How many years are we away from AI significantly reducing game development time?

F

FMX

Member
With AAA games taking so much time and money how many years away are we from having development time greatly reduced due to AI?
 
Last edited:
Already happening today, tons of tasks can be significantly sped up with the use of AI, from asset creation to code to the manager making a power point.

Tool's already quite powerful. Sure it could do other stuff like musical production or level design but I'd rather leave that to the humans.
 
Last edited:
Hard to say. Depends on the game I'd wager type I'd wager. For sports games it will speed them up significantly. Stuff with large areas needing to be filled will as well.
 
Trogdor1123 said:
Hard to say. Depends on the game I'd wager type I'd wager. For sports games it will speed them up significantly. Stuff with large areas needing to be filled will as well.
Click to expand...
We've had tech to do that for decades. Speedtree and random generation have been around forever.

I don't think it is going to meaningfully impact dev times.
 
Dude give it no more than 10 years and we'll probably be playing games that are entirely generated in real-time per user, based off of our own preferences that the AI algorithmically already knows.

Things are gonna look WILDLY different.
 
Already happening. It's not going to look like a steep drop on a graph. These are going to be gradual changes and processes added to existing workflows.
 
Games being developed right now are being impacted IMO.

I'm not sure how long until we can actually see the effect like in shorten development times. It takes a while to integrate it into workflows and to people to actually learn how to use them.

Even for typical office stuff companies are struggling.
 
nW9wzUvCkTMm3qcA.gif

What AI will enable is more

A Model T car was produced in around 2 hours, modern cars take around 17-30 hours.... why?... complexity


videogames are just going to get more complex and sophisticated and people want moooreee content and shit in them
 
I think the important part of the question is meaningfully impact dev times. AI is being used in game development to shortcut certain tasks and improve aspects, but it isn't shortening dev cycles yet (the average dev cycle is still just as long). It's being used to make better games, but there is still a supply vs demand of graphical expectations that need to be met and is rising every year. How developers use these tools to become more efficient will improve with time. Right now, we are not getting the best efficiency out of them.
 
FMX said:
With AAA games taking so much time and money how many years away are we from having development time greatly reduced due to AI?
Click to expand...
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...)
 
Last edited:
AI is already destroying the whole visual novel genre as we speak, no more mid budget handcrafted vn's because now they are all the same looking AI slop.
 
Lorianus said:
AI is already destroying the whole visual novel genre as we speak, no more mid budget handcrafted vn's because now they are all the same looking AI slop.
Click to expand...
VNs were never of any interest to me, but I always wondered.
Do people play it for the art style or for the story that is told?
 
ChorizoPicozo said:
nW9wzUvCkTMm3qcA.gif

What AI will enable is more

A Model T car was produced in around 2 hours, modern cars take around 17-30 hours.... why?... complexity


videogames are just going to get more complex and sophisticated and people want moooreee content and shit in them
Click to expand...
I concur.

The video game industry are all in the complexity addiction and AI is not going to fix this. If anything, it'll fuel even more complexity as menial and time consuming task gets brushed underneath the AI toolbox.
 
Last edited:
AI is here to stay and will play a major role in the future of game development

Dismissing that reality is simply out of touch
 
HRK69 said:
AI is here to stay and will play a major role in the future of game development

Dismissing that reality is simply out of touch
Click to expand...
Playing a major role doesn't necessarily mean that development times will get shorter. AI might speed up some areas of game development, but it might also open up completely new ones that could take just as much or even more time to implement, maintain and support. Just think of the billions of ways AI-generated NPC dialogues and player-led conversations could go catastrophically wrong, for example.
 
It's kind of already happening, but the problem is that profit motive encourages companies to do more work with less people, so they save money, but the project still takes the same amount of time, if not longer.
 
I dunno... it might be difficult to have an AI cut & paste the same 5 battle objectives over a procedural generated map the way expert Ubisoft employees can.
 
Here's my real world but anecdotal information:

For me, my game development skills are still rather "pure" from AI helping me too much because ChatGPT can't make any GameMaker code that does anything helpful. It helps much more with web development. I'll ask it "Is there a JavaScript function to do ___?" or "What CSS is used for ___ effect?" or I'll have it save me time from manually making changes to a number of HTML elements that I've already typed. It's helped me much more with PHP but that's mostly because I don't like that language and I don't know it as well as the other web related languages I do know. You might say "Why don't you just Google it?" That's easier said than done. I'm usually given a list of unhelpful stack overflow responses, or reddit posts, or overly-long blog posts. I usually typed in "___ w3" to give me a W3 Schools results since that has actually helpful information. Search engines are pretty bad for finding anything now a days. Seriously, that's the main reason why I use ChatGPT. It's because Google is so bad now.

For my wife, she has Microsoft Co-Pilot integrated into her IDE (either VS Code or Visual Studio) and it will suggest stuff you were already going to type which is pretty useful.

Any time I've ever asked it to "just do the whole thing for me" it never "just works". It still requires tons of manual modification. I've had the thought before "If I spent the time to learn this the right way instead of trying to take a shortcut, I could have had it done by now." I really don't think these things are going to get smarter and smarter. I think they will plateau hard.
 
Last edited:
Yeah I'm gonna go ahead and say this is never going to happen. While AI significantly increases productivity in some key areas of game development, the vast majority of tasks, projects and roles can't be easily replaced with some AI run or enhanced substitute. Let's say the overall productivity increase is 30% (why just 30 you might ask? Think of the huge overhead in these massive companies and then ask yourself "how much of the actual work being done to develop games can easily be replaced by AI?") - that 30% will instantly be neutralized by either salary increases or development time stopgaps introduced by some new idea (I don't know, maybe quality assurance lol). People find a way not to diminish their importance in the company one way or another.
 
Game development time and game development scope are two different things. AI may reduce the time to do certain things but if your scope is on the level of GTA6 or higher, then development time will still be pretty long.
 
games that are being developed right now one way or another is using AI and it is reducing development cost and time
 
ChorizoPicozo said:
nW9wzUvCkTMm3qcA.gif

What AI will enable is more

A Model T car was produced in around 2 hours, modern cars take around 17-30 hours.... why?... complexity


videogames are just going to get more complex and sophisticated and people want moooreee content and shit in them
Click to expand...

Over on the NeoCAF:

How many years are we away from AI significantly reducing car production time.

post #10
nW9wzUvCkTMm3qcA.gif

What AI will enable is more

GTA2 took 2 years to develop GTA6 has taken over 10......why?....complexity

Cars are just going to get more complex and sophisticated and people want mooooreee shit in them.






NeoCAF = New Car Age Forum



My point car production and development could be sped up using AI if it can find more efficient ways to do parts of production, testing etc etc, the same way CAD helped speed up the designing of it.
If lets say 80% of the textures in a game, are generated in 2 minutes, that will def speed up production. (AI texture generation is already a thing btw)
Animation is one of the most time consuming parts of game dev......let the AI do it.
QA? Boot up the game on 10,000 AI models and have them play the game over and over and over, you could get that first month of teething problems near ever game has at launch done.


And thats just me spit balling, we could find many ways to have AIs help in game dev to reduce the time it takes to actually do shit.
If the AIs get better and better which they inevitably will and dev studios train them for specific things we could cut a bunch of time out because they can do that shit super fast.



Now will scopes get larger...sure, but not across the range, Game Design is probably NOT the thing we are gonna see change massively sooner than the AIs can catchup to current and near future trends.
As sad as it may sound we are gonna be playing largely similar games over the next ~10 years.
 
Folks here that state it's never going to happen have no idea at the rapid pace of A.I. less than 5 years from now we will have A.I. that is as smart as the smartest person on this planet on every single subject matter and will be able to think thousands of time faster and there will be millions of the them. Super intelligence won't be far behind. 10 years from now they will be 1000s of times smarter than us.
 
My estimate of when AI gets seriously useful gets shorter and shorter every day.
There's an AI which can pick out detailed mo-cap motion data straight from video, no special suit with ping-pong balls needed, it can also seamlessly stitch together different motion captures. It would have been first page tech news if it happened 3 years ago. Now it's just noise drowned out by the constant AI progress.

I think 1-2 years till we start seeing dramatically cut down development times for your typical Sony walking simulators.
 
CuNi said:
VNs were never of any interest to me, but I always wondered.
Do people play it for the art style or for the story that is told?
Click to expand...
If the art didn't matter the genre would be called "novels".

Anyways I suspect the slop doesn't end at the art either, why stop with it when you can have AI 'crafted' narratives as well.

cyberheater said:
Super intelligence won't be far behind. 10 years from now they will be 1000s of times smarter than us.
Click to expand...
In this scenario the idea of 'gaming' as it exists(alongside the entirety of economic system) today is also completely obsolete.
So the OPs question no longer applies to anything either.

That said If our entertainment at that point is everyone having their personalised Neo Matrix experience, the costs of those will be astronomically different as well. Possibly not touched by humans anymore though, so also irrelevant to us in said terms.
 
Lorianus said:
6/10 not leaving footprints (the most satisfying thing when walking on that wet beach sand)
Click to expand...
It's a good start though. I'm really thinking 3D engines will be obsolete in 3-4 years, if not sooner.

There's already https://about.decart.ai/ which is kinda doing real time. We gonna get a lot of shitty games and content, until it'll sort itself out.
 
Last edited:
Do T2 wants a new GTA every 3 years ? same thing with Halo and other major publishers, if AI got expensive then we're screwed.
 
Last edited:
FMX said:
With AAA games taking so much time and money how many years away are we from having development time greatly reduced due to AI?
Click to expand...
With AAA, still a long way, what you'll see is indies and smaller studios with less money, starting to use it more and more, to create content that will rival those AAA companies in terms of production value, but at a fraction of the cost and time (think E33 but maybe even closer to AAA production value) Until basically the current AAA industry will completely go out of business because they simply cannot compete, and the new smaller studios will become the next major ones, or there will be a massive turn in the AAA industry once they realize what's going on and they are hemorrhaging money. My bet, based on historical patterns, is scenario one. The build-in culture of massive corporations are like dinosaurs, they simply lack the ability to change course because they have ingrained corporate culture, and individuals in them are more about protection their own small turf rather than what's best for long term organizational survival. This is a problem with all large corporations, all the people in it are people who can jump ship, no one really cares about the organization as a whole, only what benefit they can get out of it.
 
Dacvak said:
Dude give it no more than 10 years and we'll probably be playing games that are entirely generated in real-time per user, based off of our own preferences that the AI algorithmically already knows.

Things are gonna look WILDLY different.
Click to expand...
I came here to say this.

And when I finish playing the game I just made, I'll watch the eleventh season of Friends, trained off the data from the first ten seasons but featuring me and people I know as side characters.

Everything is going to change.
 
Small John McCarthy said:
I came here to say this.

And when I finish playing the game I just made, I'll watch the eleventh season of Friends, trained off the data from the first ten seasons but featuring me and people I know as side characters.

Everything is going to change.
Click to expand...
I can't wait for AI to remake the last few seasons of Game of Thrones.
 
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.
 
You must log in or register to reply here.

Similar threads

Sakurai: modern game development is too time consuming and unsustainable. AI is the solution.
2 3
112
1K
Felessan replied
Sony Senior Vice President: We are moving away from a hardware centric business model
11
179
I Love Rock 'n' Roll replied
[GamesIndustry.biz] 87% of game developers are using AI agents in their workflows, says new survey
Business Game Dev 
36
37
FoxMcChief replied
Like it or not, the future of game development is AI: Google Genie 3 video
40
478
poppabk replied
Genie 3 is out the bottle. New AI development.
News  2
94
792
Draugoth replied
Top Bottom