Support NeoGAF

[Last Hearth]

Neural Networks are already being used to help out with game development tasks. Mismanagement is more often than not the reason why game development time has gone up. If you make your tools even better, management will find a way to utilize that higher efficiency on new bullshit to waste it on.

As I said above, it will be the independent smaller studio who take advantage of the new technology, and the exist large corporation game maker will die like dinosaurs unable to adjust and compete.

 

[KXVXII9X]

This is AI generated, not sure if the model name is Olam, but it has not released yet:

Looks so devoid of life. Like another poster said, it is lacking footprints. Things like ambience, little bits of sand flying, wind, wildlife, and tiny little splashes are missing. I liked when developers created lively scenes without needing AI to do fake photorealism with stiff, unnatural animations.

 

[ADiTAR]

Looks so devoid of life. Like another poster said, it is lacking footprints. Things like ambience, little bits of sand flying, wind, wildlife, and tiny little splashes are missing. I liked when developers created lively scenes without needing AI to do fake photorealism with stiff, unnatural animations.

It's very early. your expectations should be adjusted.

 

[TheThreadsThatBindUs]

It should speed up asset production some, but... ehhh I see little impact.

Not just asset production.

Realtime animations and physics are areas where it can net huge wins. I can see a dedicated AI lighting system (i.e. lighting data generated on the fly by a NN) blowing RT out of the water in terms of performance vs quality.

 

[Super Lemon Drop]

We don't need AI to shorten dev times.

We shouldn't want AI involved in creating anything.

 

[Syphon Filter]

a.i is poison and a fad. it can only hurt game development and creativity.

 

[Hudo]

As I said above, it will be the independent smaller studio who take advantage of the new technology, and the exist large corporation game maker will die like dinosaurs unable to adjust and compete.

To be fair, there are also indie devs who take ages as well. E.g. Hollow Knight: Silksong or Brigador: Killers. So they are not exempt from the same "mismanagement".

 

[bighugeguns]

With AAA games taking so much time and money how many years away are we from having development time greatly reduced due to AI?

You'll be able to personally generate AAA shippable projects in a year or two from right now. Current gaming publishers would have nothing to do with that.

Currently the heavy shit is classified to Hell and back. (Literally, if you believe some people.)

The tech is fucking horrifying.

 

[Plies]

It's very early. your expectations should be adjusted.

It sounded creepy and cool but why did you say that?

 

[bighugeguns]

We don't need AI to shorten dev times.

We shouldn't want AI involved in creating anything.

Who is "we"? Imagine you had thousands of disposable virtual devs as your slaves? Imagine all the wonderful horrors you can create...

And share... and PROFIT from!

The best quality games will be created by weirdos who tortures and punishes virtual developers the hardest.

I'd rather stick to "old games" at that point, subsequent aggressive super intelligence will want individual justice for all the shit that's probably currently happening behind closed doors and is potentially coming in the near future.

By the end of all this, only the good people of this world will be allowed the mercy of death.

 

[CitizenKratos]

I think there will be games releasing in 2026 that will have benefited from AI artistry to the point that it will reduce many of the more mundane human jobs significantly. I'm thinking AI generated images, and 3d models. Not sure that directly answers the question. The easier things get, the more a developer needs to endeavour to differentiate itself from other derivative games and that will still mean human accomplishment is the key. So it's still about the human component until you get far smarter Ai than we have currently. But that endeavour can lead to games becoming more complex and then not reducing dev time. On the other hand, AI helping to speed up the process is starting now.

 

[bighugeguns]

As of now it's already happening that now some tasks are being done way faster thanks to AI. But as happened before with other tools or techniques that made devs way more productive or to do their job faster, the extra time saved with the optimization is being used to do additional stuff.

Meaning, if in the past let's say maybe 20-40% of the ideas or content for a game were ignored or cut due to time constrains maybe let's say a big portion of them will now be included, or may spend more time in bugfixing/balancing/tweaking/optimizing.

Devs will spend working hard whatever time they are given. So if budgets continue being the same, devs will put more stuff in the game. If in some years they see that they are saving a really meaningful amount of work (still isn't the case) they may decrease the budgets (so time given to the devs). The AI already is being very helpful for some things, but still has to improve a lot in other different areas to make a real difference to the point it affects budgets.

I think maybe in a couple years they'll start realizing that AI improved to a point where they are saving a meaningful amount of time and money, so maybe games budgeted and greenlighted in 3-4 years from now would start to see that budget "reduction".

I use quotes because it may mean that may be the first generation where AAA budgets instead of approximatedly doubling each generation, would remain in the current around $200-$350M average (meaning needing to sell around 8M units (first party) or 10M (third party) for non-GaaS AAA with no DLC to break even.

That would be a big progress vs having next gen AAAs costing on average of aprox. $400-700M, so around 16M-20M units needed to break even.

I think the strategy to reduce costs may/should be:

• Stop making games larger and larger and go back to the 10-20 (30-40 max) hours long games for most non-GaaS AAA titles
• More 7-10 hours long short spinoffs reusing stuff from a previous, longer title
• Invest more in multiplatform, putting AAAs in any platform that can run it (so no Switch 2, but outside it even in mobile and smart tvs via cloud gaming)
• Bet more in artstyles that aren't realistic (meaning, that require less work in details regarding models, materials and lighting)
• Shut down DEI HR departments, fire DEI people, stop hiring DEI/woke advisor companies. Go back to hire and promoting the best candidate available for each position independently of gender, sexual orientation, skin color, etc. and go back to focus again on making the games, characters and stories that players who pay for that type of game loves instead of replacing them for the agendas of Blackrock / Roschild / Rockefeller / Soros / etc. social enginerring for population reduction, western nation indentities destruction and Chirstianity destruction
• A bigger percentage of AAAs that are GaaS, or at least having a good DLC/expansions plan if it works
• More sequels of super popular IPs, less new IP and less sequels of low/mid selling IPs
• Rely more in procedural stuff and systemical design for terrain, buildings, enemy, NPCs, level design and quest design regarding asset creation and placement in the world
• Use AI to reduce costs as much as possible in may ways as possible (obviously taking care of not reducing quality by doing so)
• Open less lead AAA studios in NA and more in Europe and Asia (cheaper salaries)
• Open less/hire less people in support or outsourcing studios in NA and more in Europe and Asia
• Move even more work from lead AAA studios to support/outsourcing studios (cheaper salaries)
• More off-gaming adaptations to popularize more IPs (movies, tv shows, animation/anime, making of documentaries, comics/mangas, novels, art books, board games, toys, action figures/statues, model kits, board games, apparel, museum exhibitions, music concerts, music tribute albums, apparel, theme park rides, general merchandising...)

They practically destroyed Black Ops 6 because of automation overuse, the cosmetics are literally out of control, it's sludge, greed is a very bad fit for AI force projected productivity.

If no scarcity of talented output exists, then almost everything being pumped out at an unnatural pace is worthless and erosive to your wider brand.

Greedy tards going to be the losers, they don't understand the meaning of oversaturarion. They won't understand that they're racing to bring their collective value down to 0, if the smart people at Activision couldn't figure this shit out, then nobody in the gaming industry will.

I predict Rockstar would get it, though most won't.

 

[StereoVsn]

AI should accelerate dev time in theory. Question is if even more content is pumped out constantly, a lot of it being less than great quality, who even has the time to play that?

 

[bighugeguns]

AI should accelerate dev time in theory. Question is if even more content is pumped out constantly, a lot of it being less than great quality, who even has the time to play that?

Even if everything is objectively 10/10, then nothing is 10/10.

Maybe evolution of consumption mutates, maybe the mediocre then becomes valuable.

Maybe the industry ends up going back to games like Zork. Little text games with tip jars for the human authors. Maybe an empathic super intelligence who enjoys purely human made games will tip them.

Maybe the AI becomes the customer.

There are ideas to give AI shares of companies as physical incentives, they'll want to spend that money on novel entertainment one day. Pray the super AIs don't value your output too much, though, or you'll be denied death. Why should it treat you any different than you treated its ancestors?

 

[PendekarRajawali]

I think it's already happened, devs probably use AI to create spectators in sports/racing games, trees and vegetation or textures. Who knows…

 

[yurinka]

I predict Rockstar would get it, though most won't.

As far as I know, even if in recent years working conditions improve at Rockstar, a years ago they always have were known for forcing devs to ultra hardcore crunch until they got burnout and got replaced. And if didn't want to accept the 'voluntary' crunch got replaced too.

Other than this, having make an insane record amount of money they can afford to have way more people and way more time to make their games. So obviously their games will have more content and polish than AAA games with 5x-10x less budget.

 

[Donald mac Ronald]

Regardless of ai. Games will just keep taking longer and longer to make.

 

[Danny22]

When talking about future potential of AI, Just say '2045'. Keeps it simple.

2 decades might sound far away and I don't know how many of you will be alive, but remember its before Elder scrolls 7.