Why don't indie devs use AI-generated images as art?

[V17]

Once again you're not talking about quality, you're not talking about coherence, about training artists to be better, about making art people enjoy, nor about exceeding expectations, nor anything that makes art beautiful or touching. You're not interested on the fact that creative vision can transform a game, or that creativity is often seen as the most pleasant part of concepts design.

You're talking theoretically about something you don't understand. You don't understand AI. You don't see that most art and design is already in a quagmire, that mediocrity and uniformity have taken over all creative endeavor. Go on, boast about the fact that AI lets you shit designs quicker than an amoebia-ridden colon, rejoice in the great look-alike CGI pictures carpet-bombing, revel in your own inability to experience critical thinking and beauty appreciation. You're a bean counter. You want to optimize workflows. You care about money, not people, not even results, you're boring and bring nothing of value, you talk with internet keywords and does not seem to show any specie of original thoughts. You could be replaced by some AI.

I'm not talking about quality etc. because those are not relevant to the workflow. There is no reason why it should bring mediocrity. The rest of your post is made up bullshit, so I'm just going to say it's funny how on a forum where laughing at unproductive devs indefinitely delaying release is a meme there are simultaneously people complaining so much about tools that have the potential to allow artists to work faster with no drop in quality. But perhaps those two groups do not overlap.

This result is less useful than a Google image search. (Read the title of the thread).

Of course the generated art is rubbish and no use with hiring an artist on top of this (which begs whats the point?).

My thought is: are there *any* use for this stuff at all? I have my doubts that for anything this algorithm produces there isn't already an existing algorithm that outperforms it.

You've already been told several times, but let's try it again. When creating game art (or vfx or achviz or product viz or...) there's generally a concept art stage, which is where most of the creative work needs to be done. That creativity doesn't come from thin air, there are processes that artists use. And AIs like Midjourney can make that workflow significantly faster because with a skilled user it's a powerful and fast inspiration machine. In some cases its output can be used straight up just with some editing, in other cases it's just used as a base for something else. The AI is designed for this type of work and allows for things like uploading your own image into it and having it create several variations, recreating it in a different style, in higher resolutions etc. Don't expect people to make examples just for you because these services naturally cost money to use.

You're talking theoretically about something you don't understand. The reality is that the technology is already being used and when in a few years you see the graphics made by artists who used it, you won't even know about it.

Either take a shit or get off the toilet. A person can draw inspiration from anything, so basically its the people doing all the difficult stuff and your "AI" isn't bringing much to the table. I'm still struggling to see the use, and theres no good examples (that aren't fluke one in a million results). All the online apps I've tried have toilet results.

As for more practical uses for someone like me, a developer who has to pay for artists, this retarded algorithm can't even generate a decent face or isometric building, how on earth could I use this?

Yes, a person can draw inspiration from anything. Except this tool is in many cases much faster, more malleable and gives you reasonably original results that you can use in any way without licensing or copyright difficulties.

You're basically complaining that people who already use the tools to make money aren't polite enough to stop and use their time to show you how they work instead, for free. Just be a little patient - Midjourney and Dall-E have just been available for a couple months and Dall-E still has a pretty annoying waitlist. People love talking about their work and eventually videos showing their practical use will start popping up. I've already seen a few from archviz people, but not from games yet.

If your whole issue is "why can't *I* use this??", I doubt anyone can help you with that.

 

[The_Sloth_Sleeps]

Once again you're not talking about quality, you're not talking about coherence, about training artists to be better, about making art people enjoy, nor about exceeding expectations, nor anything that makes art beautiful or touching. You're not interested on the fact that creative vision can transform a game, or that creativity is often seen as the most pleasant part of concepts design.

You're talking theoretically about something you don't understand. You don't understand AI. You don't see that most art and design is already in a quagmire, that mediocrity and uniformity have taken over all creative endeavor. Go on, boast about the fact that AI lets you shit designs quicker than an amoebia-ridden colon, rejoice in the great look-alike CGI pictures carpet-bombing, revel in your own inability to experience critical thinking and beauty appreciation. You're a bean counter. You want to optimize workflows. You care about money, not people, not even results, you're boring and bring nothing of value, you talk with internet keywords and does not seem to show any specie of original thoughts. You could be replaced by some AI.

I'm not talking about quality etc. because those are not relevant to the workflow. There is no reason why it should bring mediocrity. The rest of your post is made up bullshit, so I'm just going to say it's funny how on a forum where laughing at unproductive devs indefinitely delaying release is a meme there are simultaneously people complaining so much about tools that have the potential to allow artists to work faster with no drop in quality. But perhaps those two groups do not overlap.

This result is less useful than a Google image search. (Read the title of the thread).

Of course the generated art is rubbish and no use with hiring an artist on top of this (which begs whats the point?).

My thought is: are there *any* use for this stuff at all? I have my doubts that for anything this algorithm produces there isn't already an existing algorithm that outperforms it.

You've already been told several times, but let's try it again. When creating game art (or vfx or achviz or product viz or...) there's generally a concept art stage, which is where most of the creative work needs to be done. That creativity doesn't come from thin air, there are processes that artists use. And AIs like Midjourney can make that workflow significantly faster because with a skilled user it's a powerful and fast inspiration machine. In some cases its output can be used straight up just with some editing, in other cases it's just used as a base for something else. The AI is designed for this type of work and allows for things like uploading your own image into it and having it create several variations, recreating it in a different style, in higher resolutions etc. Don't expect people to make examples just for you because these services naturally cost money to use.

You're talking theoretically about something you don't understand. The reality is that the technology is already being used and when in a few years you see the graphics made by artists who used it, you won't even know about it.

Either take a shit or get off the toilet. A person can draw inspiration from anything, so basically its the people doing all the difficult stuff and your "AI" isn't bringing much to the table. I'm still struggling to see the use, and theres no good examples (that aren't fluke one in a million results). All the online apps I've tried have toilet results.

As for more practical uses for someone like me, a developer who has to pay for artists, this retarded algorithm can't even generate a decent face or isometric building, how on earth could I use this?

Yes, a person can draw inspiration from anything. Except this tool is in many cases much faster, more malleable and gives you reasonably original results that you can use in any way without licensing or copyright difficulties.

You're basically complaining that people who already use the tools to make money aren't polite enough to stop and use their time to show you how they work instead, for free. Just be a little patient - Midjourney and Dall-E have just been available for a couple months and Dall-E still has a pretty annoying waitlist. People love talking about their work and eventually videos showing their practical use will start popping up. I've already seen a few from archviz people, but not from games yet.

If your whole issue is "why can't *I* use this??", I doubt anyone can help you with that.

So basically, this works like self driving cars. Google & Apple are making them, and they are just 2 months away from being released to the public. I see.

 

[gurugeorge]

The argument here is a bit analogous to the use of tracing in comics. Probably the majority of comic book artists had a lightbox and used a bit of tracing here and there to cut corners and speed up their workflow, and nobody noticed or cared much. It was only when you had an artist like Greg Land, whose work eventually became so obviously stuffed to the gills with tracing (of other artists and of porn models), that people sat up and took notice.

Similarly, with the use of AI, it's obviously going to be helpful for a good artist to be able to use it where lots of duplication with slight variations are required, or to get inspiration, or whatever. But people will notice when it's over-used, or used as a crutch by a bad artist - AI stuff generally has an uncanny valley feel, and the more of it there is the more you notice it (i.e. you might not notice it in one picture in isolation, but you'd notice it in a series, just as with AI poetry or literature, you can be fooled for a bit, but it doesn't take long before your mind hits an uncanny valley speed bump).

 

[The_Sloth_Sleeps]

The argument here is a bit analogous to the use of tracing in comics. Probably the majority of comic book artists had a lightbox and used a bit of tracing here and there to cut corners and speed up their workflow, and nobody noticed or cared much. It was only when you had an artist like Greg Land, whose work eventually became so obviously stuffed to the gills with tracing (of other artists and of porn models), that people sat up and took notice.

Similarly, with the use of AI, it's obviously going to be helpful for a good artist to be able to use it where lots of duplication with slight variations are required, or to get inspiration, or whatever. But people will notice when it's over-used, or used as a crutch by a bad artist - AI stuff generally has an uncanny valley feel, and the more of it there is the more you notice it (i.e. you might not notice it in one picture in isolation, but you'd notice it in a series, just as with AI poetry or literature, you can be fooled for a bit, but it doesn't take long before your mind hits an uncanny valley speed bump).

Well the same is true of any computer generated stuff. Those computer generated dungeons or mazes, its helpful in that not every dungeon has to be hand made and everyones experience is different. The fractal landscapes for flight simulators and so on. Yes from that point of view procedural generation is useful. (You can even chuck in a buzz word and call it "AI" generated and I would argue that 90% of AI you see is essentially the same thing as PG, and the "AI" part plays a minimal role in the generation).

Anyways, for artistic endeavors, whether or not its helpful to an decent artist....under most circumstances I doubt it - a lot.

 

[J1M]

They didn't fail because they've never even been tested. The main issue is the people constantly pushing the "SELF DRIVING CARS WILL BE HERE IN 5 10 15 20 25 30 SOMETIME SOON YEARS!!!!". These things, they take time.

Must have been over twenty years since I've first heard about self-driving cars trials. They do take time indeed and I'm sure I'll never see them in my lifetime, not without a brand new technological approach.

Current neural networks basically are brand new. They're a very old concept, but large scale GPU computation of neural networks that revolutionized their research only became relatively cheap in 2010s. We're still in the "low hanging fruit" phase of research where small breakthroughs happen pretty regularly. Judging self-driving cars (or any other current neural net applications) on that just because Musk has a big mouth does not make any sense.

AIs usable for game-ready graphics will undoubtedly happen too at some point, it's just a question if there's enough money in it for somebody to do the research and development sometime soon of if we're going to have to wait till somebody else does most of the work on the tech for creating similar but more profitable stuff. Imo even the usability of neural nets for concept art workflow is a huge thing on its own.

Also the tech evolves really fast. Artflow referenced above was launched a year ago and is quite limited, afaik they specialize on portraits and landscapes (which indeed makes it a big surprise they can't make an isometric house), where as Craiyon I think launched like two months ago and is far from state of the art, but it at least already understands what an isometric house is:

This result is less useful than a Google image search. (Read the title of the thread).

Of course the generated art is rubbish and no use with hiring an artist on top of this (which begs whats the point?).

My thought is: are there *any* use for this stuff at all? I have my doubts that for anything this algorithm produces there isn't already an existing algorithm that outperforms it.

You've already been told several times, but let's try it again. When creating game art (or vfx or achviz or product viz or...) there's generally a concept art stage, which is where most of the creative work needs to be done. That creativity doesn't come from thin air, there are processes that artists use. And AIs like Midjourney can make that workflow significantly faster because with a skilled user it's a powerful and fast inspiration machine. In some cases its output can be used straight up just with some editing, in other cases it's just used as a base for something else. The AI is designed for this type of work and allows for things like uploading your own image into it and having it create several variations, recreating it in a different style, in higher resolutions etc. Don't expect people to make examples just for you because these services naturally cost money to use.

You're talking theoretically about something you don't understand. The reality is that the technology is already being used and when in a few years you see the graphics made by artists who used it, you won't even know about it.

Either take a shit or get off the toilet. A person can draw inspiration from anything, so basically its the people doing all the difficult stuff and your "AI" isn't bringing much to the table. I'm still struggling to see the use, and theres no good examples (that aren't fluke one in a million results). All the online apps I've tried have toilet results.

As for more practical uses for someone like me, a developer who has to pay for artists, this retarded algorithm can't even generate a decent face or isometric building, how on earth could I use this?

Let's start with two assumptions, which are easy enough to infer from the history of game releases:

• Design-by-committee rarely produces anything novel or pushes design forward. Individual genius or enthusiasm among a small homogenous group is responsible for the vast majority of gaming experiences worth remembering.
• Translating the creative vision of a few people into a commercial product becomes more and more difficult as the team size grows and the percentage of the team that are visionaries decreases.

Therefore, any tool or technique that allows you to reduce the size of the team will increase the overall team homogeneity and the chances of that vision surviving to the end of the project.

TLDR: Productive tools -> smaller teams -> better games

 

[Pyke]

A big issue is just usability. You can get some really interesting designs and stills - but apart from using it as some elements in a game, its not really 'there' yet. Like - for background loading scenes perhaps, and character portraits - but if you want to take the step where you have a game character, sprite, animation, etc - AI art isn't useful. It's basically stock photographs. You could make a game if you had access to stock photography - but only very specific games could be made.

To the people it IS already useful to, they have artistic skill nessesary to USE it in interesting ways. But it's not going to replace game artists, or really help non artistic developers create usable game assets in its current form.

are you aware of any AI art gen that will produce *.AI/PSD/EPS/XCF with different design elements preserved as individual layers? Giving artists access to the layers would go a long way to increasing the usability of these machine generated images, but I imagine it's a huge computational task and the technology isn't there yet.

I've only played around with Midjourney which flattens the image. AFAIK however, the way that AI works isn't with thing like layers. It creates the images through noise algorithms, and assigns weighted values to the noise to get the images.
However, to break an image down into layers wouldnt be too difficult. Photoshop has smart functions which will fill in blank areas with aproximations of pixels. Artists have been painting out errors and creating clean plates forever. Hell, in Matte Painting its like 90% of what you do - break an image into layers, and recreate parts that appear 'behind' something.

I DO think that this tech WILL come very far very fast. Hell, in 6 months time all of the things I've mentioned as potential blockages could be completely usurped by a single program.
Where I do think this will go will be in creating variations of existing work. So like, I could produce a model of a house - and AI takes my model, and produces 100 variations with the same style and polycount and generated textures. Same with characters, scenery, etc. Hell - it could even start doing this sort of thing on the fly in real time. AI generated quests in any RPG could potentially be more intricate and reactive than anything a human could create. Now THATS exciting stuff!

 

[J1M]

It is, but I don't see these algorithm tinkerers viewing the tech as a productivity tool. They are all treating it as a novelty to drive social media hits.

 

[Zibniyat]

All the pictures of AI created things ITT look ugly. They are always invariably too dark and creepy. And just plain ugly.

 

[V17]

An experimental shmup "game" made entirely with AI images that seem to be unedited apart from alpha masks: https://unityroom.com/games/shoon
A video and images of prompts that did not make it into the game plus some information in Japanese here: https://twitter.com/Nao_u_/status/1558595111147425792

It is unsurprisingly pretty bad. There is some coherence in the overall style, but since the images are unedited, enemy ships don't seem to make much sense, plus there are no animations. And the gameplay is not very refined to say the least.

 

[Üstad]

https://beta.dreamstudio.ai/

Better than craiyon but slightly worse than MidJourney when it comes to drawing anything but humans.

 

[V17]

That's the other news.
Stable Diffusion is out. You can try it online on the Dreamstudio link above or through some google collab notebooks, but you can also just download it and run it locally if you have an NVidia GPU with at least 4 GB RAM (supposedly some people managed to run it on an AMD GPU, but it's not officially supported yet at least).
It has no user friendly GUI yet, it's used through a command prompt, but it's not difficult to run. On my GTX1060 a set of 10 512x512 px images takes 12 minutes to generate with the default setting. Apart from text to image, it can also be used as text guided image to image, which I haven't tried yet.

I haven't had the chance to try Dall-E 2 yet, but the difference between Craiyon, Midjourney and Stable Diffusion seems to be:
Craiyon is only usable for fun, the outputs are not good enough for anything, but it's very tolerant to the precision of text prompts. It tries to generate something sensible out of absolutely any text.
Stable Diffusion is close to Dall-E in quality in some ways, it's good at imitating the styles of various artists, pretty good at creating humans and all in all much more realistic. But it's very sensitive to text prompts, I've had some prompts generate what was essentially just noise. And it tends to be more realistic than creative/artistic.
"Old" (stable) version of Midjourney is more artistic and creative, but many images of things like landscapes have a recognizable Midjourney style to them (kind of a dark/magical thing). New beta version of Midjourney is based on Stable Diffusion underneath and seems to be much better in this regard, a good compromise between the two. However it costs money.

Obviously the most important thing about Stable Diffusion is that it can generate artificial titties. Some examples here.

 

[The_Sloth_Sleeps]

An experimental shmup "game" made entirely with AI images that seem to be unedited apart from alpha masks: https://unityroom.com/games/shoon
A video and images of prompts that did not make it into the game plus some information in Japanese here: https://twitter.com/Nao_u_/status/1558595111147425792

It is unsurprisingly pretty bad. There is some coherence in the overall style, but since the images are unedited, enemy ships don't seem to make much sense, plus there are no animations. And the gameplay is not very refined to say the least.

This tells us a lot about the "state of the art".

 

[V17]

Well it certainly tells you "it's not possible to make a decent game using only unedited AI pictures in 3 days", but then again, nobody was ever claiming that.

You can get world war 2 with cats, which is more important anyway.

 

[KateMicucci]

The AIs I've seen create decent thumbnail images, but once you zoom in you start to see all of the weird shit. Right now you'd still need an artist to touch them up. Right now.

They are useful for finding good-looking palettes and compositions that an artist can steal without anybody noticing. I've made a few AI images that were nonsense zoomed in but I wanted to paint them into something real.

 

[Üstad]

Stable Diffusion is out. You can try it online on the Dreamstudio link above or through some google collab notebooks, but you can also just download it and run it locally if you have an NVidia GPU with at least 4 GB RAM (supposedly some people managed to run it on an AMD GPU, but it's not officially supported yet at least).

I'm using the online link version, signed up with throwaway discord account, that was easier for me compared to grinding my GPU. Homever I still don't know what is the free trial limit. Checked the FAQ but that didn't help.

 

[The_Sloth_Sleeps]

I tried Dreamstudio. I went with a two approaches. Real life scene + abstract backdrop.

The output it appears to be not really usable if you are going for life depiction ("Medieval battle scene oil painting"). The illusion is quickly given away and you realize how shallow the algorithm is. Horrendous and nonsensical outputs (men with chicken feet, missing limbs, backwards heads etc).

It performed slightly better at more abstract prompts ("Medieval book border ornate"). Still not any use to me as I would have to spend more time fixing up the results than it would take to begin anew.

Beyond mere entertainment value of generating silliness, I do not feel this tool is much use. However, if there is a use its probably something to do with this "abstract/noise" aspect. Generating imagery with a lot of noise where the human eye cannot discern. Even then it will "that" look of being algorithmically generated, so best avoided.

 

[Outlander]

https://beta.dreamstudio.ai/

Better than craiyon but slightly worse than MidJourney when it comes to drawing anything but humans.

"Chris Avellone with his harem"

 

[Üstad]

https://beta.dreamstudio.ai/

Better than craiyon but slightly worse than MidJourney when it comes to drawing anything but humans.

View attachment 27379

"Chris Avellone with his harem"

Midjourney is worse when it comes to human drawing. Like eyes and limbs missing being too common

 

[Assisted Living Godzilla]

Well, it's only fairly recently that it looks good is probably why. However, in the future, if some developers needs to generate a bunch of painting to put in the environment, they'll probably use it.

 

[V17]

I tried Dreamstudio. I went with a two approaches. Real life scene + abstract backdrop.

The output it appears to be not really usable if you are going for life depiction ("Medieval battle scene oil painting"). The illusion is quickly given away and you realize how shallow the algorithm is. Horrendous and nonsensical outputs (men with chicken feet, missing limbs, backwards heads etc).

It performed slightly better at more abstract prompts ("Medieval book border ornate"). Still not any use to me as I would have to spend more time fixing up the results than it would take to begin anew.

Beyond mere entertainment value of generating silliness, I do not feel this tool is much use. However, if there is a use its probably something to do with this "abstract/noise" aspect. Generating imagery with a lot of noise where the human eye cannot discern. Even then it will "that" look of being algorithmically generated, so best avoided.

So, serious non-sarcastic answer:

I don't do 2D graphics, so I can't speak for myself about that, although I have talked to concept artists who use it mainly for scifi scenes and have been able to wrangle scenes good enough from Midjourney that getting the scene + painting over it offered a considerably faster workflow than painting it from scratch. It also seems to work pretty decently for semi-abstract highly stylized art.

A big part of Midjourney specifically is that you can upload any image into it and let it generate n images that are similar in style and composition. Together with the other tools it's useful for narrowing down a certain style. Stable Diffusion cannot do this yet, although it can do img2img, where you supply a base image together with a text prompt, and it tries to generate an image described with the text prompt like normal, except normally it starts with noise, whereas here it starts with the provided image which ten serves as a composition guide.


Anyway, I do 3D graphics and have been able to get useful things for 3D art from Stable Diffusion. I don't do game graphics, I do either archviz/product viz or some more artistic renders for fun, so the work is not just about modelling and texturing, a big part of it is lighting, composition, color schemes to evoke a certain feeling, styles of post-processing etc. The AI is pretty good at generating images with interesting details that can be used in that process. I don't intend to use the actual generated images, so errors in generated results are not important, what's important is that it's visually close enough to what I'm trying to create and that it can show me interesting or unexpected things that I can adopt and recreate - the imperfections often help with that.

Learning how to write prompts that work well is a bit of a pain though, it takes time.

 

[The_Sloth_Sleeps]

I tried Dreamstudio. I went with a two approaches. Real life scene + abstract backdrop.

The output it appears to be not really usable if you are going for life depiction ("Medieval battle scene oil painting"). The illusion is quickly given away and you realize how shallow the algorithm is. Horrendous and nonsensical outputs (men with chicken feet, missing limbs, backwards heads etc).

It performed slightly better at more abstract prompts ("Medieval book border ornate"). Still not any use to me as I would have to spend more time fixing up the results than it would take to begin anew.

Beyond mere entertainment value of generating silliness, I do not feel this tool is much use. However, if there is a use its probably something to do with this "abstract/noise" aspect. Generating imagery with a lot of noise where the human eye cannot discern. Even then it will "that" look of being algorithmically generated, so best avoided.

So, serious non-sarcastic answer:

I don't do 2D graphics, so I can't speak for myself about that, although I have talked to concept artists who use it mainly for scifi scenes and have been able to wrangle scenes good enough from Midjourney that getting the scene + painting over it offered a considerably faster workflow than painting it from scratch. It also seems to work pretty decently for semi-abstract highly stylized art.

A big part of Midjourney specifically is that you can upload any image into it and let it generate n images that are similar in style and composition. Together with the other tools it's useful for narrowing down a certain style. Stable Diffusion cannot do this yet, although it can do img2img, where you supply a base image together with a text prompt, and it tries to generate an image described with the text prompt like normal, except normally it starts with noise, whereas here it starts with the provided image which ten serves as a composition guide.


Anyway, I do 3D graphics and have been able to get useful things for 3D art from Stable Diffusion. I don't do game graphics, I do either archviz/product viz or some more artistic renders for fun, so the work is not just about modelling and texturing, a big part of it is lighting, composition, color schemes to evoke a certain feeling, styles of post-processing etc. The AI is pretty good at generating images with interesting details that can be used in that process. I don't intend to use the actual generated images, so errors in generated results are not important, what's important is that it's visually close enough to what I'm trying to create and that it can show me interesting or unexpected things that I can adopt and recreate - the imperfections often help with that.

Learning how to write prompts that work well is a bit of a pain though, it takes time.

I am not a concept artist. If I want inspiration I can simply go to my reference books. I don't need a mashup picture, its less than helpful. I am struggling to see the point of all this TBH. If its helpful to some people then great, but it all seems a bit...niche.

Regarding 3D graphics...can these algorithms can generate 3D models?

 

[V17]

I am not a concept artist. If I want inspiration I can simply go to my reference books. I don't need a mashup picture, its less than helpful. I am struggling to see the point of all this TBH. If its helpful to some people then great, but it all seems a bit...niche.

Regarding 3D graphics...can these algorithms can generate 3D models?

Concept art is not exactly niche. There are many areas where reference books may not be enough, and while various sources of inspiration are essentially infinite, this one is pretty useful in some ways because it's fast and not license protected.

There is nothing similar to this that generates 3D models yet as far as I know.

 

[The_Sloth_Sleeps]

I am not a concept artist. If I want inspiration I can simply go to my reference books. I don't need a mashup picture, its less than helpful. I am struggling to see the point of all this TBH. If its helpful to some people then great, but it all seems a bit...niche.

Regarding 3D graphics...can these algorithms can generate 3D models?

Concept art is not exactly niche. There are many areas where reference books may not be enough, and while various sources of inspiration are essentially infinite, this one is pretty useful in some ways because it's fast and not license protected.

There is nothing similar to this that generates 3D models yet as far as I know.

If its useful to you then great. I am in favour of making better tools to support my artistic undertakings. These algorithms, I don't really get it though. The problem I really have with this all nonsense is the hype surrounding it. That's what bugs me. Its just utterly and massively underwhelming - these hyped up things never do anything near what they claim. They are just wank fantasies. Anything hyped turns out to be utterly disappointing. Same as this. And it really grinds on me. Our supposed "replacements" can't even get the mere basics right, what chance do they have at anything superior? Its embarrassingly rubbish. I don't think the guys making these algorithms do enough to put what they are attempting into perspective. They kind of encourage this "mistique" around their mashupware. Theres not even the hope of a promise these will ever improve enough to replace artists.

 

[J1M]

The hype is simple. Collect some quotes for outsourced artwork and you will understand.

1) $$$ (obvious)
2) time per draft/iteration (less obvious)
3) an algorithm won't quit or have a busy schedule or require a salary to keep on retainer (dealing with contracts and tax law makes it hard for founders to scale up to a full team and maintain a clear creative vision)

 

[Machocruz]

For production artists and illustrators, it's the next step from running reference photos through filters and/or randomizer in Photoshop to generate ideas, color schemes, compositions, which digital artists have been doing for ages. The random nature of what it produces could be useful, another form of staring at clouds or DaVinci looking at stains and cracks on walls to evoke images of various forms in the mind.