What AI could do for MMOS

OG_Solareus

Now that I'm completely geeking on AI technologies... Here is example of what I think AI can be used for which is creature design. How many dragons or waredogs can one fight over the past 20 years. AI can help with Gen Z lack of imagination.. here is a few monster I pulled from the void..

https://twitter.com/ai_d__g/status/1700004764887535886

https://twitter.com/ai_d__g/status/1699980656221082018

https://twitter.com/ai_d__g/status/1699980247905534378

These are beast I would love to fight in an mmo. The imagination comes from actual language now, and it can be processed with other (paid) ai tools to be come 3D.  Ai is the biggest step into next generation development and it feels like not a single mmo developer is talking about it.. as I can see. Maybe one is but it is not very vocal..

KnightFalz

I don't know if I'd want to fight them. All I have to go on is their appearance. The use of AI in that aspect is literally skin deep.

What does that use of AI really change in MMORPGs that will elevate them to a new level?

OG_Solareus

KnightFalz said:

I don't know if I'd want to fight them. All I have to go on is their appearance. The use of AI in that aspect is literally skin deep.

What does that use of AI really change in MMORPGs that will elevate them to a new level?

production speed. Everything from concept to working model could be done in a few days, Adding new models to the game would be a lot easier when the pipe line is implented.

How many developer videos do we see some do tweaking a model , and working on that model for over year.  AI can do it faster and  in the near future with less mistakes..

It would also cut costs in the number of people needed to "create concepts". Cutting those costs I think would help greatly in other aspects of mmo development like various content systems.

I don't know if anyone ha snoticed, but the unionization is in key areas that AI could fill.

Scot

You changed your avatar, it's great so stick with it. (I always say that because I use avatars to zero in on posts and get thrown if people change them. )

I don't see AI doing much for MMOs in the short term, they will simply be used by studios as a cost cutting measure, like a gaming engine you buy of the rack. But long term they could start to make decisions in game, that will be a whole new ball game.

Vrika

OG_Solareus said:

KnightFalz said:

I don't know if I'd want to fight them. All I have to go on is their appearance. The use of AI in that aspect is literally skin deep.

What does that use of AI really change in MMORPGs that will elevate them to a new level?

production speed. Everything from concept to working model could be done in a few days, Adding new models to the game would be a lot easier when the pipe line is implented.

How many developer videos do we see some do tweaking a model , and working on that model for over year.  AI can do it faster and  in the near future with less mistakes..

It would also cut costs in the number of people needed to "create concepts". Cutting those costs I think would help greatly in other aspects of mmo development like various content systems.

I don't know if anyone ha snoticed, but the unionization is in key areas that AI could fill.

Currently AIs can not create working models. AI can create great static images, even on 3D, but it can barely create movement correctly for a human. For a monster like that, AI would get it totally wrong.

Perhaps in a decade we'll have AIs that understands what body's parts are and how they should move well enough.

Cactus_LFRez

AI ultimately is a tool that iterates based on inputs.
And so far it looks like that is limited to images and text.

If you use it carefully, a not so good artist could take examples of their not so good art and AI could clean it up.

But AI so far lacks the purpose and deliberate choices of human made art, it literally just amalgamates stuff.  It all looks very fake.

I would only use AI art as a way for a non artist to make concepts, like a mood board, and then pass it off to a real artist to make real concept art.  But again AI cannot create anything truly unique so it will only create a generic amalgamation.  So if you truely have a unique and specific idea AI probably won't help too much.

I would not try to use it to make game assets, I think people will be able to tell it is AI generated.

Angrakhan

Yeah near term you will see AI used to create art and sound assets for games. This will not be limited to MMOs. As mentioned it will be a cost cutter and time saver because what can be done by a whole art team in a month will take an AI a few minutes. From an artistic standpoint it will be copy/ paste / iterate crap like you see above, but it will be good enough to make a game out of because who really sits there and ponders the artistic virtues of the rat or skeleton or goblin you just killed for the umpteenth time? AI will be used similarly for sound tracks. At the end of the day what it's going to do is cut jobs. This is true across every industry, but for sure it's happening right now to artists. Why hire a human when the AI can work 24 x 7 x 365, never needs breaks or gets sick or wants a raise? So what if the quality of the work is less, the volume is much higher, and most importantly the cost is much cheaper. The art department will go from a team to a single director of art whose primary job is to babysit the AI and evaluate what it's producing to decide what is worth putting in the game and what is junk. This person won't need any actual artistic skills themselves. The actual artists will be heading to Olive Garden to apply for wait staff positions at least until robots driven by AI replace those jobs as well.

At the end of the day I'm not sure we'll like the consequences of moving to AI, but by then it will be too late.

Quizzical

You would love to fight against non-animated concept art?  I usually prefer to fight against mobs that are animated.

KnightFalz

OG_Solareus said:

KnightFalz said:

I don't know if I'd want to fight them. All I have to go on is their appearance. The use of AI in that aspect is literally skin deep.

What does that use of AI really change in MMORPGs that will elevate them to a new level?

production speed. Everything from concept to working model could be done in a few days, Adding new models to the game would be a lot easier when the pipe line is implented.

How many developer videos do we see some do tweaking a model , and working on that model for over year.  AI can do it faster and  in the near future with less mistakes..

It would also cut costs in the number of people needed to "create concepts". Cutting those costs I think would help greatly in other aspects of mmo development like various content systems.

I don't know if anyone ha snoticed, but the unionization is in key areas that AI could fill.

I don't care about production speed or cost. What I care about is the nature of that produced and how it transcends that which came before. Being faster and cheaper to make alone doesn't elevate MMORPGs to a new standard. It just allows more to be pumped out at the current standard.

KnightFalz

Vrika said:

OG_Solareus said:

KnightFalz said:

I don't know if I'd want to fight them. All I have to go on is their appearance. The use of AI in that aspect is literally skin deep.

What does that use of AI really change in MMORPGs that will elevate them to a new level?

production speed. Everything from concept to working model could be done in a few days, Adding new models to the game would be a lot easier when the pipe line is implented.

How many developer videos do we see some do tweaking a model , and working on that model for over year.  AI can do it faster and  in the near future with less mistakes..

It would also cut costs in the number of people needed to "create concepts". Cutting those costs I think would help greatly in other aspects of mmo development like various content systems.

I don't know if anyone ha snoticed, but the unionization is in key areas that AI could fill.

Currently AIs can not create working models. AI can create great static images, even on 3D, but it can barely create movement correctly for a human. For a monster like that, AI would get it totally wrong.

Perhaps in a decade we'll have AIs that understands what body's parts are and how they should move well enough.

If that's all AI can do in the next decade it is nowhere close to being transformative of MMORPGs in a meaningful way.

Vrika

KnightFalz said:

OG_Solareus said:

KnightFalz said:

I don't know if I'd want to fight them. All I have to go on is their appearance. The use of AI in that aspect is literally skin deep.

What does that use of AI really change in MMORPGs that will elevate them to a new level?

production speed. Everything from concept to working model could be done in a few days, Adding new models to the game would be a lot easier when the pipe line is implented.

How many developer videos do we see some do tweaking a model , and working on that model for over year.  AI can do it faster and  in the near future with less mistakes..

It would also cut costs in the number of people needed to "create concepts". Cutting those costs I think would help greatly in other aspects of mmo development like various content systems.

I don't know if anyone ha snoticed, but the unionization is in key areas that AI could fill.

I don't care about production speed or cost. What I care about is the nature of that produced and how it transcends that which came before. Being faster and cheaper to make alone doesn't elevate MMORPGs to a new standard. It just allows more to be pumped out at the current standard.

The devs don't really care whether you care. If they can make 80% of the game cheaply to match current standard then that's what they set out to do, and then they'll spend their time and focus to improve the 20%.

There are maybe 2 studios in the world who have big enough project budget to improve everything from what's been created before, Rockstar and Cloud Imperium Games. The rest of them choose some areas to focus on and elsewhere do intentionally only average work to stay on budget.

Cactus_LFRez

Vrika said:

KnightFalz said:

OG_Solareus said:

KnightFalz said:

I don't know if I'd want to fight them. All I have to go on is their appearance. The use of AI in that aspect is literally skin deep.

What does that use of AI really change in MMORPGs that will elevate them to a new level?

production speed. Everything from concept to working model could be done in a few days, Adding new models to the game would be a lot easier when the pipe line is implented.

How many developer videos do we see some do tweaking a model , and working on that model for over year.  AI can do it faster and  in the near future with less mistakes..

It would also cut costs in the number of people needed to "create concepts". Cutting those costs I think would help greatly in other aspects of mmo development like various content systems.

I don't know if anyone ha snoticed, but the unionization is in key areas that AI could fill.

I don't care about production speed or cost. What I care about is the nature of that produced and how it transcends that which came before. Being faster and cheaper to make alone doesn't elevate MMORPGs to a new standard. It just allows more to be pumped out at the current standard.

The devs don't really care whether you care. If they can make 80% of the game cheaply to match current standard then that's what they set out to do, and then they'll spend their time and focus to improve the 20%.

There are maybe 2 studios in the world who have big enough project budget to improve everything from what's been created before, Rockstar and Cloud Imperium Games. The rest of them choose some areas to focus on and elsewhere do intentionally only average work to stay on budget.

I am not sure how much time AI could really save,

As far as I am aware, it can't make 3d models, it can't make texture maps either.
AI can only really make 2d images and it can only amalgamate preexisting images often with nonsensical details.
It can't for instance rotate a 2d image to make a template for a 3d artist to work from, or even generate the same character twice in a different pose.

Like this


Maybe it could advance more to do some of these things but it has a long way to go and I am not sure how useful it really is right now.  You would probably spend more time fixing the things AI creates, you might as well do it yourself.

Right now I see AI being used mostly to create Steam trash.

Sensai

Vrika said:

KnightFalz said:

OG_Solareus said:

KnightFalz said:

I don't know if I'd want to fight them. All I have to go on is their appearance. The use of AI in that aspect is literally skin deep.

What does that use of AI really change in MMORPGs that will elevate them to a new level?

production speed. Everything from concept to working model could be done in a few days, Adding new models to the game would be a lot easier when the pipe line is implented.

How many developer videos do we see some do tweaking a model , and working on that model for over year.  AI can do it faster and  in the near future with less mistakes..

It would also cut costs in the number of people needed to "create concepts". Cutting those costs I think would help greatly in other aspects of mmo development like various content systems.

I don't know if anyone ha snoticed, but the unionization is in key areas that AI could fill.

I don't care about production speed or cost. What I care about is the nature of that produced and how it transcends that which came before. Being faster and cheaper to make alone doesn't elevate MMORPGs to a new standard. It just allows more to be pumped out at the current standard.

The devs don't really care whether you care. If they can make 80% of the game cheaply to match current standard then that's what they set out to do, and then they'll spend their time and focus to improve the 20%.

There are maybe 2 studios in the world who have big enough project budget to improve everything from what's been created before, Rockstar and Cloud Imperium Games. The rest of them choose some areas to focus on and elsewhere do intentionally only average work to stay on budget.

"If they can make 80% of the game cheaply to match current standard then that's what they set out to do, and then they'll spend their time and focus to improve the 20%."

Do you honestly believe that? I think it's beyond naive to believe that even with cost saving tools that they would spend any additional time or money on anything beyond the cash shop/revenue milking stream, they will just enjoy the boost to net income. 

As for the topic, the only real improvement would be using AI as dungeon master to alter mob densities and locations based on player actions, to "craft" dungeon experiences based on class/group makeup and to activate "random" dynamic world events. 

OG_Solareus

Vrika said:

OG_Solareus said:

KnightFalz said:

I don't know if I'd want to fight them. All I have to go on is their appearance. The use of AI in that aspect is literally skin deep.

What does that use of AI really change in MMORPGs that will elevate them to a new level?

production speed. Everything from concept to working model could be done in a few days, Adding new models to the game would be a lot easier when the pipe line is implented.

How many developer videos do we see some do tweaking a model , and working on that model for over year.  AI can do it faster and  in the near future with less mistakes..

It would also cut costs in the number of people needed to "create concepts". Cutting those costs I think would help greatly in other aspects of mmo development like various content systems.

I don't know if anyone ha snoticed, but the unionization is in key areas that AI could fill.

Currently AIs can not create working models. AI can create great static images, even on 3D, but it can barely create movement correctly for a human. For a monster like that, AI would get it totally wrong.

Perhaps in a decade we'll have AIs that understands what body's parts are and how they should move well enough.

AI is fastest developing technology in the history of mankind. Granted you still need humans to finish the job but soon AI will be able to it all

 for movement, there's already 2D rigging that I can use to adjust poses, it will definitely become a 3D norm  in a few years.

Here is an easy site to test some AI

https://www.genmo.ai

OG_Solareus

Cactus_LFRez said:

AI ultimately is a tool that iterates based on inputs.
And so far it looks like that is limited to images and text.

If you use it carefully, a not so good artist could take examples of their not so good art and AI could clean it up.

But AI so far lacks the purpose and deliberate choices of human made art, it literally just amalgamates stuff.  It all looks very fake.

I would only use AI art as a way for a non artist to make concepts, like a mood board, and then pass it off to a real artist to make real concept art.  But again AI cannot create anything truly unique so it will only create a generic amalgamation.  So if you truely have a unique and specific idea AI probably won't help too much.

I would not try to use it to make game assets, I think people will be able to tell it is AI generated.

I agree to an extent, but why are you deferring language as not an art. To get decent images you need to at least have the idea in your mind on what you want to see.

Creating "Prompts" to get cool images is a trial an error. but you still need to be able to iterate what it is your mind is seeking. What I like about AI is that you "can" stumble upon something even better than what you initially set out to do.

OG_Solareus

Angrakhan said:

Yeah near term you will see AI used to create art and sound assets for games. This will not be limited to MMOs. As mentioned it will be a cost cutter and time saver because what can be done by a whole art team in a month will take an AI a few minutes. From an artistic standpoint it will be copy/ paste / iterate crap like you see above, but it will be good enough to make a game out of because who really sits there and ponders the artistic virtues of the rat or skeleton or goblin you just killed for the umpteenth time? AI will be used similarly for sound tracks. At the end of the day what it's going to do is cut jobs. This is true across every industry, but for sure it's happening right now to artists. Why hire a human when the AI can work 24 x 7 x 365, never needs breaks or gets sick or wants a raise? So what if the quality of the work is less, the volume is much higher, and most importantly the cost is much cheaper. The art department will go from a team to a single director of art whose primary job is to babysit the AI and evaluate what it's producing to decide what is worth putting in the game and what is junk. This person won't need any actual artistic skills themselves. The actual artists will be heading to Olive Garden to apply for wait staff positions at least until robots driven by AI replace those jobs as well.

At the end of the day I'm not sure we'll like the consequences of moving to AI, but by then it will be too late.

This i agree with, and right now you see it with the SAG/AFTA strike in hollywood.. I don't feel sorry for them because they look like children whiing about not have enough money for private jets or limos.

I don't see art departments shrinking down more then 50% of their capicity. 24 hour cycle you need a 24 hour team. Quality control, art direction and the usual list of modeling people to make sure everythign is working.

OG_Solareus

Cactus_LFRez said:

Vrika said:

KnightFalz said:

OG_Solareus said:

KnightFalz said:

I don't know if I'd want to fight them. All I have to go on is their appearance. The use of AI in that aspect is literally skin deep.

What does that use of AI really change in MMORPGs that will elevate them to a new level?

production speed. Everything from concept to working model could be done in a few days, Adding new models to the game would be a lot easier when the pipe line is implented.

How many developer videos do we see some do tweaking a model , and working on that model for over year.  AI can do it faster and  in the near future with less mistakes..

It would also cut costs in the number of people needed to "create concepts". Cutting those costs I think would help greatly in other aspects of mmo development like various content systems.

I don't know if anyone ha snoticed, but the unionization is in key areas that AI could fill.

I don't care about production speed or cost. What I care about is the nature of that produced and how it transcends that which came before. Being faster and cheaper to make alone doesn't elevate MMORPGs to a new standard. It just allows more to be pumped out at the current standard.

The devs don't really care whether you care. If they can make 80% of the game cheaply to match current standard then that's what they set out to do, and then they'll spend their time and focus to improve the 20%.

There are maybe 2 studios in the world who have big enough project budget to improve everything from what's been created before, Rockstar and Cloud Imperium Games. The rest of them choose some areas to focus on and elsewhere do intentionally only average work to stay on budget.

I am not sure how much time AI could really save,

As far as I am aware, it can't make 3d models, it can't make texture maps either.
AI can only really make 2d images and it can only amalgamate preexisting images often with nonsensical details.
It can't for instance rotate a 2d image to make a template for a 3d artist to work from, or even generate the same character twice in a different pose.

Like this


Maybe it could advance more to do some of these things but it has a long way to go and I am not sure how useful it really is right now.  You would probably spend more time fixing the things AI creates, you might as well do it yourself.

Right now I see AI being used mostly to create Steam trash.

you are behind. AI can make texture and normal maps now. video above shows this.

The old school PNG file breaths life again as it can store information  about the image and can manipulated very quickly.

To get high rez texturing the AI upscaling is silly insane.

This was my first upscape x2

https://twitter.com/ai_d__g/status/1700062298730123652

Imagine this process of developing a scene, beable to upscale to 8k by only scaling 512px x 512px at time, it is pretty awesome. SDevelopers can use the ai upscaling now on existing texture in 2d mode and have a complete old mmo in 4k-8k in a few days.

OG_Solareus

I would concede AI animation is the slower of the groups, but we already has AI pose, and also an AI 2D skeleton ainmator much like Flash's bone system.

Scot

We're doomed I tell you, doomed!

OG_Solareus

Scot said:

We're doomed I tell you, doomed!

So you think this world would be boring if it could be generated ina playable state ?

https://twitter.com/ai_d__g/status/1700345664813613277

KnightFalz

Vrika said:

KnightFalz said:

OG_Solareus said:

KnightFalz said:

I don't know if I'd want to fight them. All I have to go on is their appearance. The use of AI in that aspect is literally skin deep.

What does that use of AI really change in MMORPGs that will elevate them to a new level?

production speed. Everything from concept to working model could be done in a few days, Adding new models to the game would be a lot easier when the pipe line is implented.

How many developer videos do we see some do tweaking a model , and working on that model for over year.  AI can do it faster and  in the near future with less mistakes..

It would also cut costs in the number of people needed to "create concepts". Cutting those costs I think would help greatly in other aspects of mmo development like various content systems.

I don't know if anyone ha snoticed, but the unionization is in key areas that AI could fill.

I don't care about production speed or cost. What I care about is the nature of that produced and how it transcends that which came before. Being faster and cheaper to make alone doesn't elevate MMORPGs to a new standard. It just allows more to be pumped out at the current standard.

The devs don't really care whether you care. If they can make 80% of the game cheaply to match current standard then that's what they set out to do, and then they'll spend their time and focus to improve the 20%.

There are maybe 2 studios in the world who have big enough project budget to improve everything from what's been created before, Rockstar and Cloud Imperium Games. The rest of them choose some areas to focus on and elsewhere do intentionally only average work to stay on budget.

If it's about maintaining current standards the extra 20% will just be increased profit. You don't improve through maintenance.

Dammam

AI tools to boost productivity are definitely an active area of development, whether to speed up coding or modeling and animation, but the current iteration has some significant limitations when creating something entirely new. Complex, constrained problems are exactly what AI excels at so it is likely that generative models for things like animations will become quite good in the near future. Simplifying it to "AI that will do it all", however, passes over the many ways AI has and will continue to impact the tools used to make games, things like motion capture, rendering, particle physics, etc. Sure, there will be attempts to replace the art team with AI, but most art teams will use tools that leverage evermore powerful AI to make their textures, models, and so on. Then there's analytics, and how AI could impact things like level design, class balance, and so on by training on past behavioral data. Again, this isn't new exactly, but the shiny, new AI architecture behind large language models like GPT could prove to be more effective in this area as well.

As for me, I am far more interested in how these types of models could be used to create interesting NPC behavior and player-driven interactions in the story. A GPT-powered DM fine-tuned to operate in a given game-world. Imagine each NPC its own AI-agent, primed with a set of characteristics that drive its decision making, and the player actions around these NPCs further skew and shape their priorities and actions. That said, training these models is not yet cheap enough for all that so I don't expect something that powerful soon. Maybe someday.

Scot

Dammam said:

AI tools to boost productivity are definitely an active area of development, whether to speed up coding or modeling and animation, but the current iteration has some significant limitations when creating something entirely new. Complex, constrained problems are exactly what AI excels at so it is likely that generative models for things like animations will become quite good in the near future. Simplifying it to "AI that will do it all", however, passes over the many ways AI has and will continue to impact the tools used to make games, things like motion capture, rendering, particle physics, etc. Sure, there will be attempts to replace the art team with AI, but most art teams will use tools that leverage evermore powerful AI to make their textures, models, and so on. Then there's analytics, and how AI could impact things like level design, class balance, and so on by training on past behavioral data. Again, this isn't new exactly, but the shiny, new AI architecture behind large language models like GPT could prove to be more effective in this area as well.

As for me, I am far more interested in how these types of models could be used to create interesting NPC behavior and player-driven interactions in the story. A GPT-powered DM fine-tuned to operate in a given game-world. Imagine each NPC its own AI-agent, primed with a set of characteristics that drive its decision making, and the player actions around these NPCs further skew and shape their priorities and actions. That said, training these models is not yet cheap enough for all that so I don't expect something that powerful soon. Maybe someday.

If they get any better they will be able to depixelate Dammam for us.

Scot

OG_Solareus said:

Scot said:

We're doomed I tell you, doomed!

So you think this world would be boring if it could be generated ina playable state ?

https://twitter.com/ai_d__g/status/1700345664813613277

Notice in that picture there are no people, not even in the cars; we became redundant code to the machines.

Nanfoodle

I don't think people still don't get where AI is taking us. The latest AI that just dropped is 155 IQ. That's twice as smart then the average person. Instine was 160 IQ. But it's still a kid. It has yet to gain the experience it needs to flex its 155 IQ. With in the next 2 to 3 years, AI won't just be twice as smart then the average human, it will be 3000 to 5000 times smarter then the average human.

How will this effect MMOs? That will be impossible to comprehend. 

Scot

Nanfoodle said:

I don't think people still don't get where AI is taking us. The latest AI that just dropped is 155 IQ. That's twice as smart then the average person. Instine was 160 IQ. But it's still a kid. It has yet to gain the experience it needs to flex its 155 IQ. With in the next 2 to 3 years, AI won't just be twice as smart then the average human, it will be 3000 to 5000 times smarter then the average human.

How will this effect MMOs? That will be impossible to comprehend. 

I think people generally overestimate how quickly any tech development will have a major impact but are very mixed on how big an impact it will have. We went from computer kit enthusiasts to social media in about thirty five years, big changes take time, but there will be big changes.