Square Enix says it wants generative AI to be doing 70% of its QA and debugging by the end of 2027
https://www.videogameschronicle.com/news/square-enix-says-it-wants-generative-ai-to-be-doing-70-of-its-qa-and-debugging-by-the-end-of-2027/Open linkView original on lemmy.world367
Comments108
Ew, sounds like a great reason to not buy any Square Enix games...
Not even from an ethically standpoint. Color me shocked if these games are like, playable
Exactly, as I don't expect QA done by something that can't think or feel to know what actually needs to be fixed. AI is a hallucination engine that just agrees rather than points out issues, in some cases it might call attention to non-issues and let critical bugs slip by. The ethical issues are still significant and play into the reason why I would refuse to buy any more Square Enix games going forward. I don't trust them to walk this back, they are high on the AI lie. Human made games with humans handling the QA are the only games that I want.
That is a very small part of QA's responsibility. Mostly it is about testing and identifying bugs that get triaged by management. The person running the tests is NOT responsible for deciding what can and can't ship.
And, in that regard... this is actually a REALLY good use of "AI" (not so much generative). Imagine something like the old "A star algorithm plays mario" where it is about finding different paths to accomplish the same goal (e.g. a quest) and immediately having a lot of exactly what steps led to the anomaly for the purposes of building a reproducer.
Which actually DOES feel like a really good use case.... at the cost of massive computational costs (so.. "AI").
That said: it also has all of the usual labor implications. But from a purely technical "make the best games" standpoint? Managers overseeing a rack that is running through the games 24/7 for bugs that they can then review and prioritize seems like a REALLY good move.
They're already not paying for QA, so if anything this would be a net increase in resources allocated just to bring the machines onboard to do the task
Yeah... that is the other aspect where... labor is already getting fucked over massively so it becomes a question of how many jobs are even going away.
Whiplashed by one of the works by great bassist and producer Bill Laswell being inadvertently mentioned in discussion of AI.
I would initially tap the breaks on this, if for no other reason than "AI doing Q&A" reads more like corporate buzzwords than material policy. Big software developers should already have much of their Q&A automated, at least at the base layer. Further automating Q&A is generally a better business practice, as it helps catch more bugs in the Dev/Test cycle sooner.
Then consider that Q&A work by end users is historically a miserable and soul-sucking job. Converting those roles to debuggers and active devs does a lot for both the business and the workforce. When compared to "AI is doing the art" this is night-and-day, the very definition of the "Getting rid of the jobs people hate so they can do the work they love" that AI was supposed to deliver.
Finally, I'm forced to drag out the old "95% of AI implementations fail" statistic. Far more worried that they're going to implement a model that costs a fortune and delivers mediocre results than that they'll implement an AI driven round of end-user testing.
Turning Q&A over to the Roomba AI to find corners of the setting that snag the user would be Gud Aktuly.
The thing about QA is the work is truly endless.
If they can do their work more efficiently, they don't get laid off.
It just means a better % of edge cases can get covered, even if you made QAs operate at 100x efficiency, they'd still have edge cases not getting covered.
And hire other people with the excess budget. Hell, depending on how badly these systems are implemented, you can end up with more staff supporting the testing system than you had doing the testing.
The repetition of "Q&A" reads like this comment was also outsourced to AI.
I was going to say, this is one job that actually makes sense to automate. I don’t know any QA testers personally, but I’ve heard plenty of accounts of them absolutely hating their jobs and getting laid off after the time crunch anyway.
What does Q&A stand for?
Quality and assurance
Usually Questions and Answers.
Ugh. QA. Quality Assurance. Reflexively jamming that & because I am a bad AI.
Regardless, digital simulated users are going to be able to test faster, more exhaustively, and with more detailed diagnostics, than manual end users.
They already have a really cool solution for that, which they talked about in their GDC talk.. I don't think there's any need to slap a glorified chatbot into this, it already seems to work well and have just the right amount of human input to be reliable, while also leaving the "testcase replay gruntwork" to a script instead of a human.
That's a stupid idea. You're not supposed to QA or debug games. You just release it, customers report bugs, and then you promise to fix the bugs in the next patch (but don't).
No better testing than in production.
Or do the Bethesda thing and let people playtest their slop and fix it for free.
Literally not how any of this works. You don't let AI check your work, at best you use AI and check it's work, and at worst you have to do everything by hand anyway.
From a game dev perspective, user
Q&AQA is often annoying and repetitive labor. Endlessly criss-crossing terran hitting different buttons to make sure you don't snag a corner or click objects in a sequence that triggers a state freeze. Hooking a PS controller to Roomba logic and having a digital tool rapidly rerun routes and explore button combos over and over, looking for failed states, is significantly better for you than hoping an overworked team of dummy players can recreate the failed state by tripping into it manually.There's plenty of room for sophisticated automation without any need to involve AI.
Not all AI is generative.
I mean, as a branding exercise, every form of sophisticated automation is getting the "AI" label.
Past that, advanced pathing algorithms are what Q&A systems need to validate all possible actions within a space. That's the bread-and-butter of AI. Its also generally how you'd describe simulated end-users on a test system.
The article is specifically talking about generative AI. I think we need to find new terminology to describe the kind of automation that was colloquially referred to as AI before chatgpt et al. came into existence.
The important distinction, I think, is that these things are still purpose-built and (mostly) explainable. When you have a bunch of nails, you design a hammer. An "AI bot" QA tester the way Booty describes in the article isn't going to be an advanced algorithm that carries out specific tests. That exists already and has for years. He's asking for something that will figure out specific tests that are worth doing when given a vague or nonexistent test plan, most likely. You need a human, or an actual AGI, for something on that level, not generative AI.
And explicitly with generative AI, as pertains to Square Enix's initiative in the article, there are the typical huge risks of verifiability and hallucination. However unpleasant you may think a QA worker's job is now, I guarantee you it will be even more unpleasant when the job consists of fact-checking AI bug reports all day instead of actually doing the testing.
If it does the job better, who the fuck cares. No one actually cares about how you feel about the tech. Cry me a river.
The problem is that if it doesn't do a better job, no one left in charge will even know enough to give a shit, so quality will go down.
That's not generative AI though. Generative AI is the SLOP machine.
its *
Ironically, that’s definitely something AI could check for.
Spell check? Yeah fair enough. The misspelling has historical value now though so I have to keep it in :P
Ask it for many R's there are in strawberry
So Square Enix is demanding OpenAI stop using their content, but is 100% okay using AI built off stolen content to make more money themselves
As a developer, it bothers me that my code is being used to train AI that Square Enix is using while trying to deny anyone else the ability to use their work
I could go either way on whether or not AI should be able to train on available data, but no one should get to have it both ways
🙇♀️🙇♀️🙇♀️🙇♀️ I came to this 🧵 to make this exact ridicule.
Capitalists really showing their true colors when it comes to self-indulgence.
QA annnnd Debugging?
LLMs have a much better chance at succesfuly replacing whoever said that.
the LLMs can barely make fucking working code, let alone test it
Well, good luck with that. Software development is a shit show already anyway. You can find me in my Gardening business in 2027.
Good Luck. When the economy finally bottoms out the first budget to go is always the gardening budget.
You can find me in my plumbing business in 2028.
I deal with shit daily so it's what we in biz call a horizontal promotion.
Market gardening isn't so bad, people gotta eat. But yeah, if you're cutting lawns you're going to suffer when the economy shits the bed.
That is a valid point, and I think i'll preemptively pivot to woodworking.
Square Enix actually has a pretty sick automated QA already. There's a cool talk about how they did that for FFVII remake in GDC vault, and I highly recommend watching it, if you're at all interested in QA.
It has nothing to do with AI, it's just plain old automation, but they solve most of the issues you get with making automated tests in non-discrete 3D playspace and they do that in a pretty solid way. It's definitely something I'd love to have implemented in the games I'm working on, as someone who worked in QA and now works in development. Being able to have mostly reliable way how to smoke-test levels for basic gameplay without having to torture QA to run the test-case again is good, and allows QA to focus on something else - but the tools also need oversight, so it's not really a job lost. In summary - I think the talk is cool tech and worth the watch.
However, I don't think AI will help in this regard, and something as unreliable and random as AI models are not a good fit for this job. You want to have deterministic testcases that you can quanitfy, and if something doesn't match have an actual human to look at why. AI also probably won't be able to find clever corner-cases and bugs that need human ingenuity.
Fuck AI, I kind of hope this is just a marketing talk and they are actually just improving the (deterministic) tools they already have (which actually are AI by definition, since they also do level exploration on top of recorded inputs), and they are calling it an "AI" to satisfy investors/management without actually slapping a glorified chat-bot into the tech for no reason.
That's why they want 70% QA from AI. Because right now their games are only 10% QA tested.
30% of the the QA team will be working at 300% capacity - can you guess which 30%
Realistic goal considering they already do so little QA.
I hope they put out the last FF VII remake part before that, so i can finally start playing them all! I don't care what they want to waste their money on afterwards lol
I wouldn't be shy about getting into Remake or Rebirth now. They both stand up as their own games (concise start/ending, somewhat distinct mechanics, each one is easily 40+ hours of gameplay). And with Part 3 targeted for 2027 release, I suspect this kind of overhaul would be outside their dev cycle to implement.
Part 2 is already using the engine from Part 1 with minor adjustments. I suspect most of Part 3 development is cinematics and world building.
Remember when square used to make great games?
That was Squaresoft, not Square Enix.
I loved Enix's Ogre Battle and Square's Final Fantasy 6 and 7. How could putting the companies together make a bad?!
Square Ewwwwnix
My memory failed me. Thanks for the correction.
Not buying their next game it will be a nightmare
Was gonna make a post with this article but this is related so I'll drop it here instead.
"Square Enix is laying off more developers in the UK and US as it refocuses on Japan The publisher has expressed interest in replacing development roles with AI"
https://www.engadget.com/gaming/square-enix-is-laying-off-more-developers-in-the-uk-and-us-as-it-refocuses-on-japan-201907305.html
Square Enix exec doesn't know what QA and Debugging entail.
"Well it works for unit testing, so just extend that out to all testing! Problem solved!" -Senior Management, probably
Who am I kidding. They have no idea what unit testing is.
Considering how the open source community is being inundated with low-quality bug reports filed using AI, I don't have much faith in the tech reviewing code, let alone writing it correctly.
Could it be a useful aid? Sure, but 70% of your reviewing is a pie-in-the-sky pipe dream. AI just isn't ready for this level of responsibility in any organization.
And I thought I had no more disappointment left to allocate
It's just that you've reached your free quota, further disappointments will be charged 0.0937 emotional stability per hour
They jumped on the NFT bandwagon a couple years ago, too. Did they not learn anything from that?
it's a bit late in the game to be making idiotic claims but I guess the default state for corpos is being out of touch
Frankly, this is good news. Whoever buys the rights to kingdom hearts in 3 years when the company falls apart might manage to create an intelligible storyline.
But then it won't be a Kingdom Hearts game!
Kingdom hearts 1 was a coming of age story with some fantastic elements tossed in
Kingdom hearts 2 was about antipathy and how it destroys the world, with some, uhh, who was that guy? And why’s the bad guy on my side?
Kingdom hearts 3 was, wait, why was he cloned? When was he cloned? When was she, and him, and him again? And his third clone was a girl? And whose heart was imbued into what? What war? What? Who? What???
I get that getting all the games as they released was hard, because the series is on so many platforms. But I really don't get the "KH is hard to understand" argument today, because you can easily find hundreds of letsplays for every game, cutscenes complications, play/watch every game on the PS4 remix disk, and even watch a fandub of the mobile games (Dark Road is a WIP) if you don't like the KHUX Back Cover recap.
So like, what's so hard? If you skip games and only read a wiki (the worst possible way to consume any sort of media, mind you), of course you're not gonna know the story and characters, and of course it'll sound confusing.
Dude, it’s multi-author-comic level bad. I’ve skipped entire sagas in several book series due to a lack of translations and ended up less confused. It’s green arrow levels of clone shenanigans.
To be clear, I’ve played most of the games and they’re still ridiculously difficult to keep track of. All besides the mobile, early non-Ventus card mechanic arpg, and the disappearing girl clone sora game.
They’d be easier to follow if they stuck to the rules of their own universe. Body and heart separate and the body persists not once but twice? What?
Be prepared for Square Enix games to fail even EA's QA standards in the near future 😅
It's your funeral
Their games going to be shit 🤣
Chrono Trigger remake, first all AI development, because we live in the bad timeline.
spits
Oh. Gross.
A year from now they'll be wondering why their games are so buggy.
"The players are obviously insane and confused. The AI is always right."
So their games will cost 70% less right?
From a tech POV, that makes a lot of sense. Use AI to find the needle in the haystack. Then let a person validate. That's probably one of the better uses for it. Although I don't love AI for any of the broad reasons to not like AI.
Wasn't AI decent at writing code, but bad at review and modifying it?
Maybe before. But it's gotten pretty damn good at detecting anomalies and issues. And every time a human QA validates the info, it gets better.
I'd still leave it to a human to fix the code though. I suspect that letting AI write the code would make it unworkable for people in the future. But maybe it can write code in a straightforward way to be managed. I don't know. It's advancing pretty fast.
Hmm. While I don't know what their QA workflow is, my own experience is that working with QA people to design a QA procedure for a given feature tends to require familiarity with the feature in the context of real-world knowledge and possible problems, and that human-validating a feature isn't usually something done at massive scale, where you'd get a lot of benefit from heavy automation.
It's possible that one might be able to use LLMs to help write test code --- reliability and security considerations there are normally less-critical than in front-line code. Worst case is getting a false positive, and if you can get more test cases covered, I imagine that might pay off.
Square does an MMO, among their other stuff. If they can train a model to produce AI-driven characters that act sufficiently like human players, where they can theoretically log training data from human players, that might be sufficient to populate an MMO "experimental" deployment so that they can see if anything breaks prior to moving code to production.
I think that the problem is that you're likely going to need more-advanced AI than an LLM, if you want them to just explore and try out new features.
We can do a reasonable job of summarizing human language with LLMs today. I think that that might be a viable application.
False positives during testing are a huge time sink. QA has to replicate and explain away each false report and the faster AI 'completes' tasks the faster the flood of false reports come in.
There is plenty of non-AI automation that can be used intentionally to do tedious repetitive tasks already where they only increase work if they aren't set up right.
insert plane crashing.gif
Lol. Good luck!
So... im a big supporter of squeenix, buy everything they make... but this tells me the quality of their games is going to go down the toilet. Knowing AI it'll come up with fake lists of bugs that didn't happen and all the real bugs will not be listed and they'll release the buggies shit. One thing I LIKE about square, being one of the few companies I do pre-orders from still, is that their products are fairly bug free on launch FFXVI had some graphics optimization issues, but I've been happy with most of what I got the past few years.
FFXIV saved them and they don't ever put any money back into the game. It's their cash cow that pays for all their other bad ideas.
Great graphics and legacy. But some crappy ideas about what the players want.
Look at their payment systems for subs. It's so confusing for no damn reason.
Believe it or not FFXI was even MORE confusing to sub
A lot of hate in the comments but IMO this is one of the few things that LLMs are actually really good for. It's a shit job nobody wants to do that LLMs are really good at. Notice that they said 70% and not 100%. Yeah that means they're probably going to have 30 people doing the work that 100 people used to do but people are still in the picture overseeing things. Automation isn't, by itself, bad. The bad part is that our whole society is built on the idea that your entire value as a person is based on being able to work and make money and job loss is way worse than it should be.
70% by what metric?
Is that going by bugs identified, fixes implemented, headcount?
Moment for silence for David "Ribs" Carillo 🪦
Well their goes FFXIV, that will be their end
more shit
Sure
dont they already have dumbbots in playtesting?
The Talos Principle certainly did in 2014.
Large companies probably do that anyway.
Take Blizzard for example. They just released a new patch, where class campaign quests for 8/12 classes do not work. Sure, it's a remixed version of older expansion, and with all the phasing stuff I can kind of imagine some of the phasing issues being caused by, I don't know, the player having a weird combination of completed stuff that's hard to properly catch in testing, since there's quite a lot of variables.
But the fact that one of the class quests requires crafted items to be completed, while crafting isn't available by design in the Remix, there's just no excuse. They either just don't give a fuck about an issue that's literally a progression blocker with 100% repro rate (while also being pretty easy to fix), or no one ever tested it even once. And it's not just some random sidequest, it's literally the main class campaign, one of the main features of the expansion.
As someone who worked in QA and gamedev, I can't imagine how could something as obvious as this ever get approved for release. That's something you catch immediately. Hell, you don't even have to play through it to realize that this might be a problem.
Work at a larger company. Most people are so used to terrible Customer Service these days that we just use our customers as the QA. Nobody complains as much as they should. As they say
Inb4 their games come out even more broken
@inclementimmigrant I'm so glad I've stopped buying AAA games.
I will continue ignoring anything they make
In my country we say ajajajajajjjj
Some AI or central computer going haywire and destroying everything is, like, the third or fourth stock RPG trope just behind the Dark Lord burning down the protagonist's village in the first act or the mysterious waif girl actually turning out to be a princess.
You really think they'd know better.
Given how much squenix struggles with changing its development practices, I would be very surprised if they actually got there.
Barf.
Grrroooossss, noooo I liked you Square Enix in spite of everything else.
I'm cautious but a little curious about this one, because QA could actually be a very good target for AIs to work with.
That said, those talking about human creativity and player expectations are still correct. An AI can report a problem with feedback that a human can say "No, that looks fine. Override that report." It will also be good to do occasional manual tests, and lament "How did the AI think this was okay??"
Well, it's not game development, but bugfixes and quality testing.
I dont know, but it does makes sense, when there's still 30% work being done by human eyes. There will still be people checking everything through.
Even if they hit 50-50, they could put more money into the development.
The argument that they will just save the money only works as long as another company doesnt use it for game devs. Otherwise you naturally fall behind.
It also only works as long as the AI can actually competently do the QA work. This is what an AI thinks a video game is. To do QA, it will have to know that something is wrong, flag it, and be able to tell when it's fixed. The most likely situation I can foresee is that it creates even more work for the remaining humans to do when they're already operating at a deficit.
To be fair, that's what an AI video generator thinks an FPS is. That's not the same thing as AI-assisted coding. Though it's still hilarious! "Press F to pay respects" 🤣
For reference, using AI to automate your QA isn't a bad idea. There's a bunch of ways to handle such things but one of the more interesting ones is to pit AIs against each other. Not in the game, but in their reports... You tell AI to perform some action and generate a report about it while telling another AI to be extremely skeptical about the first AI's reports and to reject anything that doesn't meet some minimum standard.
That's what they're doing over at Anthropic (internally) with Claude Code QA tasks and it's super fascinating! Heard them talk about that setup on a podcast recently and it kinda blew my mind... They have more than just two "Claudes" pitted against each other too: In the example they talked about, they had four: One generating PRs, another reviewing/running tests, another one checking the work of the testing Claude, and finally a Claude setup to perform critical security reviews of the final PRs.
I don't know what they were testing, but if your output is text, it will be a lot easier for the AI to know it's correct than any of the plethora of ways that video games can go subtly wrong, and that's where my lack of faith comes from. Even scraping text from the internet, my experience is more often that AI is confident in its wrong answer than it is helpful.