EA Says Generative AI Could Make It 30% More Efficient (videogameschronicle.com) 46
EA CEO Andrew Wilson believes generative AI will "revolutionize" the gaming industry over the next five years. He predicts that the technology will allow for more efficient content creation, reducing development time from months to days. From a report: Greater efficiency coupled with "deeper, more immersive experiences" will lead to significant audience expansion over the next few years and provide a "multi-billion dollar" growth opportunity, he said. Wilson said that in the past it might take six months to build an in-game sports stadium. Over the last 12 months, that time has shrunk to six weeks, and over the coming years it could maybe be cut to six days.
And while FIFA 23 has 12 run cycles for how the players move in the game, EA Sports FC 24 has 1,200 created with generative AI. Over the next five years, Wilson hopes that generative AI will make EA's development 30% more efficient, help grow its 700 million-strong player base by "at least" 50%, and lead to players spending 10-20% more money on its games. "What we've seen every time there's been a meaningful technological advancement in media and in technology, where you are able to democratise an industry and hand it over to the population at large, incredible things happen," he said.
And while FIFA 23 has 12 run cycles for how the players move in the game, EA Sports FC 24 has 1,200 created with generative AI. Over the next five years, Wilson hopes that generative AI will make EA's development 30% more efficient, help grow its 700 million-strong player base by "at least" 50%, and lead to players spending 10-20% more money on its games. "What we've seen every time there's been a meaningful technological advancement in media and in technology, where you are able to democratise an industry and hand it over to the population at large, incredible things happen," he said.
30% more loot boxes (Score:4, Insightful)
You heard it here first.
Re: (Score:2)
30% fewer EA employees, for sure. If you read this news article and you work for them, your head is now on a chopping block.
Re: (Score:2)
30% fewer EA employees, for sure.
When your workers become 30% more profitable, you don't fire them. You hire more.
It is a fallacy that there is a fixed amount of work to be done. Efficiency improvements lead to an expanded economy, not fewer workers doing the same old jobs: Lump of labor fallacy [wikipedia.org]
Tech worker employment is an example of Jevons Paradox [wikipedia.org]. Efficiency improvements lead to higher demand.
Re: (Score:2)
Tell me you're someone who has never worked for a company that is grounded in modern reality without telling me you're someone who's never... you get the picture.
As someone who's directly experienced exactly what you say "doesn't happen" more than a few times in his life and sadly observed the same in most all of my friends and family; I can say you're dead wrong.
If a company can make 30% more money with zero additional expenditure; they will... then demand an additional 5% out of their existing workers....
Re: (Score:2)
If a company can make 30% more money with zero additional expenditure; they will...
If a company can double profits by investing in new capacity, they would rather do that.
There may be exceptions to the rule; but it's still the rule.
If that were the "rule", the economy would've collapsed when the steam engine was invented.
In reality, incomes have improved twenty-fold since then, all because of the efficiency improvements you believe should've led to impoverishment.
EA Won't be Too Happy Soon (Score:4, Interesting)
One of the best uses for generative AI will be in video games. And it'll enable indie developers to build much more. Indie developers who, at least initially, will present much higher quality games without all the usual bullshit that comes with AAA studios.
I think for those who enjoy video games, this is going to enable a lot more competition. Great for us players, but the AAA studios can't see past a quarter or two. They merely salivate at laying off tonnes of staff, use AI to push out mediocre brand-related nonsense, while new, smaller indie studios push out far bigger projects than they were ever capable of.
I think we're going to enter into a golden age for a while. Then the Indies will become the new monsters and get greedy and go public and enshittification will happen. But at least there will be a few years or more of some incredible things, things not coming from EA, Paradox, Epic, Bethesda, Ubisoft, etc etc etc
Re:EA Won't be Too Happy Soon (Score:4, Funny)
For some reason knowing the authorship changes the perception of the work, at least for me. Take for example No Man's Sky, an essentially unlimited universe to explore; but knowing that it was not made by human hands cheapens it. There's tons of open world games out there, there's not much functional difference between a mostly empty world made by humans, and any given planet in NMS. Yet I'd much rather run around Elden Ring or Zelda's world than traipse around a planet in NMS.
Looking at it another way, pretend we were talking about novels here instead of video games. Would an AI generated novel ever compare to flesh and blood author? Assuming you knew beforehand it was not authored by a human.
Re: EA Won't be Too Happy Soon (Score:2)
One example: Procedural generation and generative AI will solve problems like repetitive or non context sensitive NPC dialog. A big studio has the money to pay voice actors to record thousands of phrases. A small indie studio does not.
It's not just about generating maps. There is so much more possibilities that without generative techniques, would be too cost prohibitive for smaller studios.
Re: (Score:2)
That's what I was thinking. We've had things like map generation for a long time, AI for that would be the same thing, just grander.
Now, making it so that all the guards in a city are different, rather than having a hundred or so clones of each other? That would be interesting. Especially if they can do conversations that don't sound like "pick one of three random responses" every time.
But I could also see things like designing a city, down to building layouts.
For strategy games, smarter, but not too sma
Re: (Score:2)
I think the fundamental problem is not the quality or detail of the generation but simple supply and demand.
Video games are *already* hundred-hour plus endeavors overflowing with content that 99% of players never see.
Do you really have so much time on your hands that you're going to roam around an endless digital world that just keeps generating itself the further you go? Sure, it'll be fun for awhile, but at some point you have a *a life* to get back to. The very fact that highly-detailed, context-sensitiv
Re: (Score:2)
As a minecraft player, that doesn't bother me at all. And as a minecraft modder, one of my most popular mods simply adds a small number of activities to the villagers. Players enjoy seeing the villagers doing those things.
And randomly generated dungeons and buildings are a long tradition in many games.
But supporting your point, Vault Hunters combines hand crafted 125 meter cubes randomly into a procedurally generated world. It's popular because of the human touch in the cubes.
Re: (Score:2)
Exactly, essentially you in a way connect to the people who made that which you are experiencing, be it a video game or a picture or a book. Connecting with a statistical digest of all of humanity in digital form isn't that.
Re: (Score:2)
Thank you, that's a far more eloquent way of phrasing it.
My fear is that as other's have noted, shipping off the world building (be it map/terrain or even NPC dialog or interactions) to AI/algorithmic generation is going to be cheaper for the dev studios, incur shorter dev times, eliminate or reduce bias yadda yadda
So OF COURSE developers (particularly the AAA shops) will gravitate towards using it -- but the product (verbiage quite intentional here) will be soulless and devoid of any sort of creative spark
Re: (Score:2)
What cheapens No Man's Sky is that it's highly derivative and not interesting, not how it was created. It was not a fun game.
AI can potentially fix this by adding a human element to these endless generated expanses, by adding uniqueness in excess of, "generate 1-3 baddies every 2-3 areas of character_level(player_level - 3) or such.
Re: (Score:2)
I think the difference is that if you know a human author toiled away on a book for months or even years, that is an investment that commands respect. You appreciate the author's vision and dedication. It becomes a work of art.
Meanwhile, if the whole thing is auto-generated by a computer, it's mere entertainment. Maybe it can be beautiful in the way a mountain vista or sunset can be, but with no intent behind it's just something pretty to look at.
Re: (Score:2)
Ah, you enjoy making people toil for your gratification. That's certainly not uncommon.
Many other people just like human designed stuff because it's engaging compared to procedurally generated. Make good auto-generated stuff and they won't care. In fact, they'll like it better because it's much, much cheaper.
Re: (Score:2)
Re: (Score:2)
This is true, any nerd programmer with a game idea has historically had one huge issue: graphics. You can program a great game engine, but it takes a lot of time and effort to make those graphics. Now you can contract it out, but it's still expensive and time consuming to coordinate. If it can just be generated, that will enable a whole new generation of games that aren't held back by "creative types."
Re: (Score:2)
One of the best uses for generative AI will be in video games. And it'll enable indie developers to build much more. Indie developers who, at least initially, will present much higher quality games without all the usual bullshit that comes with AAA studios.
I don't think so.
First off, Indie games are pretty damn good already... Well the good ones are, there are unfortunately a load of dross games because making a generic mobile game is cheap. To be specific, they're usually better in terms of gameplay, story or novel ideas, graphics are usually not bad but I can forgive less than stellar graphics if the rest of the game keeps me interested. Doubly so if the devs are active with their players (although with some of the toxicity around, I can't blame them if
More efficient... (Score:1)
at copying last year's game, hooray!
Well, can't say they're wrong, ML is really good at copying things and changing the name.
Re: (Score:2)
at copying last year's game, hooray! Well, can't say they're wrong, ML is really good at copying things and changing the name.
That's exactly what I figure will happen. Every blade of grass we're accustomed to seeing hand-crafted in glorious 4k will soon be an 800x800 AI-generated blending of every other blade of grass ever uploaded to the Internet, up-scaled back to 4k by another AI trained on the same models. Then the models will start training off each other's output, and every blade of grass we end up seeing will be the same lifeless lump of recursive blandness. Which admittedly is quite efficient.
Re: (Score:2)
...every blade of grass we end up seeing will be the same lifeless lump of recursive blandness. Which admittedly is quite efficient.
Sounds like we should be storing texture maps in a shared directory for our games to use, and if developers chose to ship games with their own textures the installation media handling company (EG: Steam) can compare the texture maps of both to see if the new game texture references should be changed to use existing texture library entries or the texture library updated and the new textures added as a new entry or perhaps even overwrite as a better entry.
We could call it a Texture Link Library (TLL) perhaps
Just like taking laxatives (Score:2)
Dumb CEO shit (Score:2)
Re: (Score:2)
I'd buy that for a dollar!
30% more efficient? Wishful thinking! (Score:2)
No, this means that using something like that in house is that they'll use 30% less original materials.
That first party games they make will become 30% more recycled stuff from before.
Remember these models rely on what you feed them to be creative. Nothing original comes out of them. If used for games then anything new or different or interesting will be an effect of other workers doing that. Not what the AI is.
Whatever savings they get will certainly not be passed on to the workers or the public who buys.
Re: 30% more efficient? Wishful thinking! (Score:2)
Games will have massive install bases with that much detail. They should be 30% cheaper.
I can better imagine games using LLMs for NPC dialogue. This will require online access and thus reduce piracy.
Re: (Score:2)
That is MBA-thinking: They can make producing one thing 30% cheaper. Which is probably realistic. What they overlook, since MBAs do not really understand anything, is that other things will get more expensive, because quality drops. MBAs do not understand quality at all.
And obviously, this is a straw-fire. At the moment, for example, the defects and general sameness of AI generated pictures gets accepted. But in a few years, everybody will immediately recognize them as "AI crap". If you have fired all your
Someone watched the last season of West World (Score:2)
I can imagine a gaming company CEO watching that scene in West World where the main character is telling the machine what she wants in the game as assets and salivating over the possibilities.
I look forward to more soulless cash-grabs in the gaming industry. Because what we need is a quicker way to update old games and concepts with slightly better graphics, not better gaming experiences overall.
AI-generated positions sounds scary (Score:2)
Considering AI's "understanding" of human anatomy, you'd probably injure yourself attempting to pull off AI-generated sex positions.
Re: (Score:2)
Considering AI's "understanding" of human anatomy, you'd probably injure yourself attempting to pull off AI-generated sex positions.
You'd have to injure yourself to get the fourth joint in the arms.
Re: (Score:2)
Hehehe, waiting for the first personal injury lawsuit claiming "that AI generated porn movie made us think it could be done!".
Potential (Score:2)
The potential isn't for saving time through "efficiency". That way of thinking demonstrates the thought that what they're providing today is sufficient. It isn't.
Most games produced today, especially by EA and other big studios, feels very uninspired, with little direction or focus. They often feel incomplete, with a lack of attention given to maturity or completeness.
If they want to be effective, they need to increase the depth of the game. The in-game notes left laying around, computers that get hacked, r
Re: (Score:2)
Incidentally, same with the movie industry.
Good entertainment needs _ideas_ and _stories_, not cheaper technology. Cheaper technology is a plus, but it is only one piece of the puzzle.
Re: (Score:2)
I'm not playing "AAA" games these days, mostly going for smaller studio stuff, but the last game I played as a person (IE not the 4X I'm playing now), had not only flushable toilets, but working showers.
But then, with Unity or a number of other engines, you can probably just buy a toilet asset and get flushing action included.
Months to days... (Score:2)
Suuure. Having ideas and telling a story is soooo yesterday. Yes, some things will get easier, but making good games will still be hard.
Ok But (Score:2)
30% of 0 efficiency is still 0. Not to mention EA Sports does not do any work year to year. Every year it's the same games (bad), same mechanics (bad), same graphics (really bad), and new rosters - which should be DLC updates instead of entirely new games.
30% more efficient is EA speak for 30% less staff (Score:2)
Do we need people ? (Score:2)
CEOs must be stressing about their jobs, after all, any AI could probably generate similar stock hyping statements followed by the inevitable rationale based on uncontrollable externalities for why the improvements never materialized a year later. But, wait for it..., (rotates Magic-8-ball) Quantum computing will deliver 30% more shareholder value real soon, especially when combined with crypto.
Think how quickly AI will get developed as it captures C-suite level salaries and bonuses
Dear AI (Score:2)
Dear AI, please give me an idea for a video game that will be massively popular and make us a lot of money.
Make the same sports game but update the names of the players and teams to be current.
Wow this thing is genius!
Re: (Score:2)
The way AI could be ethical and efficient (Score:1)
You have to think the thing that EA thinks of as "more efficient" is to get rid of artists, and use generative AI to just spit out art.
However - they way it could be used that actually respects artists and original content is, to have personal AI's attached to each artist hired, that is trained to their body of work - the instruct it what to produce along with doing any number of original works as well, also added to the AI...
Now you really are more efficient, because one artist can produce a lot more work