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.
Exactly, as I don’t expect QA done by something that can’t think or feel to know what actually needs to be fixed
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
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.
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.
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.
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.
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.
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.
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.
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.
“Because I would love to be able to start up 10,000 instances of a game in the cloud, so there’s 10,000 copies of the game running, deploy an AI bot to spend all night testing that game, then in the morning we get a report. Because that would be transformational.”
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.
One former Respawn employee who worked in a senior QA role told Business Insider that he believes one of the reasons he was among 100 colleagues laid off this past spring is because AI was reviewing and summarising feedback from play testers, a job he usually did.
We can do a reasonable job of summarizing human language with LLMs today. I think that that might be a viable application.
Worst case is getting a false positive, and if you can get more test cases covered, I imagine that might pay off.
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.
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.
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&A QA 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.
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.
I mean, as a branding exercise, every form of sophisticated automation is getting the "AI" label.
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.
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.
videogameschronicle.com
Najstarsze