- Razer's Project Ava is a gaming AI copilot that analyzes gameplay and provides suggestions to improve skills.
- It takes photos of the screen, identifies patterns, and offers tips in real time.
- It can also offer feedback post-game to help you strategize for the next match.
Back in 2017, Nvidia shared a video on the occasion of April’s Fool Day, showcasing a plug-in drive called GeForce GTX-Assist. It was designed to analyze your gaming skills and take over the game on your behalf with AI’s help. Well, I never in a million years could have imagined the Razer would actually be working on something like it, until they announced the new Project Ava.
Project Ava aims to be the copilot that will “coach” you through games to help you get better at them. To accomplish this, the AI will take thousands of photos of your screen as you play. And then analyze them to give you suggestions on what you can do to beat bosses or have a leg up at competitive matches.
According to The Verge, Project Ava was shown a prerecorded boss fight from last year’s Black Myth: Wukong. In response, the AI replied with pretty responses like, “Get ready to dodge when his blade spins or glows with an orange tinge” or “Keep a close eye on his health, once you shave off 10 to 20 percent, get your dodge fingers ready”.
They also demoed League of Legends, where Ava suggested what spells and items you need based on the current scenario. It also gives out the enemy position since it captures multiple shots of the mini map, making it seem like you are almost cheating. Even after the match, the AI will continue suggesting tips, and details about your gameplay. It will offer feedback on how to improve next match.
However, its responses were a bit delayed on the prototype at CES. Possibly because it was running locally on a pair of Nvidia RTX 4090 laptop GPU with Meta’s Llama 3.2 LLM. Razer is going all in with Project Ava with no signs of slowing down. Razer also has a patent-pending algorithm on the way to suggest gaming tips and will run with its proprietary hardware.
Honestly, all these details look interesting on the surface, but to me, it feels like Mircosoft Recall but for games. At the moment, Razer is training Ava with many gaming guides that are available online. But Razer isn’t crediting those websites to make said guides. But that has become a standard with AI in tech. I feel it needs to do more than suggest loading-screen tips if it really wants to help you improve gameplay.
