AI Assistants Like Project Ava Would Turn Gaming Into A Chore

To the chagrin of pretty much everyone, generative AI has wormed its way into all levels of game development. Not content to just let it infect the way games are made, some companies are now pushing to hand over the controller to generative AI and let it play the games, too.

This week, fancy mouse manufacturer Razer showed off an “improved” version of its Project Ava at the 2026 Consumer Electronics Show. Ava is an AI-powered assistant designed to analyze game footage and offer real-time advice while you play, and while the project isn’t new, its latest iteration is a freestanding peripheral that displays a cat-eared anime girl trapped in a glass tube who just kind of stands there while Ava robotically delivers advice.

First impressions of the new Ava model are about as good as you’d expect. Reports from Gizmodo, PC Gamer, and The Verge mention Ava being unable to answer basic questions about the game it’s supposed to be analyzing, offering blatantly incorrect commentary, constantly veering off-topic with pointless asides, and generally skeeving out the writers interacting with it thanks to its flirtatious personality. Oh, and it’s also running on Grok, the chatbot model that’s mostly being used to create nonconsensual sexual material of unsuspecting women and girls over on X.com.

So, as you could probably guess based on every other application of generative AI, Project Ava doesn’t seem to actually deliver on its promises. But whether it works is secondary to a bigger question — why does this exist?

Around the same time that Razer’s AI catgirl prison was revealed, a patent for Sony’s latest proposed intrusion of AI into games also surfaced. Filed in 2024, the tech would use AI to offer varying levels of help to players stuck on difficult parts of games, as VGC reports. Options range from displaying a ghostly avatar pointing the player toward goals to the AI taking the reins and completing difficult sections on its own. Even assuming it could work as intended, it, like Ava, is a technology that seems to miss the point of gaming entirely.

The best-faith reading of Sony’s patent is that it could be used as an accessibility aid to get players through portions of games they physically can’t complete, or as a way to keep difficult sections from halting progress entirely. We’ve argued for better difficulty options and accessibility features before, but those are best handled by developers thinking about what works best for their games, rather than by a third party imposing ideas from the outside.

Both Razer and Sony are trying to solve problems that aren’t really problems. Difficulty can be a barrier, but it’s not a bug to be fixed. When a game is too challenging to complete, it’s a sign that you’re approaching it the wrong way or need more practice. In either case, the solution is to rethink what you’re doing, adopt another strategy, and try again. That process is what it means to play a game, and if you’re not enjoying it, there’s no shortage of other titles out there.

While a game that plays itself sounds absurd on its face, it’s Project Ava’s “helpful” tips that actually seem more objectionable, even ignoring the inherent problems with its AI model. Playing a goal-oriented game like Battlefield, Razer’s CES example, is about experimentation. Figuring out why what you’re doing isn’t working is just as important as getting it right in the end, and skipping straight to the solution means treating games like problems to be solved as efficiently as possible, rather than art to be enjoyed.

There’s no shortage of advice you could be finding on your own. Countless other players are eager to share their own strategies online, and the asking and answering of questions is a core part of what makes gaming a social activity. Project Ava presumes that querying a chatbot is preferable to pausing a game to seek the advice of your fellow humans, but that’s missing the point entirely. Any part of a game that encourages you to think about it or to connect with other players is a gift, not a burden to be solved with technology.