Support NeoGAF

[kicker]

Let us assume that the current average AAA game development cycle is 4 - 6 years (I don't have exact numbers). What are the factors that might suggest that number reduces in the future?

Possible generalized factors could be new tech, changing delivery model, changing prices, changing development model, so on.

For example: Raytracing was supposed to be one such factor, but for all it's worth has not seemed to help dev time yet. Studios still rely on baked lighting for core development, with raytracing being a post-release update most of the time.
That might be because the processing power required for RT necessitates it being an extra option for those with expensive cards alone.

Any other ideas?

 

[Sephimoth]

I expect AI might have some play in this.

I mean I've found myself using ChatGPT to write/fix code and shaders for Unity. Sometimes takes some pushing but often gets there.

 

[Elder Legend]

Let us assume that the current average AAA game development cycle is 4 - 6 years (I don't have exact numbers). What are the factors that might suggest that number reduces in the future?

Possible generalized factors could be new tech, changing delivery model, changing prices, changing development model, so on.

For example: Raytracing was supposed to be one such factor, but for all it's worth has not seemed to help dev time yet. Studios still rely on baked lighting for core development, with raytracing being a post-release update most of the time.
That might be because the processing power required for RT necessitates it being an extra option for those with expensive cards alone.

Any other ideas?

AI will be the solution. Many programmers in a variety of fields already use it to solve coding problems or weird bugs. Even my SEO agency that I hired that works on my website used AI to fix a line of code they were having issues with on their end. It's already happening, only a matter of time.

Imagine how fast devs can find and fix lines of codes that are hidden in god knows where.

AI will be used for a variety of aspects including:

Dialogue lines for RPGs
Much faster writing for story/lore/character backgrounds

I wouldn't even be surprised if AI will be used to generate random quests on the fly.

As a matter of fact I believe there is a Skyrim mod that already does that and is in testing.

https://www.roadtovr.com/chatgpt-skyrim-vr-ai-mod/

 

[Zannegan]

I expect AI might have some play in this

Honestly, I could easily see some next gen AI creating games and gameworlds on the fly for the kids, a la The Giant's Drink in Ender's Game... only, instead of trying to give overseers psychological insight into the minds of future leaders, the whole point will be to massage their little brains into buying more of whatever it is that takes the place of skins/hats and possibly using them as little psychological guinea pigs for protecting future trends.

Roblox meets Dreams meets ChatGPT, but with a point.

 

[ChorizoPicozo]

Raytracing

 

[Fuz]

I expect AI might have some play in this

Came here to post this.

 

[kicker]

I expect AI might have some play in this

AI will be the solution. Many programmers in a variety of fields already use it to solve coding problems or weird bugs. Even my SEO agency that I hired that works on my website used AI to fix a line of code they were having issues with on their end. It's already happening, only a matter of time.

Imagine how fast devs can find and fix lines of codes that are hidden in god knows where.

AI will be used for a variety of aspects including:

Dialogue lines for RPGs
Much faster writing for story/lore/character backgrounds

I wouldn't even be surprised if AI will be used to generate random quests on the fly.

As a matter of fact I believe there is a Skyrim mod that already does that and is in testing.

https://www.roadtovr.com/chatgpt-skyrim-vr-ai-mod/

I agree that AI will have varied use in quest design, npc implementation and player interaction, but I'm more not sure that alonr will significantly cut down dev time. I don't knoe, maybe.

The mods I've seen so far, while amazing, don't seem to be stuff that really takes that much dev time.
Theoretically, using completely artificial voices could mean writers are less dependent on VA schedules and production timelines can be more free. It could also increase the scope of the game in terms of writing, but I don't see that alone taking off 2 years off of Fallout 5 development, for example.

One area AI could definitely help with is engine interfacing and debugging. Maybe a local model could be trained on the unreal engine, so technicians could more easily work around bugs and free upndesigners to include more crazy scenarios in games.

Exciting times, really.

 

[Elder Legend]

I agree that AI will have varied use in quest design, npc implementation and player interaction, but I'm more not sure that alonr will significantly cut down dev time. I don't knoe, maybe.

The mods I've seen so far, while amazing, don't seem to be stuff that really takes that much dev time.
Theoretically, using completely artificial voices could mean writers are less dependent on VA schedules and production timelines can be more free. It could also increase the scope of the game in terms of writing, but I don't see that alone taking off 2 years off of Fallout 5 development, for example.

One area AI could definitely help with is engine interfacing and debugging. Maybe a local model could be trained on the unreal engine, so technicians could more easily work around bugs and free upndesigners to include more crazy scenarios in games.

Exciting times, really.

You are aware there are AI already that can copy a persons voice? Have you not seen all the the gaming Joe Biden, Trump and Obama AI voice memes on YouTube?

They sound very close to their original voices. They'll pay the VA to have their voice listened by the AI so it can learn it so then it can be replicated. This will most likely a conflict of interest though. Weird times ahead for a folks in the industry.

And all of which you have listed will 100% cut down dev time. I don't know exactly by how much, but it will def help. The people don't need to sit there for hours anymore taking their time writing tons of dialogue for NPCs and so on. I wouldn't even worry about Falloit 5 yet... Elder Scrolls 6 hasn't even started development yet or is at least in pre-production. Elder Scrolls 6 will most likely be one of the very first few games that utilizes AI tech for development and Fallout 5 will take it a step further.

 

[Badlucktroll]

Maybe they can automate some shit with AI

 

[feynoob]

https://www.ign.com/articles/major-publishers-report-aaa-games-can-cost-over-a-billion-to-make
According to the report, AAA games that are greenlit now with potential releases in 2024 or 2025 typically receive development budgets of $200 million or higher — Call of Duty has already surpassed $300 million in development costs alone, and the next Grand Theft Auto title will likely require a development budget of $250 million or more. When considering marketing costs, this number can jump to over $1 billion, with one large studio reporting that a major franchise's development cost $660 million and marketing cost nearly $550 million

 

[nemiroff]

Not my problem. I don't take part in the development, I play the product.

 

[kicker]

Not my problem. I don't take part in the development, I play the product.

And you don't care that the big budget product is taking longer and longer to come out when they don't even innovate that much anymore, while citing that same development as the reason the prices keep going higher and higher?

I mean, maybe you really don't, that's fine.
Still, I would very much prefer it wasn't this way.

 

[Umbasaborne]

At its current rate, its not sustainable. Major franchises like gta and the elder scrolls did not get new releases last gen because there is literally only enough time in a console cycle for a studio to make one new game now. Shit is fucked. Rocksteady havent put out a game in almost a decade

 

[James Sawyer Ford]

Better tools and AI

I think it will get a lot better

 

[James Sawyer Ford]

At its current rate, its not sustainable. Major franchises like gta and the elder scrolls did not get new releases last gen because there is literally only enough time in a console cycle for a studio to make one new game now. Shit is fucked. Rocksteady havent put out a game in almost a decade

More mismanagement than anything else

Studios like Insomniac dont have this problem

 

[Umbasaborne]

More mismanagement than anything else

Studios like Insomniac dont have this problem

Insomniac and capcom are putting the rest of the triple a games industry to shame these past few years. High quality releases on a consistent schedule. Like even fucking ubisofts pipe line is so backed up and mismanaged that they havent released a single new multiplatform game in over a year! Ubisoft, who used to put out 3-4 big games every year

 

[mdkirby]

We’re maybe a year or so away from ai generative 3d art tools, probably integrated into unreal engine, and utilising their existing tools too such as metahuman. Need a cowboy? Ask for it. Need 30 guns for your cowboy? Hey presto. Now need him in a bar? Ask for it, then ask for a whole town. Then ask it to populate the town…then use ai to help with a lot of the npc natter from scripts to voices…and you can very quickly start to see where huge huge amounts of time can be saved.

And before anyone says stuff will look stale and unimaginative, I’ve seen significantly more unique and interesting looking characters on ai art communities than anything in any game I’ve played in recent memory.

 

[rofif]

gta 4 took 3,5 years to make.
gta 5 took 5 years to make.
Mass Effect took 3,5 years to make. Sequel took barely over 2 years.
Same with gears and uncharted 3-2 years between games.

 

[nemiroff]

And you don't care that the big budget product is taking longer and longer to come out

I don’t know about that. But even if it's true It doesnt affect me, so why would I care. The market is abundant with games in my view. HOWEVER, I do of course sometimes find the stories around development interesting.

 

[DaGwaphics]

Isn't UE5 supposed to make things a little quicker?

Automation and improved tools that reduce the time needed to do the work would seem to be the only things that would make things quicker.

 

[CGNoire]

https://www.ign.com/articles/major-publishers-report-aaa-games-can-cost-over-a-billion-to-make
According to the report, AAA games that are greenlit now with potential releases in 2024 or 2025 typically receive development budgets of $200 million or higher — Call of Duty has already surpassed $300 million in development costs alone, and the next Grand Theft Auto title will likely require a development budget of $250 million or more. When considering marketing costs, this number can jump to over $1 billion, with one large studio reporting that a major franchise's development cost $660 million and marketing cost nearly $550 million

Skeptical.....someones pocketing a huge chunk of that.

 

[Varteras]

I expect AI might have some play in this.

I mean I've found myself using ChatGPT to write/fix code and shaders for Unity. Sometimes takes some pushing but often gets there.

 

[feynoob]

Skeptical.....someones pocketing a huge chunk of that.

It's mostly due to wages.
The wages aren't the same as 2010s. People now have a choice to leave their current company and go to another developer who will pay higher money.

It's either spend a lot of money retaining those people or lose them to your competition.

 

[Knightime_X]

AI will play a HUGE roll in development across the board.
I think some people have no idea what's coming.
It's going to be insane.

 

[Varteras]

AI will play a HUGE roll in development across the board.
I think some people have no idea what's coming.
It's going to be insane.

The next 5 years alone will be transformative for a lot of industries. Nevermind 10 or 20 from now. Like you said. Buckle up. Shit is about to get wild.

 

[March Climber]

Smaller example clip:

 

[{insert code here}]

AI generated art assets (3D models, textures, animations) should do it but that's going to put 2/3 of the industry out of work.

 

[fran.fernandez]

Unreal engine 5

If Sony and epic start doing more support together development would increase drastically. But that would mean an unreal tweaked specially for Sony platform.

Anyway. Less own engine more unreal

 

[Glass Prison]

I want to see more triple A studios outsource projects to smaller studios (think of Bethesda and Obsidian for example in Fallout 3 and New Vegas). This wouldn't shorten the development time of the Triple A game itself, but we would see more spin offs of the games we love. For example: while we wait for the next 'proper' God of War game, I'd love to play a spin off that is more in line with the PS2 GOW gameplay.

I don't know. I'm all in for more creative takes on already existing titles, but I know not everyone would like that.

 

[MiguelItUp]

Some QA departments have already dabbled with automation for awhile. Rather than manual user bugging/logging and verification.

I could see using AI in to help with running test cases, and maybe even creating test plans.

Unsure of where other departments may thrive with the use of AI. Aside from QA, I suppose Art, Sound, Coding/Programming, and Marketing could all benefit from it in some way. But in the long run, I don't really see the overall development time drastic changing or anything. Maybe lighting some loads.

 

[Draxon Gamer]

No, that would require console makers that make it easier for developers to create games at a faster pace which hasn't happened on consoles...ever, and on computer hasn't happened since the 80s.

They give developers new technology they can use, but developers have to spend more money to take advantage of it each time, so the costs only go in one direction until someone stops and says, maybe like every other industry Video Games should also come up with new ways to cut costs. No one has figured out how to do that yet strangely.

 

[onnextflix5]

With ai and better tools, hopefully we can get back to 3 year dev cycle, as current model is too expensive and not sustainable.

 

[clarky]

This is what happens when you stop mandatory overtime for the softies. Broken games with massive development times. Hopefully AI can pick up the slack

 

[TheAmazingSpiderMan]

https://www.ign.com/articles/major-publishers-report-aaa-games-can-cost-over-a-billion-to-make
According to the report, AAA games that are greenlit now with potential releases in 2024 or 2025 typically receive development budgets of $200 million or higher — Call of Duty has already surpassed $300 million in development costs alone

I don’t know how game development works but I’ve never understood this, cod games clearly reuse animation, assets, and other bits and bobs in the next mainline cod with a fresh lick of paint, I mean what’s costing so much? All the gun animations gun models etc all pretty much the same as previous games also same engine being used, I’d imagine making the graphics look good may take a lot of resources.

 

[bender]

Hopefully as studios and publishers continue to swallowed up by mega corporations that will lead the true creatives to break off and work on smaller more personal projects.

 

[kikkis]

I don’t know how game development works but I’ve never understood this, cod games clearly reuse animation, assets, and other bits and bobs in the next mainline cod with a fresh lick of paint, I mean what’s costing so much? All the gun animations gun models etc all pretty much the same as previous games also same engine being used, I’d imagine making the graphics look good may take a lot of resources.

I dont think cod reuses a to a meaningful degree, I mean even exact same guns were re modeled and animated in mw19 vs mw22. I think biggest issue is just at least some games are much bigger than games used to eg consumers expect more. And that takes time.

 

[StreetsofBeige]

Reduce:

- Open world scope. Bigger isn't always better

- Cut scenes and dialogue. Funnel those costs to the rest of the game

- Reliance on 4k and nitpicky textures. Most people cant tell the difference and meticulous details are only appreciated by a small number of gamers who want to star at barrels and if the bolts in the wheels of a car are accurate

- I dont know know how much processing power high end Dolby Atmos, Tempest and such take up. But if it does eat up a lot then scale it down. Most people dont care or even have the audio gear to even take advantage of it

 

[Comandr]

Reduce:

- Open world scope. Bigger isn't always better

- Cut scenes and dialogue. Funnel those costs to the rest of the game

- Reliance on 4k and nitpicky textures. Most people cant tell the difference and meticulous details are only appreciated by a small number of gamers who want to star at barrels and if the bolts in the wheels of a car are accurate

- I dont know know how much processing power high end Dolby Atmos, Tempest and such take up. But if it does eat up a lot then scale it down. Most people dont care or even have the audio gear to even take advantage of it

Worst suggestion in the thread by a mile.

The solution isn't to reduce fidelity and scope, but to develop new tools to facilitate faster development and deployment of such features. I don't want to have to go back to PS2 games because it takes too long to make them now.

 

[lefty1117]

Was just talking with my boy about this. The industry needs some fundamental change in how games are put together. Maybe it's AI or maybe it's some other tech revolution, but something has got to give because the industry is in a bad state right now, imo.

 

[OCASM]

AI will not help. It will simply allow for the creation of bigger and more complex AAA games.

In economics, the Jevons paradox (/ˈdʒɛvənz/; sometimes Jevons effect) occurs when technological progress or government policy increases the efficiency with which a resource is used (reducing the amount necessary for any one use), but the falling cost of use increases its demand, increasing, rather than reducing, resource use.

Jevons paradox - Wikipedia

en.wikipedia.org

 

[Zannegan]

I think a large part of why some recent "safe bet" games have bombed is that they just missed their window.

Sure qualiity control has been a factor, and some concepts are just bad in and of themselves, but with studios having to invest 5+ years to get a game out, it's entirely possible that a game that would have been a koderate success two years prior bombed because that style of gameplay (or that tone, character design, etc.) is played out by then. Not to mention, the hype cycle both creates unrealistic expectations and diminishes enthusiasm the longer it stretches on.

This puts immense pressure on studios to try to predict the future and guess trends years ahead of time or to pivot and try to bolt on or cut out systems and features to respond to the changing demands of the gaming public.

For its own good, the industry has to learn how to be more nimble again. Whether that means planned crunch compensated for by long vacations post-release (sadly, we're more likely to get the former without the latter), leaning heavily on AI generated assets and code, reducing the scope of projects in general, or some combination of the above, I don't know, but they've got to figure it out.

 

[analog_future]

UE5 seems to have a lot of tools designed to simplify and streamline the development process.

 

[GHound]

Ever think about how many games we hear about being rebooted during development multiple times to chase whatever shitty trend is currently mucking about? Yeah. Stop that.

 

[Playboi Carti]

Raytracing

 

[Smether]

I think pandemic and gross gen period have stunted growth and adoption of Unreal Engine 5. If not, I think devs would have been farther along in learning the tech and now to use it properly. This will naturally lead to time savings and more efficient turn around in projects. It will improve as developers get more and more familiar with consoles also.

 

[Kataploom]

Four things comes to mind that WILL reduce dev times a lot:

1. Devs knowing not always more is better, since many players avoid lengthy games like AC in order to finish games within a life time

2. AI

3. Geometry shader. LOD manual optimization workflow should be gone by late this gen... Also it would allow artists could work without polygon budget limitations unless for specific stuff.

4. Raytracing. When hardware is good enough to push full RT lighting models, baked lighting setups should stop slowing down artists on their workflow.

I could be wrong but that's as far as I can see.

 

[vaibhavpisal]

Nope, not coming down.

Unless you are ok with rehashing games.

 

[Daneel Elijah]

Unless you are ok with rehashing games.

This is why I believe that the AAA will be easier to do in the future. With the tech slowing down and photorealism being easier than ever to do, we will have in one or two generations enough games that did the job to simply push from the base for a long time. TLou 3 or 4 will be good enough as a base to make super DLC tech wise if you see what I mean. Or GTA being good enough for 10 years now. This will be the path going forward, for those who already have great IP. The FIFA model of incremental upgrades. For new IP I can see them using early access more and more with mods used to push them up to parity with older AAA until a game made enough to warrant the work needed to be modernised. For consoles Sony will either go the Nintendo route of being one of the few to make it work in their consoles, or go GAAS all the way ( ) and be like GTA with 10 years plans. Microsoft will continue Gamepass so early access in all but name too.

 

[lh032]

focusing on gameplay instead of graphics i guess.

 

[X-Wing]

There is so much to play. This doesn’t worry me the slightest. I don’t mind waiting for great work.