If you respect the art of game development in any way, please, do not use AI Art.

[123edc]

AI as a tool become several times better with the knowledge. That's why I think it's important to learn them.

in the end, it's a question about individual goals ...

if i get 100 pictures out of my ai, i don't need to "learn" which color evokes which emotion ... i'll know, which picture fits best to the story and scene i have in my head ... if you get what i mean

could it help me, to know the fundamentals? maybe
is it neccacary, if my goal is, to get my hands on affordable artwork fitting towards the scene, i want to portray? ... no

It means an indie dev either has to find another person, or change genre, or settle with less options.

and exactly that's the point of the entire conversation,
ai - art opens up another option aside from buying a artist or changing the game/vision you have

not more - not less,

I play with ai text to image sometimes, from my experience text prompt almost never accurately solve my problems.

1] yes, it still is in it's children shoes ... but the progress made over the past couple of month can't exactly be denied

2] then we have different experiences, becouse from my experience i already got some quite nice anime CG's out of it ... sure, it still takes time ... but it is a] more custom tailored towards your setting then the generic dlc's and b] more affordable, then hiring an artist for each individual picture you need ...

and sure, i could invest 9 years part-time studying art theory in university ... but then my game probably would never get finished ...

I often need to fix AI output in photoshop, because AI output often has result that is illogical, sometimes they don't apply many design ideas that I have in mind even though they should.

that is true, especially the hands are often messed up and need adjusting [though, they are working on it]
... however, as said above, the technology gets better and better over time

and there are actual "solutions" to some problems ...
sometimes, you can use pre-made hands ... or use your promts in a way, that don't show the wobbly hands ...
also, there are tags, to include the "good" sample pieces and exclude the "bad" samples [like those, tagged by ppl as anatomically wrong] which can massively increase the outputs quality

also, one realy interesting approach one of these AI's is trying to go recently is, to allow the user to "scribble" over the painting [like a small child would xD] and let the ai fix up the parts, you overwrite ...

so yeah ... we're just at the start of it,

 

[Tamina]

and sure, i could invest 9 years part-time studying art theory in university ... but then my game probably would never get finished ...

The good news is....You don't have to spend 9 years in university to learn art fundamentals. I didn't go to art university myself, my knowledge came from short 1-3 hr online lessons about composition and color theory, which explains why things work the certain way.

game designer is a position that needs to learn a little bit of everything, since we need to handle everything in game dev kinda, especially in small teams. In the end I feel the time investment is worth it.

that is true, especially the hands are often messed up and need adjusting [though, they are working on it]

To me the problem is more than hands error: that is easy to fix and I think it will be completely fixed soon with some of the new tools that I read about.

My problem is.... I often have a very specific composition, gesture and focal point in mind but it is incredibly frustrating to communicate all that info using text to image accurately. AI often likes give me 10 images but none of them matches nor beat the solution that I already have.

AI often make mistakes on lighting logic, values and man-made object structure, such as tons of buildings on the street....

Then I have to use the art skill and sketch them or fix them in the end........

Maybe after many years the tools will improve to be 100% accurate with built in knowledge to solve all the problems with 1 click. But I would rather work on my games now than wait and pray.....

 

[123edc]

I often have a very specific composition, gesture and focal point in mind but it is incredibly frustrating to communicate all that info using text to image accurately.

ah, i see ... that's true,

the main issue in this case is, that most ai's are still trained on just a few of the realy, realy big websites out there

and at the moment, you still mostly need their exact right wording, they use ...
a simple _ instead of a space bar can realy mess things up ^.^ [and sometimes the tagging of certain stuff can be a bit ... well]

but then again ... that's a matter, that will be resolved, wit more samples and better technology,

AI often make mistakes on lighting logic, values and man-made object structure, such as tons of buildings on the street....

hmh, ok ... i haven't realy tried large buildings or stuff like that [since i don't realy need it anyway ... that stuff is already covered via the tilesets ^.^]

but as for anime characters / busts ... large landscapes and especially close ups it works realy well in my opinion ... also, yet again ... you need to figure out the right tags for the right things ... which is it's own science in and on itself xD


... but than again, back to squere one ...
in the end it's all about the question, what you actually need and what you actually (expect to) get out of it ...

 

[Tai_MT]

I think "learning art fundamentals" isn't really all that necessary for most people. I took 3 years of Art Classes in High School. The most I learned about "fundamentals of art" were simply about "which colors accentuate and clash against each other and why".

But, I have that same "baseline" from music as well. Which instruments tend to produce which emotional reactions.

I've met artists, even on these forums, who have been college educated on the subject and who still make artwork that clashes with the inherent design/direction of their games and what they want the players to be feeling at the time.

I think it's less "something you need to learn" and more "something you need to feel".

Let me use the recently remade "Dead Space" as an example.

The original game evoked a sense of dread in me everywhere I went, everything I did. Even in the brightly lit places with lots of greenery. Much of it was very "minimalist". Detail only existed on things that mattered, and when it didn't matter, it was allowed to "fade into the background". This meant, you could tune out a lot of the nonsense and focus on the "feeling of dread". The game was less about "let's shove you in the dark all the time to scare you" and more about "You aren't safe anywhere and the color palette tells you that instantly and immediately. Much of the game had this strange disconcerting "cleanliness" about it. Like you were walking in an abandoned and sterile environment... where it was OBVIOUS very bad things went down... but there wasn't a lot of evidence for it. Maybe a blood splatter here or there... some debris on the ground here or there... but not much else. It evoked a sense of "I have no idea where the threats are coming from".

The new one has some serious problems. The only "evokation of dread" I have in the new one is "am I low on ammo?". Every location where a threat looms is obvious. The area is splashed in red, or bodies, or growths. The lights will flicker constantly or go out altogether. The areas where fights will happen tend to be filled with the most clutter and detail now. Nothing looks sterile or clinical anymore. It all looks like a hurricane blasted through the ship, which removes the existential threat of "when will I be attacked?". The answer is usually "once per room, and maybe a 50% chance every hallway". The game uses its art assets to communicate to you that "you are going to be attacked constantly", which removes the existential dread of "when am I going to be attacked?". The remake is very "busy" in terms of detail and art assets. This "busyness" tends to overshadow and get in the way of any feeling of "dread" you may have had. It plunges you into darkness frequently which is either annoying since there's nothing to fight... or predictable because darkness is usually a cue to throw enemies at you.

I mean, just the change of art direction and color palette removed the ever present feeling of "dread" from the remake that the original has. I'm not an art person, and I knew that just by FEEL of the game. Dead Space original felt very oppressive and minimalist and empty and lonely and depressing at all times. Dead Space remake feels very much like just traversing from A to B and checking things off the list, with a couple of more effective "jump scares" than the original. It's a game that feels like everyone took everything on the ship and just threw it into the middle of the floor, whether it made sense to have it there or not. Blood Graffiti everywhere it's possible to have it. Corpses everywhere it's possible to have them. Random goop and gunk everywhere to the point of desensitizing the player. Chairs everywhere overturned. Tables too. Or anything on those surfaces, strewn about the floor. Everything is now "dark and gritty". Made to look "dirty". Which doesn't evoke the appropriate feelings. Instead it evokes "this is all normal, ignore it".

If a place looks safe and like the monster hasn't been there. Then it ramps up the dread when you know it COULD be there at some point. That it CAN get you there. But, if the place looks like a war zone and the monsters have been through there a dozen times... then you're on guard, expect to be attacked, and it moves from "anxiety" into "just patiently waiting for the inevitable attack". All from the art style.

It is also possible to take "warm colors" and make them "terrifying colors". One, again, need only look at Dead Space original. Oranges, yellows, and reds, in hues that would normally evoke "sunrise" or "sunset" are suddenly very sinister in certain places of that game just because of what those hues are attached to. They can induce panic since they're strapped to some of the enemies that explode. They can induce caution and dread if they're meant to signify that you're in a room with a very dangerous thing that can kill you if it hits you (like a moving piece of machinery).

The color, itself, can have its meaning changed via context.

But, if you just have the color as "ambience", then it'll only ever evoke the baseline "ambient" feeling of zero association with anything.

Then you get into weird things like, "Oh, yellow evokes feelings of hunger, that's why a lot of restaurants use yellow". Really? The color of a restaurant or even the packaging has never "made me hungry". Does that "warm color" really evoke "hungry feelings"? I'd likely argue "no". But, art schools might tell you otherwise.

Does blue really evoke "feelings of depression"? I'd argue "not unless it's nearly grey blue and not visually striking at all... but you can achieve the same effect with browns... or really any color if it's overused."

You "could" learn art, but it's just really not all that necessary unless a person has issues with pinning down how their own emotions work, is somewhat color blind, or intends to do their own artwork. There's not really a reason to learn anything beyond "what colors go together and which ones clash" in terms of game design. The rest is just learning player psychology.

 

[gstv87]

what you actually need and what you actually (expect to) get out of it ...

money.

if that's the question, the answer invariably gravitates towards *money*
and when you have *money* locked behind the requirement of *work* by a society that's been raised under the premise of *work for money*, and suddenly along comes a tool that requires no *work* for the purposes of providing *result* to paying *client* who provides *money*, then you have a problem.

the real problem is much deeper than simply "this is/isn't art", and somehow people keep avoiding the matter because it would show them how flawed the rest of their views is, across the board.
if you know where to look, you'll address the root of the problem directly (you know, as *real* intelligence would allow?) and submit "Well, then don't pay for the piece, pay for the show of artistry behind it."
using AI art will only out the users who try to sell it as genuine art pieces, as frauds.
when you require something tailor-made for something engineered in advance like a game, AI art just doesn't have the finesse required.

 

[123edc]

and suddenly along comes a tool that requires no *work* for the purposes of providing *result* to paying *client* who provides *money*, then you have a problem.

ther's a old story about a small town,
they wanted to build a big railway station within it ... it could have made the ppl realy wealth ... but they protestet against it, becouse "the devil lives within it ... just look, how these things smoke" ... well ... that's ancient history

but if we actually go by youre argumentation wit the "problem",
then i have to ask ... wouldn't we have to argue the exact same way for ...

the cashier in the supermarket [it already get's automated]
the person refilling the good [it already get's automated]
the truck driver [china already has experimental public roads]
heck, even tooth implants [no joke, - though still under expert control - already got made via robot already]

i don't even start to talk about all these callcenter and office worksers - who's [repetetive] jobs will break away ...


so, yes, we have two options ...
we can say "we have a problem" ... and protest against change
or we can be happy about not having to take all these mundane repetitive steps any more ... and enjoy our plus in free-time ...

using AI art will only out the users who try to sell it as genuine art pieces, as frauds.

i know, we start walking in circles but i can just repeat myself

in the end it's all about the question, what you actually need

and ppl working on rpg maker-games ...
especially the hobbyist faction out there,

usually don't want to sell "genuion pieces of modern art" ...
they want to create a game, that others can enjoy ...

 

[gstv87]

we can say "we have a problem" ... and protest against change
or we can be happy about not having to take all these mundane repetitive steps any more ... and enjoy our plus in free-time ...

correct.
now, who's gonna provide for the free time of that people? you?
because, all we're having is less work, with no pay, and the same taxes.

you never understood my exposition: *the problem* is the people who take AI art and sell it as genuine pieces, for profit.
and *adding* to the problem, is the people who pay for it, without question, in a context where they're told that money equals value.
it doesn't.

SO. in that situation, the ones with a genuine reason to oppose AI art, are the artists who make digital pieces to be sold for profit. Specifically.
not the artists who make assets for games, and not the artists who make other types of pieces from different techniques.
and as long as people keep buying those pieces without question, I say let them pay, and let them add to the bubble bursting out in the end.
eventually one side of the deal will realize it's unsustainable.
meanwhile, support the artists for their skill, not their product.... because if you ask a product from their skill, they WILL deliver.
whereas if you ask these AI users for just a product, they'll give you whatever the AI comes up with.
and when you ask for refinement, they'll walk away.


EDIT 20 minutes after the fact.
I was just browsing my mail, and found a news article about this person:

Jos Avery Portraits (@averyseasonart) • Instagram photos and videos

50K Followers, 1,716 Following, 222 Posts - See Instagram photos and videos from Jos Avery Portraits (@averyseasonart)

www.instagram.com

they just confessed their work is AI generated.

Viral Instagram photographer has a confession: His photos are AI-generated

Artist wants to "come clean" and highlight a new media process.

arstechnica.com

I don't know who this person is, I don't know how good their work is... they're just someone who illustrates my explanation.
go in there, and look at the amount of praise from people who are totally clueless, and consider the fact that instagram is a monetized platform, and tell me there is not someone out there profiting from this without being called out.

 

[123edc]

meanwhile, support the artists for their skill, not their product....

meanwhile, support the taxi driver for his skill ...
not for bringing you safely from point A to point B ...

same argument ...
and again - i know, i repeat myself -
in the end it is a question of your individual goals ...
of what you actually want ...

and i do want to get the best possible visualization,
for the ideas within my head ...
with the budget i can afford to spend for it


which means i effectively have three choices:
1] buy bulk - dlc, that every other game out there uses en-masse,

getting only a single static bust per character,
without even having the option, to actually change the characters clothing [for different time periods, sttings, locations] ...

2] spend hours upon hours learning to actually draw [in a way, that others would actually want to look at] ...

all the while not beeing able to work on the thing,
that i enjoy, that actually made me want to start creating a game
[the story and mechanics]

3] abuse my budget for the next couple of years and go into unpaid dept,
to let someone else draw for me

or 4]
feed the ai with the input, about the vision within my head ...
let it create hundreds of pieces,

play around with the ai and inputs, to get the picture the way i want ...
and then pick out the one, best fitting towards my idea ...

all the while beeing able, to get my character with different clothings, different hairstyles, different locations, different poses and actions and even some useable cg's


i don't want to "support" my taxi driver for his skill,
i want to safely arrive at my destination ...

i don't want to "support" my artist for his skill,
i want to get the idea i have within my head visualized within my game ...

different priorities [and possibilitys, becouse even if i could afford a taxi driver, i would still use the subway ]

whereas if you ask these AI users for just a product, they'll give you whatever the AI comes up with.

if i buy (!) ai artwork from someone else - yes,

but if i commission a piece from an artist,
don't i also just get, what these artist comes up with?
don't i also just get, what the artist thinks within his head, while reading my description?

that's not so different, if you ask me!


and what's the objective [!] difference
[by which i mean for: if i use picture A in my game, the player would have a better experience while looking at it, then he would have with picture b]

between bulk - sold "dlc" picture A [human drawn] and bulk - sold "dlc" picture B [AI - generated] ... if i see both pieces before buying ... if i can actively choose ... which one fits my vision better?

again, i don't see any objective difference for the player ...


besides, when thinking about ai,
then i think mainly about using the options, it opens up for myself

i think mainly about generating pieces and then actively picking out those, that best describe the actual scene i want to portray ... those, that best fit the vision i have within my head ...

and yes, if i can realize my exact vision with 0$ cost ...
then i won't need to spend 1000$ to get someone else realizing it close to my vision ...

and tell me there is not someone out there profiting from this

and *adding* to the problem, is the people who pay for it, without question, in a context where they're told that money equals value.

i don't say, that there's nobody profiting from this,
but "value" is something highly subjective,

what "value" does a piece of paper with some ink on it have?
what "value" does a electronic file consisting of 0's and 1's have?

there are pieces of art, worth millions upon millions ...
meanwhile the exact same image ... on the exact same type of paper ... with the exact same type of print ... is worth a mere 50 cent ...

the objective "value" of both of these is the image, your eye sees ...
and the worth of the materials used within that piece ...

yes, ppl do subjectively assign value to things,
and a thing is only worth, what ppl are willing to pay ...


and if someone can give the ppl the things, they want to see ...
why shouldn't it be "rewarded" for it? that's how it has always been!

correct.
now, who's gonna provide for the free time of that people? you?
because, all we're having is less work, with no pay, and the same taxes.

yes, certain (historically grown) systems will have to change,
to accomodate the changes, that technology will inevitably mean for all of our lives!

and if you actually look at the statistics
[avarage hours worked per day/month/year, number of part-timers]
you can see a clear trend,

technology and ai will massively pull on these already existing trend [especially for all of the white collar jobs]

but then again, ...
what objective [!] reason would there be,

to produce something with 40 hours of human labor ...
when i can get the exact same thing with 10 hours of human labor?

that - in and on itself - realy shouldn't be something, to be afraid of ... but rather, a thing, to be happy about ...
__________
oh, and to stay with the taxi - driver example,

in art, you have to learn the "fundamentals", to draw a picture,
you have to learn "driving a car" to be allowed to drive,
[in this case - you even have to pay for youre licence]

you have to practice "drawing", to create a piece, that someone else would buy,
you have to practice "driving", to get your taxi licence,

it takes a certain ammount of time, to create a picture for someone else
[according to his specifications]
it takes a certain ammount of time, to deliver someone else to his destination,

there exists cheaper [but a little less comfortable] alternatives ...
like the pre made dlc's ... or the bus and subway

there still needs to be someone, creating these cheaper alternatives
[you still need an artist to make dlc's and a bus or subway driver]

and AI threatens, to make both of these jobs obsolete
[self driving cars / ai-generator]

so, if my self driving car costs me 3$ to the airport,
and my taxi would cost me 80$,

then the taxi would have to offer me some kind of "service", that would justify me spending 77 additional dollars on it

and if my ai-art generator costs me 3$ for the piece i want
and my artist would cost me 80$,

yes ... he would have to offer something, that would justify these additionale expense ... simple as that

 

[Byzantiumn]

AI art is the world's biggest art theft that everyone publicly use for free with no repercussions to create their art. I'm not gonna deny that I haven't used it, but I have because it was interesting at first but I never used it for a game or in developing a game. Knowing the truth behind AI art is just absolutely wrong.

Imagine making a game that you spent days, months or years creating. Only for someone to completely steal it without your permission, publicly distribute it for free and make it so that anyone can change it freely.

That's what AI Art is.

Anything that you generate with AI art are stolen art from DeviantArt, ArtStation, Pixiv, real life people, dead or alive. Every single AI art has no soul, far from being called an actual art. This is the biggest art theft and everyone simply accepts it because it creates "art" in a matter of seconds.

What AI art is doing is harmful, wrong and is straight up just a disrespect for any artists that are drawing tirelessly just for a random person to feed it to their AI without the artist's consent. Another fact is that at the death of Kim Jung Gi, a few days after his passing not even his memorial, AI programmers took his art and used it to make an AI Art generator of his art style. This is the AI "art" community.

Proof:


It is ethically wrong, unprofessional and is straight up a violation of anyone's copyright rights.

Never use AI art. Never advertise AI art. If you do, all you're doing is promoting the use of art theft. I'd rather see people create something whether its objectively good or bad rather than relying on an AI that stole creativity from everyone.

If you respect the art of game development in any way, do not use AI Art.

I found myself out of work this week after one year in the professional world after 8 years of school (bachelors and masters), I was laid off directly because of AI, the CEO (direct boss) said it outright to me, AI did my job so quick that the fact it did it worse did not really matter. If we support AI in any way now, none of us will have a future. I am living proof, DO NOT SUPPORT AI.

meanwhile, support the taxi driver for his skill ...
not for bringing you safely from point A to point B ...

same argument ...
and again - i know, i repeat myself -
in the end it is a question of your individual goals ...
of what you actually want ...

and i do want to get the best possible visualization,
for the ideas within my head ...
with the budget i can afford to spend for it


which means i effectively have three choices:
1] buy bulk - dlc, that every other game out there uses en-masse,

getting only a single static bust per character,
without even having the option, to actually change the characters clothing [for different time periods, sttings, locations] ...

2] spend hours upon hours learning to actually draw [in a way, that others would actually want to look at] ...

all the while not beeing able to work on the thing,
that i enjoy, that actually made me want to start creating a game
[the story and mechanics]

3] abuse my budget for the next couple of years and go into unpaid dept,
to let someone else draw for me

or 4]
feed the ai with the input, about the vision within my head ...
let it create hundreds of pieces,

play around with the ai and inputs, to get the picture the way i want ...
and then pick out the one, best fitting towards my idea ...

all the while beeing able, to get my character with different clothings, different hairstyles, different locations, different poses and actions and even some useable cg's


i don't want to "support" my taxi driver for his skill,
i want to safely arrive at my destination ...

i don't want to "support" my artist for his skill,
i want to get the idea i have within my head visualized within my game ...

different priorities [and possibilitys, becouse even if i could afford a taxi driver, i would still use the subway ]


if i buy (!) ai artwork from someone else - yes,

but if i commission a piece from an artist,
don't i also just get, what these artist comes up with?
don't i also just get, what the artist thinks within his head, while reading my description?

that's not so different, if you ask me!


and what's the objective [!] difference
[by which i mean for: if i use picture A in my game, the player would have a better experience while looking at it, then he would have with picture b]

between bulk - sold "dlc" picture A [human drawn] and bulk - sold "dlc" picture B [AI - generated] ... if i see both pieces before buying ... if i can actively choose ... which one fits my vision better?

again, i don't see any objective difference for the player ...


besides, when thinking about ai,
then i think mainly about using the options, it opens up for myself

i think mainly about generating pieces and then actively picking out those, that best describe the actual scene i want to portray ... those, that best fit the vision i have within my head ...

and yes, if i can realize my exact vision with 0$ cost ...
then i won't need to spend 1000$ to get someone else realizing it close to my vision ...



i don't say, that there's nobody profiting from this,
but "value" is something highly subjective,

what "value" does a piece of paper with some ink on it have?
what "value" does a electronic file consisting of 0's and 1's have?

there are pieces of art, worth millions upon millions ...
meanwhile the exact same image ... on the exact same type of paper ... with the exact same type of print ... is worth a mere 50 cent ...

the objective "value" of both of these is the image, your eye sees ...
and the worth of the materials used within that piece ...

yes, ppl do subjectively assign value to things,
and a thing is only worth, what ppl are willing to pay ...


and if someone can give the ppl the things, they want to see ...
why shouldn't it be "rewarded" for it? that's how it has always been!





yes, certain (historically grown) systems will have to change,
to accomodate the changes, that technology will inevitably mean for all of our lives!

and if you actually look at the statistics
[avarage hours worked per day/month/year, number of part-timers]
you can see a clear trend,

technology and ai will massively pull on these already existing trend [especially for all of the white collar jobs]

but then again, ...
what objective [!] reason would there be,

to produce something with 40 hours of human labor ...
when i can get the exact same thing with 10 hours of human labor?

that - in and on itself - realy shouldn't be something, to be afraid of ... but rather, a thing, to be happy about ...

The purpose of life is to live and all things therein are in relation to that. Seek to live more fully, not less, let not machines take your burdens but learn the value of yours.

 

[123edc]

The purpose of life is to live and all things therein are in relation to that. Seek to live more fully, not less, let not machines take your burdens but learn the value of yours.

so, youre working 16 hours a day for the pay of a dry bread and a bed in the hey?

 

[Byzantiumn]

so, youre working 16 hours a day for the pay of a dry bread and a bed in the hey?

Pretty much actually besides the internet access. I spent the first 14 years of my life living homeless with my family so yes, I understand poverty you little piece of trash who thought to instigate this.

 

[gstv87]

feed the ai with the input, about the vision within my head ...
let it create hundreds of pieces

please, do present to me the AI that will generate isometric-styled tiles of a specific material, with the pixel I require where it needs to be in order to tile correctly, at a specific camera angle that is not the usual, and I'll gladly use it.
or, the AI that will generate isometric characters with the outfit I specify in order to be recognizable, with the set of animations that I specify, in the output size and framerate that I specify.

if what you want is character busts, you'll get character busts... but character busts alone don't make a game.

 

[123edc]

Pretty much actually besides the internet access. I spent the first 14 years of my life living homeless with my family so yes, I understand poverty you little piece of trash who thought to instigate this.

first of all, we can discuss about everything here,
but there realy is no need to throw around with insults here!

in that case, let me rephrase the statement above:
if you would have the option, to work for 8 hours a day ...

would he still joyfully work your 16 each and every day,
if you wouldn't have to?

please, do present to me the AI that will

we're running in circles here, becouse we're - yet again - back to square 1 of our little discussion ... so i repeat myself again:

in the end, it's a question about individual goals ...

there are things, the ai can already do well,
there are things, the ai struggles with and
there are things, the ai at it's current state can't do,

technology evolves over time - that's, how it has always been!

and yes, you can't use it, if it can't produce the vision you have (yet) ...
but that is no argument, for NOT using it, if it CAN produce the vision you have ... is it?

also, we were talking about pictures here ...
not about tilesets and sprites ...

if what you want is character busts, you'll get character busts... but character busts alone don't make a game.

no, they don't

but character busts with different expresions, anime styled character pictures with background ... CG's take away a huge junk of the workforce needed [and with that the time and cost] ... especially for story heavy games ...

and yes, you still need your tilsets and sprites
[though, at least for the latter, there are already attempts to actually make it work ... but that's realy still in it's child shoes]

 

[Tamina]

money.

I always believe the current AI business model isn't sustainable from business point of view. The art industry which is the data source provider
really should get a cut from this new AI market, and subscription fee for AI service should increase for commercial use so AI service developers start making more money to develop better tools for AI users.

My ideal model whould be something more like Shutterstock. Everyone involved in AI service gets a cut. Users are the one paying. In return they get higher quality service for their end product.

This will make AI generated products more acceptable for the end product customers as well, as ethical concerns are removed if artists are getting paid. which increases the profit for game dev using AI in their pipeline.

The current model will likely lead to art industry start using hidden watermark technology to protect their work from being used as data. Then this AI industry won't grow without new data.

But copyright law exists to generate more profit for all parties involved....so I am pretty sure this new industry will eventually find a way that everyone gets their money and be happy.

the real problem is much deeper than simply "this is/isn't art", and somehow people keep avoiding the matter because it would show them how flawed the rest of their views is, across the board.

There is no shame in discussing money and business ....we all work for money and some of us sell games or game dev services...it's part of our life.

 

[123edc]

as ethical concerns are removed

in my place there already is a law in place,
regarding text and data mining

... when the image's data itself has the machine-readable code/massage for it, then it is not allowed to be used for it

that in and on itself already replace all ethical concerns in my opinion,
as anyone, who doesn't want this, can opt-out from it during the uploead of his picture


... yes, you can think about ppl opting out and private platforms, where these opted outs post their stuff for a fee ... but that's not realy something, that needs state - regulation anymore!

 

[gstv87]

but that is no argument, for NOT using it, if it CAN produce the vision you have ... is it?

right.
but, like I said earlier: the only ones with allowance to say anything about AI, are the artists who make artwork for a living.
and, specifically, digital artwork, sold online... canvas? don't count, it's not AI generated.
pottery? carpentry? soldering? none of that counts.... it's still artistry, but it's not in the argument.
neither is game-making, as IT *doesn't require the result of the use of AI art* as it's fundamental resource.

OFF that argument, the other one: the so-called "AI art" isn't technically "art", it's a mish-mash of brush strokes that somewhat resemble a recognizable piece, directed by a script that tried to copy other works of art with actual technique and purpose behind THEIR mish-mash of brush strokes.... at best. And, at worst, a blatant copy-paste of actual pieces of artwork uploaded to the cloud.
and off THAT, the argument about money.

you want to not go in circles? go back to the root of it: "if you're a game maker, don't use AI art".
to which I say:
one, AI art is not art... it doesn't have a technique or a purpose.
and two, game-making doesn't require AI art to exist, and no AI is refined enough to provide the assets required by the ocean of an industry that is game-making.

hidden watermark technology to protect their work from being used as data

Not Finding That hard to believe at all.
ahem.
nudge-nudge, wink-wink.

 

[123edc]

but, like I said earlier: the only ones with allowance to say anything about AI, are the artists who make artwork for a living.

yeah, and the only ones with allowance to say anything about self driving cars are the taxi, bus and truck drivers, who drive cars for a living ...

that's a somewhat stupid argument regarding technological advancements!

OFF that argument, the other one: the so-called "AI art" isn't technically "art", it's a mish-mash of brush strokes that somewhat resemble a recognizable piece, directed by a script that tried to copy other works of art with actual technique and purpose behind THEIR mish-mash of brush strokes.... at best. And, at worst, a blatant copy-paste of actual pieces of artwork uploaded to the cloud.
and off THAT, the argument about money.

yeah, and the so called "self driving car" isn't technically "driving", it's a mish-mash of collected data that's made to resemble a certain type of behavior, directed by a script that tried to copy the behavior of other drivers with actual driving licence and purpose behind their ,,,, you get the argument?

and two, game-making doesn't require AI art to exist,

it doesn't "require" it ...

but if my only options are between having a single static picture without clothing options ... and having different types of hairstyles and clothing ... as well as backgrounds and cg's for my character

why wouldn't i use the latter, if it's available to me?


you don't need a self driving car, to travel from point A to B ...
but why wouldn't you use it, if it costs you only 3$ instead of the 80 you'd have to pay for your taxi driver?


and the "purpose" of art within a game,
is to vizualize the idea/story you want to tell

the artist (or ai) doesn't need a "purpose" ...
it needs to deliver me the object i want, to reach my goal ...

 

[gstv87]

yeah, and the so called "self driving car" isn't technically "driving", it's a mish-mash of collected data that's made to resemble a certain type of behavior, directed by a script that tried to copy the behavior of other drivers with actual driving licence and purpose behind their ,,,, you get the argument?

uhm... no.
it's not a mish-mash of techniques, it's the constant evaluation of the immediate environment, on the fly, based on mathematical models.
you can't drive "abstractly" or "impressionistly", or "expressivistly"... you drive within the laws of the road!
one is a strict model, the other is subjective... it's two different things.
and, more to the point, if we were to put both arguments at the same level, there would need to be an AI that can generate complex models on the fly, to compare with what is usually called a real-time system, like that of self-driving cars!
there isn't one.
so, stop mixing apples and oranges.

the artist (or ai) doesn't need a "purpose" ...

no.
but what you yourself are doing, could be considered art in and of itself.
so, right back at you.

 

[123edc]

no.
but what you yourself are doing, could be considered art in and of itself.
so, right back at you.

it could be, ... maybe

but that doesn't change the fact,
that i - as someone, who creates the game - doesn't need the "purpose" behind a piece ... i need the "quality" and "availability" of the product, i want to use

neither the commissioned artist [who draws according to my specifications],
nor the creator of bulk dlc content [who draws, what he beliefes, that will sell]
nor the ai-generator [who draws according to my specifications]

need a "purpose" ...
yes, they may have a purpose ... they may have certain thoughts [which essentially are biochemical / electrical impulses] as to why they do something this and that way ...

but in the end of the day, what i need, to fullfill my goal,
it is the product ... not some imaginary, highly subjective, "purpose" interpretet into it ...

 

[gstv87]

i - as someone, who creates the game - doesn't need the "purpose" behind a piece

. . . . . ?

then why do you order a piece from an AI that exists solely for the purpose of creating it?

artists create art to consolidate ideas.
that's THEIR reason to be, which they materialize in the form of a piece, of which the purpose is to hold that reason and idea beyond the lifespan of the artist (hopefully)
AIs don't have that... their sole reason to be IS to make those pieces.
the pieces themselves are the purpose!

when you order a custom drawing, the one time that the AI presents a selection of drawings before you, that's the only one of a kind there's ever going to be of each one!
if you don't pick one and re-shuffle the lot, all those pieces are gone forever.
results are inconsequential to the existence of the AI!

i need the "quality" and "availability" of the product, i want to use

that can be interpreted in two ways:
-you need the complexity and readiness of the system at a moment's notice, for a task that involves complex input.
-or, you want large colorful pictures, for free.

if you want complex solutions to complex inputs, AI can get you only half way there, as no AI system is complex enough.
even self-driving cars crash from time to time, and airline pilots take manual control of planes during takeoff and landing, because no autopilot can possibly know about anything beyond what their own sensors can read... and we're talking about THE most complex autonomous systems ever designed.
if you want pictures for free, then the matter becomes about money.
I just pointed that out.