"CES 2026 has officially begun, and Razer is using this year's showcase to reveal a slew of AI-powered gadgets. One of the most intriguing debuts is Project Motoko, an AI headset powered by Snapdragon. Equipped with a pair of cameras near each earcup, it's capable of analyzing your surroundings and giving you audio feedback on what it sees. Razer says Project Motoko can provide all sorts of feedback to users."
"For example, it can be used while gaming to get feedback about a boss fight, detecting what's on the screen and giving you immediate tips. It can translate street signs when traveling abroad, summarize documents, or handle tasks as simple as counting your reps at the gym. Equipped with dual far- and near-field microphones, it can both pick up your voice commands and listen to your surroundings for environmental cues, ensuring it gives you as accurate a response as possible."
"Motoko is built to connect seamlessly with most AI platforms, including OpenAI, Grok, and Gemini. It can also be used for machine learning, passively gathering data to improve the technology as the device is worn. It's worth noting, however, that this is a fully auditory system for the user. Unlike smart glasses, which sometimes offer in-lens visual feedback, Motoko looks like your standard headset."
Project Motoko is an AI headset powered by Snapdragon with cameras near each earcup that analyze surroundings and deliver real-time audio feedback. Use cases include in-game boss-fight guidance by detecting on-screen content, translating street signs, summarizing documents, and counting gym repetitions. Dual far- and near-field microphones capture voice commands and environmental cues to improve response accuracy. The headset interoperates with AI platforms including OpenAI, Grok, and Gemini and can passively gather data for machine learning while worn. The design is audio-only with no visual display. Project Motoko is a concept with no release date; Project AVA is a 5.5-inch animated avatar expected in the second half of 2026.
Read at GameSpot
Unable to calculate read time
Collection
[
|
...
]