Is AI good or bad for video games? Will these tools usher in new amazing games, or be yet another disappointment to gamers already feeling heavily burned by major gaming studios?
Artificial Intelligence (AI) has been a massive disruptor both in and outside of the Tech industry, and gaming hasn’t been immune to these changes. The job market for skilled programmers, artists, designers, and writers in the video game industry always has some degree of flux, but the introduction of AI has potential to become one of the most important advancements in game development.
But is this a good thing for those with jobs in the video game industry? Is it a good thing for gamers?
Well the messaging on that one is going to be a bit mixed because like many things involving AI, there’s a lot to look at and consider, and separating facts from marketing speak is always easier said then done.
AI isn’t going a way, and for better or worse it’s only a matter of time until it is used in some way by every major studio in the development of games from free to play mobile games to major Triple A studios looking to make ever more expansive and interactive worlds.
So will AI help usher in a veritable Golden Age of gaming, helping developers create better worlds, characters, and experiences for gamers – or is it a Trojan Horse that promises a lot but will actually result in yet more broken games and terrible experiences for gamers who are sick and tired of dropping a lot of money on a new game that is junk?
What does AI mean for those working in the video game industry and are their jobs at risk of being automated or eliminated?
I’m going to dive into these questions in-depth, talking to experts in the gaming industry and AI industry, as well as many affected individuals, to try to paint a more complete picture (at least as much of one as we can have at this point) of what the introduction of Artificial Intelligence tools to game design means for the industry, the workers, and the consumers.
Is AI Good or Bad for Video Games?
AI in and of itself is neither. There have been various programming advancements over the years that already do the things that AI companies that clearly aren’t loaded with gamers claim that AI could do for gaming. Does that mean the AI threat to gaming is overblown? In some cases, yes. In other cases there are potential ways that AI could do really interesting things that move games forward in an interesting way.
Like so many things the answer to this question will hinge on some other potential issues such as:
- How well will AI continue to advance and develop as a tool?
- Can further developments improve the glaring current limitations of AI?
- Will AI be used as a tool to help overworked developers, writers, and other workers – or will bean counters try to use it to replace these essential workers?
- Will large studios make the biggest developments in video games the next decade…or will the torch continue to be carried by indie developers?
Several gaming companies, notably the large AAA studios, have hailed AI as a way to take care of less essential but heavily time-consuming tasks allowing more workers to spend more time and energy on the more interesting and essential parts of games, implying that by doing so they will improve the overall quality of the game.
The big question is: Do you believe this?
Based on the many scandals with huge companies and CEOs in the gaming industry alone the past few years, it’s understandable if you seriously doubt their intentions. Add in the number of games that were released in unforgivably broken states like Cyberpunk 2077 or Fallout 76 and the concern for many workers in the video game industry is very real.
If they don’t care about releasing outright broken games – then why would they care if they broke a game by using AI in place of human coders, writers, and artists?
There are some undeniably interesting potential applications of AI, but there are just as many questions about its use, as well.
There are many people pointing out that not only is AI and the way it’s being used not necessarily good, but there’s a good case that the so-called AI Revolution is rotten to the core.
AI Has Legal Gray Areas
One thing that Valve, the company that owns and runs Steam, is deleting games that use AI content citing (rightfully) legal concerns because since AI draws from tons of sources and credits none of them, that this opens them up to copyright and plagiarism issues.
More than once AI created content, when tested and put into a plagiarism checker, didn’t pass. Meaning it changed text so little that it was easy to track down the original source where the tool ripped the text from whether it was Jasper AI, Chat GPT, or any of the others. None of them are 100% immune to this potential problem.
Even beyond the straight up copyright violations via plagiarism there are multiple suits going on with AI having used copyrighted works to create the knowledge based that AI uses…without compensation or permission.
It’s hard to see the genie being put back into the bottle, but that doesn’t mean we’re through with legal challenges and entanglement, and that will matter when it comes to how AI affects the gaming industry and those who work within it.
These will need to be ironed out before the future of AI in gaming will be clear.
AI Isn’t What You Think It Is
AI is an advanced data aggregator and parser, not a fluently thinking and reasoning machine. Why is this important? Because most people are assuming that AI can reason and create – it can’t. Certainly heads of AI companies are not going to correct this assumption when it hurts their business, but it is crucially important to understand.
AI tools are great at collecting tons of existing data, training up, and spitting out something bland, middle of the road, and reasonable to requests based on the data available. It can NOT create new ideas from scratch, new concepts, or “brain storm” in the way that people would conventionally understand those actions.
This is the difference between statistical aggregation versus actual reasoning, and that matters because while AI can do some pretty impressive things within a box – those things are within a box.
Keeping in mind that AI must learn from existing behaviors, this also brings up concern of the many AIs from top companies that became very racist and sexist very quickly like this one, or this one, or this one. Or this modern example of a ChatGPT bot playing Runescape becoming more and more aggressive against human players as shown in this YouTube video.
How Could AI Be Great For Gaming?
There’s a reason AI tools are seen as exciting. They offer new possibilities for dealing with work flow, brainstorming ideas, and even some practical use with NPC characters in massive RPGs that tend to have many NPCs – but too many to realistically give them all interesting lines or dialogue.
“Fully Living” NPCs
Having every NPC remember interactions with you, having random interactions with other NPCs, and changing dialogue or interactions based on that would be pretty amazing. While our article on why are games like Stardew Valley so popular praised that, and other, cozy games, a major point was the fact that at a certain point there was nothing new with the NPCs.
In theory, AI would be able to change that, keeping a degree of originality or freshness that currently isn’t generally possible. This would be a neat addition as it would also make each game very different. Imagine a Fallout New Vegas where NPCs didn’t just change interactions based on your stats or overall karma, but also based on literal individual interactions with others or with them in the past.
A paper released told about a project between Google engineers and many Stanford professors to create a “Interactive Simulacra of Human Behavior” – i.e. a living world of NPCs who are all growing, “living” their own lives, and developing accordingly. Matthew Berman does a good job showing the excitement of what this could mean in his video.
And this undeniably could be an awesome boost to giant worlds.
Great Brainstorming Tool for Writers
While I’m extremely dubious of replacing writers with AI, giving AI to writers who can use it to brainstorm past writer’s block is a legitimately good use. Come up with 200 neutral NPC dialogue options. Seriously, try doing that off the top of your head for greeting a new player. It’s rough. Being able to put in a prompt to get 20+ ideas, and use one of them or have that fire off ideas in the writers’ heads to get them going again is an actual excellent use of this tool as an assistant tool.
Uprooting “The Meta” in Simple Game Mechanics
There are some games where it doesn’t take long for player (looking at you, Super Auto Pets) where a certain method of play becomes the default because it is the most efficient and is virtually unbeatable the majority of the time because of the current setup. Introducing AI could help to mix and match the parameters to open it up so there are always multiple paths, builds, or metas that can work – making it more interesting instead of funneling all players towards one build.
Make Worlds Feel More Interactive
In theory this could make video game worlds feel more interactive. Do animals react to heavier foot traffic around villages? Do enemies raid NPC villages? Do certain places grow based on what you sell and improve while others stagnate? These are all possibilities though I’ll point out that all of this type of programming is already viable through other means.
More Focus on Important Aspects on a Tight Deadline
If an internal AI tool could be trained to fix minor coding glitches, identify potential conflicts in code, or coming up with more low level NPC dialogue, that can all be useful for teams that are stretched to the limit trying to fix larger problems before a general release. This would be great if it cut down on the serious glitches, crashes, or problems that have plagued many recent releases.
If used well, it’s not out of the question that AI could be a good tool for helping to clean up many of these annoying glitches more quickly and easily so the human workers can fix those game-breaking bugs BEFORE the game would be released, which would be a nice change of pace in the minds of many gamers.
This could also make the job slightly less stressful for workers – who know when a bad game is coming out and the blowback that will come as a result.
How Could AI Be Terrible For Gaming?
There are several ways that Artificial Intelligence could be terrible for gaming, and it’s safe to say that there are plenty of fears around this not only by the players but also video game companies who are dually afraid of what AI means for them and falling behind the AI surge.
So what are some of the obvious potential red flags?
Have You Seen AI Learn Human Habits?
We’ve already shared some of the many egregious examples of how AI loves to pick up the absolute worst habits and beliefs of people. Until this is programmed out (and can that really even be done when trying to push the limits of generative AI?) that will be a problem. Not to mention the almost certain appearance of mods encouraging this behavior that will come out from people who just want to watch the world burn.
Unfixable Errors Are Almost A Certainty
There are certain bugs, errors, or problems that are hard to fix in the best of times because there’s so much programming that goes into a game from hundreds if not thousands of programmers. But if problems arise because of AI learning, what if the AI programming refuses to be “re-taught?” For online games like MMOs or seasonal passes, how could you stop the AI from doubling down on the bad influences that caused those problems to begin with?
That means there’s some serious potential for errors from AI learning and interaction that will only become more and more cemented – and eventually even unfixable without changing the entire AI input engine…which is likely impossible for a large scale game.
Will Video Games Become Impossible?
Computers can think and react faster than people can. That’s just a fact. One of the reasons that the Dark Souls games are so hard is excellent programming with Enemy AI (long used in video games) that make quick decisions based on likely player mistakes or set within patterns, but it’s still a pattern.
In Dark Souls it’s hard to deal with the pattern because the actions are taken so quickly and it is very unforgiven of player mistakes – but it’s not reactive to independent strategies put forward by players.
If adaptive AI decides it wants to always win – what if it learns to the point of using superhuman reflexes to do just that: to win every time? How do you prevent that new AI from overriding the old systematic enemy AI and just reading the player input and reacting to it?
At some point the problem has potential to pop up – and it has potential to not be great.
Some Studios Will Replace Writers, Artists, and Programmers
This isn’t just a bad because of job loss thing – this will unquestionably lead to worse video games. The A.I. will continue to become more and more standardized, no new ideas will emerge, and fewer human workers will be there to fix wonky code, input new ideas, or put in the heart that actually makes a game truly stick out in a good way.
You Will Own Nothing – But Have to Keep On Paying
The fact that major gaming companies are looking to push A.I. into board games is a way to say that board games require subscriptions…and I don’t know any board game fans who are happy with that idea at all. Many old school video gamers already lament the fact that being banned from an account pass or a platform means games you bought are gone forever – you’re really just renting access in good faith as opposed to actually owning the game.
A proliferation of A.I. is moving in a direction to make this the norm, and it’s not a great feeling for gamers.
Is AI Really That Big A Change To Gaming?
While there have been heavy advances in AI in the past couple years, it’s important to note that AI in games is nothing new. The original “Shogun: Total War” all the way back in 2000 bragged about adaptive computer intelligence that made the game more challenging than other strategy games out on the market.
This clearly is programming and not AI as we currently see it, but it shows that sometimes in gaming the two terms can be used (incorrectly) interchangeably during discussions in the gaming niche.
Generally gaming AI is split into:
- Deterministic AI – This is predictable in-game programming. Enemies only reacting with a certain sightline, walking along pre-determined routes and the like.
- Nondeterministic AI – This is adaptive AI that isn’t set in action but reacts based on player interactions with it.
- Generative AI – This is the Chat GPT tool and newer style of AI that is getting attention and has some decent potential for certain parts of game programming but which is also introducing the potential problems this article is discussing in-depth.
This article by inworld does a pretty great job of outlining the various types of programming and AI in video games and is worth a read.
It’s also worth noting there are types of programming that are sometimes referred to as AI by gamers or designers as sort of a “game AI” but it actually isn’t technically AI but just a complex bit of programming that can appear to users to be so amazing as to be easily mistaken for intelligence.
A prime example of this would be procedural generation – which was made huge by No Man’s Sky to create a game that has a seemingly infinite universe to explore. This is not actual AI, but it is a tool in the game designer’s arsenal that really opens up their ability to create vast universes and games that seem to fully interact with the player while they’re actually being built up right before creation/interaction.
So if someone asks: Is A.I. really that big a deal or change when it comes to video game design and programming?
The response to this might be: Maybe, maybe not.
Will AI Actually Make Better NPCs?
The biggest seen application for a Chat GPT style of A.I. in games would be for writing less important dialogue or making NPCs more responsive – but is this really something that would be new?
Baldur’s Gate 3 showed that an incredible amount of responsive dialogue can be programmed in the traditional way, and it creates an incredibly responsive world that adds to the replayability and feeling of living in a real vibrant world. In theory, this is fully doable the traditional way although admittedly this took a ton of work – and more than many teams of programmers will be able to pull off on a normal design schedule.
If a company trained up their own internal AI tool and input hundreds of lines of dialogue for every type of an NPC, in theory this could make at least a good surface level variation that makes NPCs seem much more like real acquaintances than basic programming with a very limited number of lines.
However, either programmers will still be needed to customize the many potential lines of NPC dialogue to shift or customize them based on the game…or eventually the games are all going to seem dull and familiar because every single game a game studio comes out with will be pulling lines of dialogue from.
There’s also the chance no one bothers to change those dialogues – meaning dialogue from a Medieval RPG is used in a futuristic Science Fiction RPG and would just not fit. Based on how many games have released from Triple A studios broken in recent years, this is a legitimate concern of gamers and game workers alike.
AI NPCs Are Often Jerks
The many stories of AIs becoming jerks aren’t isolated. This actually seems like the norm unless a major amount of restrictive programming comes in. Even with restrictions put in, they still tend to be rude trolls at some point – as the video below helps to demonstrate.
It’s also worth noting that if all NPCs are given learning AI and keep interacting with players online, you can’t cut off the same influences that have turned many other AI NPCs into jerks.
How Will AI Affect Jobs in the Video Game Industry?
There’s a wide gulf between what upper management of big companies in the video game industry are saying and what many in the industry expect to happen, and that gulf makes it hard to determine the exact level of impact that the explosion of artificial intelligence on the gaming industry and jobs within the industry, but its hard to believe there won’t be a major effect.
One of the best ways to start with understanding the differing opinions on how gaming jobs will be affected by A.I. is to look at the official corporate line on A.I. from video game companies versus what most gamers are “hearing.”
The Official Corporate Line: AI like ChatGPT can take care of unimportant but necessary dialogue like from minor NPCs found throughout RPGs. This means writers can work on more important story points.
While I don’t have a problem with this in concept, in reality there are a couple problems. One can be figured out by talking to any long-time writer who has messed around with AI writers. They can write basic and bland in an impressive way, but at the end of the day they are a glorified article spinner which means that dialogue it produces will be bland.
But if the dialogue is unimportant that shouldn’t be a big deal, right?
If it’s that unimportant, why is it in the game?
One of the prime examples of things that set truly incredible games apart from the flops is minding the details, and the runaway success of Baldur’s Gate 3 shows that attention to details matter. In that game even the most basic NPC dialogue changes based on your level, class, reputation, what other classes are in your party, and previous interactions. All programmed in to create a stunning experience for individual players.
Doesn’t sound that unimportant to me – nor to the millions of players who have enjoyed this game. Like Elden Ring, it’s another game that has millions of sales, unexpected success, and then gets mocked by Triple A game studios claiming their terrible games setting unrealistic expectations…and those really obvious lies are coming from the people who control most of the gaming industry and will be controlling how AI is used in gaming.
This is important to keep in mind.
What Gamers and Game Workers Hear (and Believe): We’re going to experiment with replacing as many programmers and writers with AI as possible because as long as the fired workers salaries is less than the money lost from people not buying the game, we’re good with it.
Considering the many fiascos that have come out with broken games, refusal of refunds, and microtransaction pummelings on players, it’s not hard to see why the trust would be so low.
Sort of like for D&D players who were assured by WOTC that no one was working on an AI DM and that those were lies, and less than a year later you have some questionable if not outright conflicting information releases about acquisitions suggesting a move to AI integration into the VTT, the massive AI Art scandal showing WOTC employees were already replacing artists with AI generated images, videos on AI DMs coming in 2024, and those have kicked off discussions on if players are being groomed for DM-less games.
While that’s D&D and D&D online versus video games – it is a clear example of many of how what the official word from public relations is versus what is actually happening in reality.
That example of what they’re doing vs what they’re saying is why many programmers, artists, and writers in the video game space are extremely nervous about the future role of AI in their industry – and frankly the workers have every right to be nervous based on these other instances of public relations saying one thing, and a whole lot of actions suggesting the complete opposite is true.
Hiring Freezes – Will They Thaw?
A combination of multiple big title busts causing major losses, widespread workforce cuts in the tech sector, and the explosion of A.I. have all contributed to major hiring freezes in many tech-related industries, including gaming. One of the major questions is whether or not these hiring freezes will thaw at any point.
Reports from multiple independent anonymous sources with companies have confirmed that many, many companies are experimenting with replacing any potential hires with A.I. first, seeing if they can walk away from hiring new talent and focusing on develop further automation instead.
So will the hiring freezes in gaming thaw…or is it going to be a struggle to overcome anything execs can be convinced that A.I. can do “good enough” at?
That’s going to be a major question that may take a few years to play out.
AI: A Helpful Tool or Worker Replacement?
We discussed this a little bit earlier but it doesn’t change the fact that one of the major forces deciding just how much of an effect A.I. will actually have on the job market in the video game industry will be: Is A.I. a tool to help developers, writers, and artists – or is it going to be used to replace them?
That is going to be a major driver that changes how A.I. will be perceived overall when it comes to the gaming industry, but even if Triple A studios go heavily into replacing staff, there are reasons why this might be a temporary hit to the jobs market and not a long term trend.
Consider the following potential scenarios:
- Triple A studios adopt A.I. to replace workers but most indie studios don’t
- Triple A studios see limitations of A.I. and curb new hiring but don’t replace workers
- Both Triple A and indie studios use A.I. to replace workers
- Neither Triple A or indie studios use A.I. to replace workers
All of those scenarios are potentially in play, and it’s worth noting that while a AAA studio often has hundreds or even thousands of workers, they also release far fewer titles than indie studios do now. From solo developers to those who are 50 workers or less, indie studios are putting out tens of thousands of games annually – meaning there is still an open pathway to work through indies.
Would mass lay-offs of Triple A studios increase competition? Sure – but this can also result in indie studios of 50 or so people forming into new companies. This has happened multiple times in the past decade and a half and some of those studios are quite successful and well-known now.
Will A.I. create turbulence in the job markets for gaming? Yes, almost certainly. That said, there’s no guarantee that even in an apocalyptic scenario with the big studios that more jobs wouldn’t open up in the long run.
It’s also worth noting that if primarily generated A.I. games (which as of now still can’t be sold on Steam – the largest marketplace for gamers) fall flat consistently (and at this point they likely would as players don’t like being gaslighted by nonsensical AI) studios would be forced to hire back more human workers or face going under – meaning studios that made better games that made massive amounts of money and received critical acclaim would grow to take their places.
At least in the short-term that will act as a hedge against massive A.I. upheaval, but long-term? It’s hard to say if that will be true 10 years from now.
What’s This Mean for the Future of Jobs in the Video Game Industry?
Will there be fewer jobs for certain workers in major video game companies because of AI? In the short-term the answer is almost certainly yes. For the long-term, time will tell.
There is even a chance that A.I. will be irrelevant in a few years when it comes to gaming. Not a trend I would bet on, but let’s look at some of the other “game changers” in gaming – those that were lauded and failed, those that changed things for a time, and those that will just sort of a side show before the rest of the industry continued to move on.
Remember NFTs?
Yeah. NFTs, the crypto-bros’ dream, a tech “advance” that made no sense to the majority of us in and out of gaming, but Triple A studios were all over that to create “games to make money” and create another way to monetize video game.
Ubisoft at one point even went so far as to say NFTs were the future of gaming, but surprise, surprise, video game fans want to play video games for fun, not have a boring, repetitive “game” that is a thinly veiled advertisement rife with upsells hammering away at their wallet during a recreational activity.
Those who were dubious about NFTs were right as around two years after they broke onto the scene, they’re all pretty much worthless now. While A.I. definitely has more utility than NFTs or cryptocurrency, it’s a prime example of being cautious about jumping too all-in on a current trendy tech.
What Historical Jumps Actually Moved the Gaming World?
One of the best ways to have a look at whether or not AI will have a massive long-term effect is to look at some of the other major jumps or shifts in gaming and whether or not those were long-term changes to the gaming industry, or if they had their effect and then everything moved on. This can give us major insights into the various ways artificial intelligence may or may not have a big effect long-term on the gaming market.
Modding
This is arguably the most important move that ever created a permanent jump ahead in gaming and once again – this came from the development of independent programmers who had the blessing of some early open-minded indie game developers.
Modding is one of the biggest alterations to gaming ever. For indie games they are often a great way to continue to keep a game fresh and add new content or wrinkles, creating new playing experiences. Mods like Stardew Valley Expanded are even famous because of just how much new content they add to a classic game.
While there is legitimate and justified criticism that many Triple A studios rely on modders to repair their broken games for free, it’s better than simply having a broken game and no refund.
Nexusmods is the most popular of many mod homes, and has over half a million game mods covering over 2,500 games.
This changed the state of gaming and will likely remain one of the all-time great influences on gaming.
Nintendo Wii
The Nintendo Wii was huge when it came out. Not only was this a system that was so different from others before it, and offered a level of real life interactivity from Wii Sports that no one had seen before, but the fun yet simple games that played to the unique controllers made this one of the first game systems that was attractive to many older players who would traditionally be non-gamers.
While really cool – this was relatively short-lived. Nintendo Wii is still around and there are many games for it, but it feels very niche at this point as even the handheld Switch feels more “mainstream” to Nintendo development than the Wii does.
Nothing wrong with that – but it wasn’t the permanent industry changer that many were imagining.
Sony PlayStation 2
There was a reason the PlayStation 2 was so popular for so long, and a major reason is that when it comes to graphics, processing power, voice acting, and a variety of other things people consider mainstream now, there was arguably never a bigger jump than N64 to PlayStation 2. The change from the previous generation to the newest one was never so pronounced, blowing the door off of what was possible in video games.
There are PS2 games that hold up 20 years later as far as being fun, challenging, and great to watch. You can’t say that about most other systems, and while undeniably newer game systems are stronger, faster, etc, they’re building off of the various options and history of games that already exist.
Super Nintendos were still popular when the PS2 came out – to further emphasize how big a jump this was. Whether or not this type of jump will happen again is debatable – and arguably the only jump that could be that big has already happened and hasn’t been the game-changer that most of us expected it to be.
Speaking of which…
VR Gaming
Virtual Reality seemed to make giant leaps forward, and there’s little question that there are some very cool things still going on in the realm of VR gaming…but does it feel like there’s less talk about it now? You’re not imaging things.
According to Google Trends interest in VR Gaming peaked in 2019. It has slowly been waning in the years since. There hasn’t been a huge drop since where it stabilized after 2019, and although there have been peaks and valleys, despite the increase in quality software, games, and VR headset gear, the interest is slowly waning away.
That shows even the most remarkable of technologies, that easily fits in the gaming industry, isn’t necessarily a hit. While I wouldn’t say VR is dead by any stretch, it did not completely dominate the way you would expect.
So What Will The Long-Term Affect of AI Be on the Gaming Industry?
I want to be optimistic, but the truth is that this is likely going to be a mixed bag. While projecting how a new technology is going to affect in industry is always a bit tricky, there are several likely outcomes at least short-term (next 10 years) that will affect gamers and gaming industry workers, as well.
Some of the most likely affects of AI on video games:
Expect More Terribly Bland and Boring Triple A Games
Unfortunately, we’ve seen the past few years that Triple A studios, even formerly beloved ones, like EA, Blizzard, and Bethesda are more than willing to cut corners to save money, to the point of putting out unfinished (see YouTuber Upper Echelon’s Cyberpunk 2077 outsourcing disaster videos like this and this for prime examples) or broken games to get short-term earnings off marketing hype. Most gaming companies are run by suits and executives who don’t play games.
And it shows. With that in mind, it’s hard to believe gaming companies run by spreadsheet warriors and accountants versus gamers will develop AI tools for writing, art, or basic programming without looking to eventually fire as many writers, artists, and programmers as they can.
As we discussed earlier, AI tools aren’t truly thinking or reasoning – that’s not how they work. These tools are going to keep drawing from the same prompts, the same database, the same seeded information. They also look at amount of repetition for how likely or strong the information is.
Meaning all AI tools for dialogue, for prompts, for quest ideas, come from the same pool. No new ideas, no originality. The longer the AI tools work from the same pool, the more homogenous it gets as the more unique options that don’t occur as often get used less than others, causing AI to further push them out of the periphery.
Expect a lot of bland and boring stuff from dialogue to quest mechanics, to player interactions. Even if companies don’t mass fire these workers they believe can be replaced by A.I. they’ll cut back, so all those workers who were promised to focus on more important dialogue or storylines will no doubt be overworked and understaffed – and that will continue to push a negative downward spiral on quality.
Expect More Triple A Games to Look Like Bland Repetitive Clones of One Another
Not only will AI tools developed for writing, art, or programming draw from the same resources over and over to make some new iteration of that existing information (which introduces repetition problems in and of itself) but it’s not like an AI tool used by a major gaming studio gets new inputs for a new game.
They don’t. They still pull data via whatever setup was used for a previous game. Not only that, but if the tool works like most AI, those bland middle of the road options for dialogue or art design made from AI the first time will be put into the pool of data it draws from, further homogenizing the information that it draws from.
That means the more it gets used, the more “same” the majority of brought back responses will be.
Writers who used early AI writers can attest to the consistent issues with:
- Repetition of words, phrases, formats, “mini-conclusions,” and recurring topics out of place from normal construction
- Middle of the road writing that is pretty but shallow
- Forgettable text that will not be long remembered
- Making up facts
- Making stupid mistakes humans wouldn’t (like naming filet mignon as a spice for steak)
- Making arguments that don’t make logical sense or are outright nonsensical
There’s also been multiple instances where those testing AI commented that in some cases the more they used the AI, the more repetitive and less creative the tool became.
Imagine it like a YouTube algorithm after you watch 60 videos on one topic by one creator completely different from everything else you used to watch. You won’t be able to find those old video topics as almost all your recommendations will be around the new rabbit hole. If you’re not careful with A.I. tools the same happens.
The more creative ideas get cut out as the most commonly occurring ones, including those created by A.I. are put into the large group of text to draw from, the more bland and middle of the road each idea, piece of text, or output will be. A huge shift to AI will add all these problems to major studio games – especially in genres where this matters a lot like RPGs.
Expect Breakout Hits to (Continue to) Come from Indie Studios
The indie renaissance is almost certain to continue. Indie games need to rely on new ideas, fun gameplay, and are almost always run by gamers wanting to scratch their itch for a certain type of game. Any A.I. used with these games are going to be tools to help out writers or designers – not replacements.
That will allow them to stick out in a gaming market that has every chance of big studio games becoming more run-of-the-mill, similar, or clones of each other, which means the massive W’s that seem to mostly be for small developers and indie studios in recent years is a trend that is almost certain to continue.
Expect More Broken Big Studio Games
If there is one light of hope to this, it’s that after seeing the stunning comeback of No Man’s Sky, and the incredible disaster that were games like Fallout 76, Anthem, and Cyberpunk 2077, not to mention a slew of live gaming pass games that lost millions and then were shut down completely, that the stock value drops might cause some studios to start making better long-term decisions.
But until we see that consistently, there’s every reason to think that Triple A gaming studios run by Wall Street Bros will look to use imperfect tools beyond their capacity to cut payroll and make end of quarter numbers look good…to the detriment of a game that eventually gets released in bad shape.
And if A.I. is heavily involved in some of these games don’t forget that fixing errors may not be possible especially on online games where you can’t shut off, filter, or control the inputs to A.I. learning. Very bad news for gamers potentially, especially casual gamers who don’t keep up with the industry news as much and so don’t know to be wary.
Expect The A Tumultuous Time for Workers in the Gaming Industry
There will almost certainly be lay-offs. There have already been hiring freezes because of testing A.I. tools. That said, there will also be more opportunities. New game studios will form, long-time veterans in game design inexplicably let go in layoffs will create new studios or spearhead new projects in existing ones, and there will be work options.
Some may choose to go their own way, looking for funding to develop their own game. There will be options out there, though as with any major transition period, there will be built-in challenges and uncertainty.
AI in Gaming: Concluding Thoughts
There’s no question that how A.I. tools continue to grow, change, or adapt will influence just how big an effect they have when it comes to the future of both video games and video game jobs. These are going to be interesting times, and there are likely to be both good and bad things that come from it.
Will AI be good or bad for video games? It appears that only time will tell, but there are few questions that major effects will be felt throughout the industry, and some of those tremors are already here.
Proud to embrace the locally created moniker of “Corrupt Overlord” from one of the all time great Lords of Waterdeep runs, Shane is one member of the Assorted Meeples crew and will be hard at work creating awesome content for the website. He is a long-time player of board games, one time semi-professional poker player, and tends to run to the quirky or RPG side of things when it comes to playing video games. He loves tabletop roleplaying systems like Dungeons & Dragons, Pathfinder, Werewolf, Fate, and others, and not only has been a player but has run games as DM for years. You can find his other work in publications like Level Skip or Hobby Lark.