closures which has begged the natural question of... why? In this episode, Danny dives into how the pricing models and development costs of games might be the culprit.
I also feel its the heavy marketing investment which cuts into development costs. Development is usually kept behind closed doors citing competition concerns, and most games marketing is focused on release sales, which is usually when games make their money. Live service seems to be a good work around since it keeps players involved and the game relevant and it also prioritises retention and expansion. Though in many instances it still hasn't stopped layoffs.
Larger studios should be invested in other areas to supplement the cost and risk of game development and marketing. And they need to respect artists more, though artists also need to realise that studio's are second homes. We are providing a service for a fee, and its easy to get lost in the illusion that we are truly valued for our talent since in the end its a product being sold, so the real value proposition is in appealing to the customer who can be a diverse and at times toxic bunch.
Like the 400,000 people boycotting games on the SBI detected curator don't really care about developers being affected by studios losing revenue. They do however see it as developers are better off not working for these studios and starting studios that care about the customer. Its an extreme viewpoint that I don't feel takes into consideration everyones perspective, but they were likely the first in line to pay full price for a game they want to own.
There is consensus that most gamers want fun games regardless of socio-political messages embedded in the experience. I feel that modding should also be made standard in games so it gives an opportunity to innovate.
Like the 400,000 people boycotting games on the SBI detected curator
don't really care about developers being affected by studios losing
revenue. They do however see it as developers are better off not
working for these studios and starting studios that care about the
customer. Its an extreme viewpoint that I don't feel takes into
consideration everyones perspective, but they were likely the first in
line to pay full price for a game they want to own.
There is consensus that most gamers want fun games regardless of socio-political messages embedded in the experience.
I apologise for the bluntness, but can you please stop trying to derail the thread to grind your axe about SBI and conspiracy theories of left wing infiltrators?
The first half of your comment was fairly on-point. Second half, not so much.
That aside......
Surely the budget is coming mainly from just the sheer demand there is now? Like, it's expected that you push graphical fidelity to the maximum as much as you can, and that's pricey as hell, right? Not just the environments, but the complexity you can go up to with characters in terms of the way their clothes deform in movement, nuances of expressions....
And not just that, but that games be longer and deeper, especially the big releases. It seems to me, though it's a simplistic way to look at it, the price is ultimately coming from the fact that gamers demand more.
And the more games try to rise to meet those expectations, the more the expectations seem to rise. Surely at some point we have to start saying "no, your expectations are unreasonable"?
Though, having said that, I also see people online who seem to be happier to take lower fidelity games if it means they can be released faster, or be less expensive.
Thinking about it, perhaps it's one of the problems of being this medium. Showing off how good your gameplay truly is takes time, and nobody will pay attention that long. But if you give it style and flashy visuals to knock the audience's socks off, that impression might stick. When you only have enough time in people's attention span to make a superficial impression, you'll have to rely on your superficial elements, I'd think.
About the SBI curator, I mentioned it since even that group seems to be against the current AAA approach, of course there's that whole DEI conspiracy which is their primary motivation but they too want to play games that are fun and don't break the bank.
I liked the idea of a large AAA company promoting smaller more focused teams to create AA budget games. In many cases AAA company's include this in their portfolios as publisher, many have a flagship title that usually follows the recipe (like FC for EA, or Assassins Creed/FarCry for Ubisoft) and these titles sustain other projects.
Not sure how much risk a studio can take with experimental titles, but they do have considerable room and do allocate resources to these projects. Its just that because of NDA's no one outside usually hears of them. In this case, indies and AA's that take a more audience focused approach with more transparency have greater leverage, so maybe for AAA they should move more into the publishing space and licensed properties which does seem to be the general direction.
And each studio usually focuses on what they do best or have more experience in, like Motive (EA) has a great level design, narrative and programming teams, so much of their character model is outsourced while they focus on integration and gameplay, I'm not seeing much initiative on building internal teams of junior/mid level artists to take on the tasks they can hand over to interns/outsourcers.
I do wish that they would balance the work load more, every senior I've met feels that they end up being burdened with a lot of work they would prefer to delegate to someone in studio rather than an outsourcer and many willingly take on this workload since they do feel they could do it best, so you need good leadership to balance this.
The visual aspect seems more a mix between style and fidelity and needs to run alongside gameplay, though it can stand on its own too. Palworld is a good example of hitting an audience niche, the graphics animations are horrible, and there are allegations that they plaigarised assets and used AI, but the gameplay is fun which is what their mostly programmer focused team is good at.
And its also on brand, like if Naughty dog replicated Palworlds approach it would likely face a boycott. Their studio has to live up to that image which is also what the artists want.
With the layoff's we're seeing, I think that AAA studios are prefering to focus on core properties rather than developing new projects. There is no real pressure or incentive on them to innovate so long as they are guaranteed to make sales on release and through their live service windows.
Games can be made on a much learner budget and directly marketed to an audience so this is likely where the AA, indie space will stay going forward. Here are some games to check out for their development approach,
They all focus on a very specific niche and marketing based on their original elements. I find AAA marketing to be more brand focused, personally I can't believe how much star wars outlaw's cost or why anyone would pay that much day 1, but the branding means that many will. AAA studios are also promoting subscription models so that's one way to rent a game to play through if you don't see yourself playing it again.
How about a little diversion creating stuff NEW stuff instead of 'pustulising' errant hyperchondrial ticks that tend to scratch out text walls of diatribe that seem too inhabit your imagination of late.....
Promise, you'll feel a lot better maybe score another gig into the bargain
How about a little diversion creating stuff NEW stuff instead of 'pustulising' errant hyperchondrial ticks that tend to scratch out text walls of diatribe that seem too inhabit your imagination of late.....
Promise, you'll feel a lot better maybe score another gig into the bargain
I'm not sure if we'll see more innovation in the AAA space given how risk averse they've been off late.
I am intriguied by the recent controversies we have going around, lot of entertaining content which actually motivates writing. Also with so many out of work its much easier to network, reachout and improve enagement on linkedin which has proved very useful.
I have a few ideas outside of making artwork for job portfolio's though they'll likely be outside the game dev space since I've been experimenting more with 3D printing for board games.
For portfolio, atleast following my consultation with a senior artist, they said some variety would be good i.e a female character, otherwise what I have does provide more than enough evidence that I'm capable of doing the role I'm targeting (mid-senior character artist)
The hiring market is a challenge, but its good to have more free time to try new things.
you don't understand Francois, the art is the walls of text themselves. A super rockstar artist (ex Blizzard, ex EA, ex Sony, ex Bioware, ex Pornhub) told me so, then complimented me on my ability to hold my breath for 30 seconds, and also do not just 1, but 2 chin ups
Now that's getting a bit mean. As much as I really, really disagree with some of the stuff that NkhilR has posted, and have my own opinions, this whole thread is going to get locked if it keeps derailing, especially if it starts turning into just slagging people off.
I'd really like to keep this thread going for the sake of the on-topic discussion and updates on the state of AI fucking up the industry. I mean, the topic is also depressing, but it's good to know.
Or to put it another way, I suggest we keep our eyes on our true enemy.
Speaking of, I saw this article. It's not related to the games industry, but it is gross.
Off Topic : Do you present art when applying for Character Artist jobs or walls of text mentioning DEI or SBI , Nikihlr
Lol, why art of course, though I am wondering how much representation I should have in my portfolio when it comes to the race and age of the characters. I try to think of this aspect more as showcasing that I can create a good range of different characters.
DEI is supposed to impact my application positively since I'm south asian but I'm not sure if applies equally when it comes to representation among visible minorities. Like I find that candidates with physical and learning disabilities aren't as represented if they don't also have other minority attributes and even then may not be selected over other more able bodied minorities.
I used to be a teacher for game dev students with autism hoping to working in the game industry so saw this first hand. The most we were able to do is to have them interact with game studios through social initiatives for community building, like a studio visit or a talk by studio employee, but it didn't translate into job offers and there weren't any inititiatives targeting these groups specifically which would really help increase representation.
For example this initiative. https://www.ubisoft.com/en-us/company/careers/interns-graduates/develop-at-ubisoft Provides representation for women and non binary candidates and its unsettling to be asked by candidates with disabilities if they ought to identify as non binary to apply since there is no seperate initiative for them and they don't stand a chance of being hired in the regular stream.
Lol did AI paint her dress on her? This is one situation where I fully approve AI being used and I wish AI could have turned back time considering what she did.
Speaking of diversity and portfolio, I saw up this thread that you could produce a character turnaround and have look fairly "hand-made". However the exemples I saw were always in the riot/disney/blizzard soup. Have we seen AI produce actual interesting character designs? (Sorry if this has been discussed earlier, I only skimmed the topic)
NikhilR said:
Lol did AI paint her dress on her? This is one situation where I fully approve AI being used and I wish AI could have turned back time considering what she did.
Assuming it is, as it looks to be, AI, the technology is being used for historical revisionism. You can't just call yourself a documentary and then start showing photos of things that never happened. Especially using it to make this narrative of her being happy and smiling when, as far as I'm aware, she was very stressed out and unhappy. That was the motive, right?
Even if she had been happy, it's fucked up to call itself a documentary and then show fake images. That shouldn't be legal, AI or not. Especially when it's about something like that; it can completely change the impression the audience gets of someone if you depict them falsely.
A third! That's crazy.
I'm not surprised at the amount who changed engines, or considered changing, considering that Unity fiasco...
I really wonder that 1/5 that isn't worried about generative AI is thinking, though. I can't understand that at all.
Speaking of diversity and portfolio, I saw up this thread that you could produce a character turnaround and have look fairly "hand-made". However the exemples I saw were always in the riot/disney/blizzard soup. Have we seen AI produce actual interesting character designs? (Sorry if this has been discussed earlier, I only skimmed the topic)
Personally, I haven't seen any. They tend to be......Very overdesigned-looking. They have a lot of detail, but they tend to lack the sense an actual designer would have about where to put the detail; they don't think about having a strong silhouette, or guiding the viewer's eyes to any parts in particular, or telling anything about the characters through design elements because, of course, it's not able to think to imagine those and work out the best way to represent them.
Plus, they tend to have extremely generic faces. Asian characters in particular always seem to end up with that offputting Kpop idol look, with the obviously-had-loads-of-surgery-sameface, big doe eyes, small features, etc.
NikhilR said:
Lol did AI paint her dress on her? This is one situation where I fully approve AI being used and I wish AI could have turned back time considering what she did.
Assuming it is, as it looks to be, AI, the technology is being used for historical revisionism. You can't just call yourself a documentary and then start showing photos of things that never happened. Especially using it to make this narrative of her being happy and smiling when, as far as I'm aware, she was very stressed out and unhappy. That was the motive, right?
Even if she had been happy, it's fucked up to call itself a documentary and then show fake images. That shouldn't be legal, AI or not. Especially when it's about something like that; it can completely change the impression the audience gets of someone if you depict them falsely.
Oh by approval I meant that about AI being used to turn back time to better times as indicated by that image they modified before she went postal.
And certainly she wasn't always happy, she has some serious problems but the image does indicate a time when everyone was in a far better situation than the horror she caused. The Netflix documentary has been understood to be more suspense driven and sensationalist, so this approach was expected.
On a more whimsical note, I felt that someone like Jeniffer didn't deserve a more curated upscaling of her photos, AI is as good as it going to get and the result is as twisted as she happens to be, so it kinda fits her sociopathic personality.
Also I'm still looking into what they did with that image but it seems it was run through an upscaler, so its not fully AI generated but enhanced albeit poorly
Speaking of diversity and portfolio, I saw up this thread that you could produce a character turnaround and have look fairly "hand-made". However the exemples I saw were always in the riot/disney/blizzard soup. Have we seen AI produce actual interesting character designs? (Sorry if this has been discussed earlier, I only skimmed the topic)
Personally, I haven't seen any. They tend to be......Very overdesigned-looking. They have a lot of detail, but they tend to lack the sense an actual designer would have about where to put the detail; they don't think about having a strong silhouette, or guiding the viewer's eyes to any parts in particular, or telling anything about the characters through design elements because, of course, it's not able to think to imagine those and work out the best way to represent them.
Plus, they tend to have extremely generic faces. Asian characters in particular always seem to end up with that offputting Kpop idol look, with the obviously-had-loads-of-surgery-sameface, big doe eyes, small features, etc.
In effect, we see a permanent weakness of the AI: it can't produce meaningful designs. That requires a combination of design thinking, some real-world grounding, wanting said design to follow a theme, knowing how to balance detail, connect with the audience, art principles such as unity and variety, etc. Keep in mind we're only talking about the visual aspects here. We've not even begun to get into the other core parts with things like a backstory, said artist/author/writer wanting to give the character flaws that require human understanding (i.e. character might have a phobia of spiders which many people do and can connect with), etc. These things have just as much of an impact on the character as does the visual aspect and is all part of character design.
The AI can't do any of that stuff because it lacks an actual brain and human empathy. The AI can do design of course, but not anything meaningful. Its limited.
In effect, we see a permanent weakness of the AI: it can't produce meaningful designs. That requires a combination of design thinking, some real-world grounding, wanting said design to follow a theme, knowing how to balance detail, connect with the audience, art principles such as unity and variety, etc. Keep in mind we're only talking about the visual aspects here. We've not even begun to get into the other core parts with things like a backstory, said artist/author/writer wanting to give the character flaws that require human understanding (i.e. character might have a phobia of spiders which many people do and can connect with), etc. These things have just as much of an impact on the character as does the visual aspect and is all part of character design.
The AI can't do any of that stuff because it lacks an actual brain and human empathy. The AI can do design of course, but not anything meaningful. Its limited.
That's something I wish AI advocates would get. I see them talk often as if it learns and behaves just like a human when it's blatantly not like that at all.
It strikes me as being like trying to learn how to sculpt characters just by looking at people and never trying to understand what you're looking at. Maybe you can make something passable on first glance, but because you don't actually know why it looks like that and what's going on under the surface, you'll fall apart quite quickly.
But I wonder how much that can be mitigated. You can never get this AI stuff on the same level as a human designing with intent all the way through, but what about making some changes? I remembered that email from Daz; one of the features they advertised was basically painting in areas the AI should redo. I wonder how much mileage you could get selectively regenerating parts of an image. Enough to scrape by?
In effect, we see a permanent weakness of the AI: it can't produce meaningful designs. That requires a combination of design thinking, some real-world grounding, wanting said design to follow a theme, knowing how to balance detail, connect with the audience, art principles such as unity and variety, etc. Keep in mind we're only talking about the visual aspects here. We've not even begun to get into the other core parts with things like a backstory, said artist/author/writer wanting to give the character flaws that require human understanding (i.e. character might have a phobia of spiders which many people do and can connect with), etc. These things have just as much of an impact on the character as does the visual aspect and is all part of character design.
The AI can't do any of that stuff because it lacks an actual brain and human empathy. The AI can do design of course, but not anything meaningful. Its limited.
That's something I wish AI advocates would get. I see them talk often as if it learns and behaves just like a human when it's blatantly not like that at all.
It strikes me as being like trying to learn how to sculpt characters just by looking at people and never trying to understand what you're looking at. Maybe you can make something passable on first glance, but because you don't actually know why it looks like that and what's going on under the surface, you'll fall apart quite quickly.
But I wonder how much that can be mitigated. You can never get this AI stuff on the same level as a human designing with intent all the way through, but what about making some changes? I remembered that email from Daz; one of the features they advertised was basically painting in areas the AI should redo. I wonder how much mileage you could get selectively regenerating parts of an image. Enough to scrape by?
Its kinda the reason why atleast in AAA studios at the moment I've learned from colleagues that AI's primary use has been moodboards and to an extent generative fill.
So far there isn't a push to AI based texturing or modelling though studios are looking to develop inhouse solutions to automate many of these processes and generally the AAA pipeline is more clean up, management and maintanance than creative innovation across the board, so work will be there at whatever salary base a company maintains to remain competitive.
I'm hoping that if AI does come to that point, creativity and innovation becomes paramount or atleast artists are retained for higher project volume.
Like for example a game that took a year and 100 million to make now takes 6 months and 10 million, but the profit is still 200 million but they can have another game begin in 6 months instead of 1 year.
But seeing how AAA companies operate I'm not very optimistic about this future. I doubt governements who provide incentives to AAA studios would be keen to see large scale unemployment because of AI, then again maybe they will have to revise terms of grant funding and provide more to smaller studios and start ups.
That's something I wish AI advocates would get. I see them talk often as if it learns and behaves just like a human when it's blatantly not like that at all.
It strikes me as being like trying to learn how to sculpt characters just by looking at people and never trying to understand what you're looking at. Maybe you can make something passable on first glance, but because you don't actually know why it looks like that and what's going on under the surface, you'll fall apart quite quickly.
But I wonder how much that can be mitigated. You can never get this AI stuff on the same level as a human designing with intent all the way through, but what about making some changes? I remembered that email from Daz; one of the features they advertised was basically painting in areas the AI should redo. I wonder how much mileage you could get selectively regenerating parts of an image. Enough to scrape by?
The core part of the problem is most AI advocates don't know how the human brain works, particularly in regards to creative matters. Thus their whole argument is being made from a faulty assumption that they understand when they don't. They're partially right in that the AI can learn what its being fed, but it can't comprehend what it just learned. Now will AI get to that point to where it can? Dunno. That I believe would require it to become sentient and if that happens we're gonna have waaaaaay bigger problems than worrying about which painting it will steal next.
Its plausible one could gain good mileage using that approach as the AI can do bits and pieces well, but when it comes to cohesively making sense of something that has multiple elements it fails due to no understanding. That means the prompt guy who is selectively regenerating stuff would have to have design knowledge to piece things together in a cohesive way. How much mileage.....is hard to say. I think that would depend on a lot of things like how far along is the original image in comparison to what you want the final result to be.
Its kinda the reason why atleast in AAA studios at the moment I've learned from colleagues that AI's primary use has been moodboards and to an extent generative fill.
So far there isn't a push to AI based texturing or modelling though studios are looking to develop inhouse solutions to automate many of these processes and generally the AAA pipeline is more clean up, management and maintanance than creative innovation across the board, so work will be there at whatever salary base a company maintains to remain competitive.
I'm hoping that if AI does come to that point, creativity and innovation becomes paramount or atleast artists are retained for higher project volume.
Like for example a game that took a year and 100 million to make now takes 6 months and 10 million, but the profit is still 200 million but they can have another game begin in 6 months instead of 1 year.
But seeing how AAA companies operate I'm not very optimistic about this future. I doubt governements who provide incentives to AAA studios would be keen to see large scale unemployment because of AI, then again maybe they will have to revise terms of grant funding and provide more to smaller studios and start ups.
AI might have some usefulness for moodboards or just general noise to help start a design (i.e. avoiding a white canvas to start on), provided you already have the idea jotted down to know what direction you want to go in.
Something people tend not to think about is in this artist vs company battle that seems to be looming over the horizon is the elephant in the room: gamers. They're the ones purchasing these games to begin with which gives them some degree of power over the companies. What if gamers for example say they don't like this/that game because they think heavy AI usage makes the game visually generic/unappealing? They might just decide to boycott it or the company which could force them back into hiring human artists. Either that or they refuse to change and eventually go under due to lack of sales. Although this is being overly optimistic, one company being downed due to this might set off ripples in other companies to not get rid of their human staff or atleast not entirely.
Gamers want a cool looking game that makes sense visually and artists want to make a living at what they do. There is mutual benefit for both parties to keep AI-abusive companies in check.
This popped up on Bluesky, through X-Twitter, through facebook. I'm afraid it's authenticity is pretty much impossible to confirm.
But if inauthentic the author paints a pretty vivid (hard to argue with) picture of what trying to work as a senior dev, over ai prompters would be like. Just from what we know about ai; the direction, "Give me a side view of this character", or "Give me seven expressions", are going to be impossible for the ai to perform.
It has the stressed, overworked, cynical tone of an art director. I think it's 100% legit. That company not only hired AI bros, but ones that don't know how to use in-paint
It has the stressed, overworked, cynical tone of an art director. I think it's 100% legit. That company not only hired AI bros, but ones that don't know how to use in-paint
Inpaint will not fix lack of knowledge. It won't let you fix perspective or composition or anything bigger. If you don't know what you are doing, ai tools wont either.
Well, apparently tons of VFX artists are out of work these days. Surely there will be at least some among them willing to pick up 'prompting' and combine that with their existing skillset. From then on it's going to be more about lack of functionality in the toolset, I suppose. So the verdict in this writeup seems a bit premature to me at least.
But yeah, typing something into a prompt to create images doesn't sound like the way forward for anything but fooling around. Either the toolset can be made to fit into existing workflows - or perhaps all of that will need to be reassessed. What about a tool that takes a sketch as an input? One where you can leave labels and remarks (treat this part as xyz, leave marked area empty, etc). Uses the lineart as definitve guide, fills in everything else.
I guess what's really going to hold this back in our favour is that likely none of the big tool makers who have the experience will bother with an intricate toolset for artists when they can instead sell something semi automated and simplistic to a much larger market of non-artists.
Either the toolset can be made to fit into existing workflows - or perhaps all of that will need to be reassessed. What about a tool that takes a sketch as an input? One where you can leave labels and remarks (treat this part as xyz, leave marked area empty, etc). Uses the lineart as definitve guide, fills in everything else.
See this is what really perplexes me about this. LLM's are the tools we've been waiting for for designing interraction with NPCs in a game. Responses that conform to anything your character says! Realistic reactions when the player inevitably harrasses someone! "Oh, that character was going to give you the item you need to beat the game, but because your first word to them was 'beeeeeeeeeeeeeeeewbs!' they have opted on their own to never talk to you again." Get on the forum and the devs are like, "You're only hope is a clean new game, we have no control over what our ai thinks of you." I mean you could remake Bioshock but with a system that gives 10,000 potential endings instead of becoming UltraPitler or JesusX2.
I mean, you couldn't make Mass Effect fun to play, but almost anything else is possible!
in the time it takes to read that tweet anybody can play with midjourney or similar and reach the same conclusion. Seems kind of crazy that this would not be done before hiring a bunch of people just to use it. I mean how would a company not have their employees evaluate and report on some new toolset before hiring a bunch of people to use it? This is far fetched, right? I hope it is.
Seems more like a "story time" way to explain how AI in context of game development works to people who don't know anything about game development or AI.
It's the NFT shit again. So many people wanted to make a game with NFTs in it. The issue isn't creating a game with NFTs, that's the easy part. The hard part is making a good game in the first place, which is what even seasoned companies struggle with.
I'm personally not against AI-powered tools that aims to make an artist's work easier. Clip Studio Paint has a bunch of tools like generating shadows and other stuff I forgot about. Even though they're wonky and don't work on most artstyle, I can see myself using such tools. For the autodesk AI shape thingy, it could be nice to generate a base mesh for sculpting.
I'm personally not against AI-powered tools that aims to make an artist's work easier. Clip Studio Paint has a bunch of tools like generating shadows and other stuff I forgot about. Even though they're wonky and don't work on most artstyle, I can see myself using such tools. For the autodesk AI shape thingy, it could be nice to generate a base mesh for sculpting.
I really don't think anyone gives a crap about the tool itself. I think it's pretty neat they got the thing up and running. When Adobe put out a version stating the dataset was made from artists who were in their stable, it sounded ok. Cept the artists were never asked if they wanted to take part in providing their artwork for ai use and it turned out Adobe spread their scanning activity to all artist on the web. Like all the other bastard ai companies out there. So I don't give a flying f... about ai itself, but I do wish the fleas of a thousand camels to infest the nether regions of every prick who participated in the theft of every online creatives means to make a living.
AI Personal statement Writer | Unique Statement in a click!
Generating a Personal Statement Requires Just 5 Simple
Steps! ... Since you need a customized statement, you need to enter
details like your name, email ID, and other simple crap even an idiot like you can't stuff up.
Not just illustration but many other creative activities. The spin is obvious though abounding with oxymorons: AI Personal and Unique Statement.
the loneliness of modern society does make people weird. It is a very common thing for people to develop paranormal relationships with the characters they see on TV and other entertainment like games, whether it's actually a real person (a human actor) or totally fictional (a cartoon character)
this isn't just creepy neckbeards in the basement, it's normal people you have normal conversations with and like.
using the AI to drive wedge between working class people because of their odd quirks isn't very useful IMO. The AI is used as a weapon to take leverage away from all working class people, that's really the only thing that matters. What sexy skin it arrives in or by what algorithm it calculates responses isn't really pertinent to working class persons ability to attain work and live with some dignity. Anything which suggest people compete more and cooperate less isn't working for the benefit of people who have to work for a living.
@CyberdemoN_1542, have you tried that out? And have you made art for games?
If you done both then you'll probably realize that it's not very useful, though I'm sure it is a fun toy for people who don't already know how to model.
@CyberdemoN_1542, have you tried that out? And have you made art for games?
If you done both then you'll probably realize that it's not very useful, though I'm sure it is a fun toy for people who don't already know how to model.
I Yes to both questions. It's not great YET but who knows where it will be one year from now.
i mean the 2d AI still isn't very useful. how many years has it been with that?
About a little over a year and a half. Check out the google trends for chatGPT. Notice something? About 5 months after it debuts it's at the highest point. It goes up and down, but it never gets higher than in April of 2023. That means that pretty much everyone who is interested in chatGPT has already investigated it. For a tool that's about to overthrow the industry, they aren't exactly killing it with their user base.
But the goalposts are all set to the other side of the horizon. This thread is a great example. People keep posting doomer posts about how the end is nigh, and then someone points out that their example is actually useless garbage. Then we get told, "Well it's useless garbage now, but think what it will be in a week month year decade after the hyper trees take over and enslave the flesh beings. How about we evaluate the tool that's supposed to be ready to replace every human in the entire game industry based on what we can see it doing now?
that graph is about on point with my experience using chatgpt.
as a tool it is pretty handy for a few things but it's not in the ballpark of a game changer for me. I still easily forget to use it a lot of times where it might be handy. I don't find enough value to pay the subscription except for on and off - like if I am going to be doing a lot of c++ this month I might pay for one month because making chatgpt translate cryptic error messages saves some frustration.
midjourney and the like sometimes augment google search for me in developing a moodboard but if i was paying for some artist by the hour i'd be upset if they were spending more than few minutes getting a few references from it, because it's one of those things that can give you the veneer of making progress but actually it's just like playing a slot machine. If I could afford a full time concept artist I'd much rather have that because, for me at least, concepting is all about thinking in context and being creative, but the machine can only translate exactly what you tell it into what the most generic version of that would be.
all in all my take at this point is basically, it adds 10% more convenience to my daily work but if governments decided it was too unethical and shut it all down tomorrow, I wouldn't miss it. It handles occassional bitch work but it can't be used to help make any important decisions, and it can easily become a procrastination slot machine. It is excellent, excellent math tutor.
Based on what we've seen so far it looks like the future of AI will just be that it gets a bit faster slowly and more accurate even more slowly.
i mean the 2d AI still isn't very useful. how many years has it been with that?
And that'll likely be the limit of it. The value of AI art serves its purpose for concepting and moodboarding. It can be used for more than that though its functionality will be limited. The people who will be most impacted from the AI "taking over" will be those who do art as their day job as in being employed by someone else. Yeah the AI will likely wipe those guys out (so will a lot of other jobs to be fair) although in the grand scheme of things....most of us will be okay. Artists will still be creating whatever they want, selling their work, still getting seeked out by others to do artwork for them, etc. If anything the AI becoming commonplace makes artists more valuable. Traditional painters are more valuable than they've ever been simply because not as many people can do it anymore, but people still value a good traditional painting more than any digital painting right?
the more i think about it, the only thing i really see AI doing is giving more power to the anti-social types and owning class.
for 99% of human history, a persons value was in their character. What they do daily, and this is summarized as reputation and known locally. Reputation has to be maintained daily - you can't accrue billions of reputations and then hold that as leverage over others.
industrialization killed the age of character and now the only means of leverage a worker has is what they know. They have to try and learn something valuable and rare so they can gain enough leverage to be worth keeping around by the owning class.
if work is gone, the only thing i see being left is what effect a person can have on how others feel. This will skew value even more to anti-social types and make life for working class even more volatile in an attention driven economy as the modern consumer has the attention span of a gnat.
Basically it's an attempt to monopolize the value of labor in a knowledge-based market - the ultimate power grab - because value of labor is really only thing people have in a system where means of production are owned by a few. Whatever the large scale effect will be, I doubt it will make life more stable, more just, more fair, or more enjoyable for the majority of people.
Replies
https://www.citationneeded.news/ai-isnt-useless/
Development is usually kept behind closed doors citing competition concerns, and most games marketing is focused on release sales, which is usually when games make their money.
Live service seems to be a good work around since it keeps players involved and the game relevant and it also prioritises retention and expansion.
Though in many instances it still hasn't stopped layoffs.
Larger studios should be invested in other areas to supplement the cost and risk of game development and marketing.
And they need to respect artists more, though artists also need to realise that studio's are second homes.
We are providing a service for a fee, and its easy to get lost in the illusion that we are truly valued for our talent since in the end its a product being sold, so the real value proposition is in appealing to the customer who can be a diverse and at times toxic bunch.
Like the 400,000 people boycotting games on the SBI detected curator don't really care about developers being affected by studios losing revenue.
They do however see it as developers are better off not working for these studios and starting studios that care about the customer.
Its an extreme viewpoint that I don't feel takes into consideration everyones perspective, but they were likely the first in line to pay full price for a game they want to own.
There is consensus that most gamers want fun games regardless of socio-political messages embedded in the experience.
I feel that modding should also be made standard in games so it gives an opportunity to innovate.
I liked the idea of a large AAA company promoting smaller more focused teams to create AA budget games.
In many cases AAA company's include this in their portfolios as publisher, many have a flagship title that usually follows the recipe (like FC for EA, or Assassins Creed/FarCry for Ubisoft)
and these titles sustain other projects.
Not sure how much risk a studio can take with experimental titles, but they do have considerable room and do allocate resources to these projects.
Its just that because of NDA's no one outside usually hears of them.
In this case, indies and AA's that take a more audience focused approach with more transparency have greater leverage, so maybe for AAA they should move more into the publishing space and licensed properties which does seem to be the general direction.
And each studio usually focuses on what they do best or have more experience in, like Motive (EA) has a great level design, narrative and programming teams, so much of their character model is outsourced while they focus on integration and gameplay, I'm not seeing much initiative on building internal teams of junior/mid level artists to take on the tasks they can hand over to interns/outsourcers.
I do wish that they would balance the work load more, every senior I've met feels that they end up being burdened with a lot of work they would prefer to delegate to someone in studio rather than an outsourcer and many willingly take on this workload since they do feel they could do it best, so you need good leadership to balance this.
The visual aspect seems more a mix between style and fidelity and needs to run alongside gameplay, though it can stand on its own too.
Palworld is a good example of hitting an audience niche, the graphics animations are horrible, and there are allegations that they plaigarised assets and used AI, but the gameplay is fun which is what their mostly programmer focused team is good at.
And its also on brand, like if Naughty dog replicated Palworlds approach it would likely face a boycott. Their studio has to live up to that image which is also what the artists want.
With the layoff's we're seeing, I think that AAA studios are prefering to focus on core properties rather than developing new projects. There is no real pressure or incentive on them to innovate so long as they are guaranteed to make sales on release and through their live service windows.
Games can be made on a much learner budget and directly marketed to an audience so this is likely where the AA, indie space will stay going forward.
Here are some games to check out for their development approach,
https://www.pcgamer.com/help-im-completely-mesmerized-by-this-relaxing-castle-building-sandbox/
https://www.canarymedia.com/articles/culture/solarpunk-is-going-mainstream-this-couples-1m-kickstarter-proves-it
https://www.rockpapershotgun.com/black-and-white-bloodbath-mouse-looks-like-cuphead-as-a-first-person-shooter#:~:text=Take the Fleischer-style cartoons,hell of a stylish look.
They all focus on a very specific niche and marketing based on their original elements.
I find AAA marketing to be more brand focused, personally I can't believe how much star wars outlaw's cost or why anyone would pay that much day 1, but the branding means that many will.
AAA studios are also promoting subscription models so that's one way to rent a game to play through if you don't see yourself playing it again.
I am intriguied by the recent controversies we have going around, lot of entertaining content which actually motivates writing.
Also with so many out of work its much easier to network, reachout and improve enagement on linkedin which has proved very useful.
I have a few ideas outside of making artwork for job portfolio's though they'll likely be outside the game dev space since I've been experimenting more with 3D printing for board games.
For portfolio, atleast following my consultation with a senior artist, they said some variety would be good i.e a female character, otherwise what I have does provide more than enough evidence that I'm capable of doing the role I'm targeting (mid-senior character artist)
The hiring market is a challenge, but its good to have more free time to try new things.
I try to think of this aspect more as showcasing that I can create a good range of different characters.
DEI is supposed to impact my application positively since I'm south asian but I'm not sure if applies equally when it comes to representation among visible minorities.
Like I find that candidates with physical and learning disabilities aren't as represented if they don't also have other minority attributes and even then may not be selected over other more able bodied minorities.
I used to be a teacher for game dev students with autism hoping to working in the game industry so saw this first hand.
The most we were able to do is to have them interact with game studios through social initiatives for community building, like a studio visit or a talk by studio employee, but it didn't translate into job offers and there weren't any inititiatives targeting these groups specifically which would really help increase representation.
For example this initiative.
https://www.ubisoft.com/en-us/company/careers/interns-graduates/develop-at-ubisoft
Provides representation for women and non binary candidates and its unsettling to be asked by candidates with disabilities if they ought to identify as non binary to apply since there is no seperate initiative for them and they don't stand a chance of being hired in the regular stream.
You can read into GDC's statistics in the matter here among other statistics,
https://reg.gdconf.com/state-of-game-industry-2024
Lol did AI paint her dress on her?
This is one situation where I fully approve AI being used and I wish AI could have turned back time considering what she did.
I recommend this documentary over the Netflix one,
And yes better to stay on topic.
The state of the industry document does talk about AI impact in the game industry.
This article gives a good summary,
https://www.gamedeveloper.com/business/gdc-2024-state-of-the-game-industry-devs-discuss-layoffs-generative-ai-and-more
And certainly she wasn't always happy, she has some serious problems but the image does indicate a time when everyone was in a far better situation than the horror she caused.
The Netflix documentary has been understood to be more suspense driven and sensationalist, so this approach was expected.
On a more whimsical note, I felt that someone like Jeniffer didn't deserve a more curated upscaling of her photos, AI is as good as it going to get and the result is as twisted as she happens to be, so it kinda fits her sociopathic personality.
Also I'm still looking into what they did with that image but it seems it was run through an upscaler, so its not fully AI generated but enhanced albeit poorly
In effect, we see a permanent weakness of the AI: it can't produce meaningful designs. That requires a combination of design thinking, some real-world grounding, wanting said design to follow a theme, knowing how to balance detail, connect with the audience, art principles such as unity and variety, etc. Keep in mind we're only talking about the visual aspects here. We've not even begun to get into the other core parts with things like a backstory, said artist/author/writer wanting to give the character flaws that require human understanding (i.e. character might have a phobia of spiders which many people do and can connect with), etc. These things have just as much of an impact on the character as does the visual aspect and is all part of character design.
The AI can't do any of that stuff because it lacks an actual brain and human empathy. The AI can do design of course, but not anything meaningful. Its limited.
So far there isn't a push to AI based texturing or modelling though studios are looking to develop inhouse solutions to automate many of these processes and generally the AAA pipeline is more clean up, management and maintanance than creative innovation across the board, so work will be there at whatever salary base a company maintains to remain competitive.
I'm hoping that if AI does come to that point, creativity and innovation becomes paramount or atleast artists are retained for higher project volume.
Like for example a game that took a year and 100 million to make now takes 6 months and 10 million, but the profit is still 200 million but they can have another game begin in 6 months instead of 1 year.
But seeing how AAA companies operate I'm not very optimistic about this future.
I doubt governements who provide incentives to AAA studios would be keen to see large scale unemployment because of AI, then again maybe they will have to revise terms of grant funding and provide more to smaller studios and start ups.
The core part of the problem is most AI advocates don't know how the human brain works, particularly in regards to creative matters. Thus their whole argument is being made from a faulty assumption that they understand when they don't. They're partially right in that the AI can learn what its being fed, but it can't comprehend what it just learned. Now will AI get to that point to where it can? Dunno. That I believe would require it to become sentient and if that happens we're gonna have waaaaaay bigger problems than worrying about which painting it will steal next.
Its plausible one could gain good mileage using that approach as the AI can do bits and pieces well, but when it comes to cohesively making sense of something that has multiple elements it fails due to no understanding. That means the prompt guy who is selectively regenerating stuff would have to have design knowledge to piece things together in a cohesive way. How much mileage.....is hard to say. I think that would depend on a lot of things like how far along is the original image in comparison to what you want the final result to be.
So far there isn't a push to AI based texturing or modelling though studios are looking to develop inhouse solutions to automate many of these processes and generally the AAA pipeline is more clean up, management and maintanance than creative innovation across the board, so work will be there at whatever salary base a company maintains to remain competitive.
I'm hoping that if AI does come to that point, creativity and innovation becomes paramount or atleast artists are retained for higher project volume.
Like for example a game that took a year and 100 million to make now takes 6 months and 10 million, but the profit is still 200 million but they can have another game begin in 6 months instead of 1 year.
But seeing how AAA companies operate I'm not very optimistic about this future.
I doubt governements who provide incentives to AAA studios would be keen to see large scale unemployment because of AI, then again maybe they will have to revise terms of grant funding and provide more to smaller studios and start ups.
AI might have some usefulness for moodboards or just general noise to help start a design (i.e. avoiding a white canvas to start on), provided you already have the idea jotted down to know what direction you want to go in.
Something people tend not to think about is in this artist vs company battle that seems to be looming over the horizon is the elephant in the room: gamers. They're the ones purchasing these games to begin with which gives them some degree of power over the companies. What if gamers for example say they don't like this/that game because they think heavy AI usage makes the game visually generic/unappealing? They might just decide to boycott it or the company which could force them back into hiring human artists. Either that or they refuse to change and eventually go under due to lack of sales. Although this is being overly optimistic, one company being downed due to this might set off ripples in other companies to not get rid of their human staff or atleast not entirely.
Gamers want a cool looking game that makes sense visually and artists want to make a living at what they do. There is mutual benefit for both parties to keep AI-abusive companies in check.
Seems more like a "story time" way to explain how AI in context of game development works to people who don't know anything about game development or AI.
AI Personal statement Writer | Unique Statement in a click!
this isn't just creepy neckbeards in the basement, it's normal people you have normal conversations with and like.
using the AI to drive wedge between working class people because of their odd quirks isn't very useful IMO. The AI is used as a weapon to take leverage away from all working class people, that's really the only thing that matters. What sexy skin it arrives in or by what algorithm it calculates responses isn't really pertinent to working class persons ability to attain work and live with some dignity.
Anything which suggest people compete more and cooperate less isn't working for the benefit of people who have to work for a living.
https://www.meshy.ai/
https://www.linkedin.com/posts/luc-schurgers-1965063_skibidisam-sigmagpt-zoomerspeak-ugcPost-7199094600131563521-hEbL?utm_source=share&utm_medium=member_desktop
@CyberdemoN_1542, have you tried that out? And have you made art for games?
If you done both then you'll probably realize that it's not very useful, though I'm sure it is a fun toy for people who don't already know how to model.
Notice something? About 5 months after it debuts it's at the highest point. It goes up and down, but it never gets higher than in April of 2023. That means that pretty much everyone who is interested in chatGPT has already investigated it. For a tool that's about to overthrow the industry, they aren't exactly killing it with their user base.
But the goalposts are all set to the other side of the horizon. This thread is a great example. People keep posting doomer posts about how the end is nigh, and then someone points out that their example is actually useless garbage. Then we get told, "Well it's useless garbage now, but think what it will be in a week month year decade after the hyper trees take over and enslave the flesh beings. How about we evaluate the tool that's supposed to be ready to replace every human in the entire game industry based on what we can see it doing now?
'Cause what it can do now ain't shit.
as a tool it is pretty handy for a few things but it's not in the ballpark of a game changer for me. I still easily forget to use it a lot of times where it might be handy. I don't find enough value to pay the subscription except for on and off - like if I am going to be doing a lot of c++ this month I might pay for one month because making chatgpt translate cryptic error messages saves some frustration.
midjourney and the like sometimes augment google search for me in developing a moodboard but if i was paying for some artist by the hour i'd be upset if they were spending more than few minutes getting a few references from it, because it's one of those things that can give you the veneer of making progress but actually it's just like playing a slot machine. If I could afford a full time concept artist I'd much rather have that because, for me at least, concepting is all about thinking in context and being creative, but the machine can only translate exactly what you tell it into what the most generic version of that would be.
all in all my take at this point is basically, it adds 10% more convenience to my daily work but if governments decided it was too unethical and shut it all down tomorrow, I wouldn't miss it. It handles occassional bitch work but it can't be used to help make any important decisions, and it can easily become a procrastination slot machine. It is excellent, excellent math tutor.
Based on what we've seen so far it looks like the future of AI will just be that it gets a bit faster slowly and more accurate even more slowly.
And that'll likely be the limit of it. The value of AI art serves its purpose for concepting and moodboarding. It can be used for more than that though its functionality will be limited. The people who will be most impacted from the AI "taking over" will be those who do art as their day job as in being employed by someone else. Yeah the AI will likely wipe those guys out (so will a lot of other jobs to be fair) although in the grand scheme of things....most of us will be okay. Artists will still be creating whatever they want, selling their work, still getting seeked out by others to do artwork for them, etc. If anything the AI becoming commonplace makes artists more valuable. Traditional painters are more valuable than they've ever been simply because not as many people can do it anymore, but people still value a good traditional painting more than any digital painting right?
I think the majority of us will be fine.
for 99% of human history, a persons value was in their character. What they do daily, and this is summarized as reputation and known locally. Reputation has to be maintained daily - you can't accrue billions of reputations and then hold that as leverage over others.
industrialization killed the age of character and now the only means of leverage a worker has is what they know. They have to try and learn something valuable and rare so they can gain enough leverage to be worth keeping around by the owning class.
if work is gone, the only thing i see being left is what effect a person can have on how others feel. This will skew value even more to anti-social types and make life for working class even more volatile in an attention driven economy as the modern consumer has the attention span of a gnat.
Basically it's an attempt to monopolize the value of labor in a knowledge-based market - the ultimate power grab - because value of labor is really only thing people have in a system where means of production are owned by a few. Whatever the large scale effect will be, I doubt it will make life more stable, more just, more fair, or more enjoyable for the majority of people.