Mind-reading cap? This AI cap can turn your thoughts into text

When people talk about controlling devices with thoughts, Neuralink usually comes up first. But Neuralink involves surgery. Now, a California-based startup called Sabi is trying something simpler. Instead of implants, it’s working on a wearable cap or beanie that can read brain signals and turn them into actions. Also Read: Your ChatGPT, Claude, Gemini chats can be used in court; Here’s why

But the question is – how does this “mind-reading” cap work? Sabi’s approach is based on EEG (electroencephalography), which basically records electrical signals from your brain through sensors placed on the scalp. The difference here is scale. While typical EEG devices use a limited number of sensors, Sabi claims its cap could use tens of thousands of sensors to capture more detailed signals. Also Read: Google’s Gemini AI app comes to Mac: How to download and use

What can you actually do with it?

The company says the first use case is simple: turning thoughts into text. So instead of typing on a keyboard, you just think and the system converts that into words on a screen. Right now, the expected speed is around 30 words per minute, which isn’t fast yet. But like most AI-based tools, it’s expected to improve with usage and training. Over time, this could extend to controlling apps, devices, or even communication tools without physical input. Also Read: Google launches Gemini 3.1 Flash TTS, a new AI voice model that supports 70+ languages

The startup is backed by investor Vinod Khosla, and was co-founded by Rahul Chhabra. Sabi is also building an AI model trained on large amounts of brain data to better understand patterns and convert them into usable output. According to the company, privacy remains a focus, with user data expected to stay secure.

Is this practical yet and when can you try it?

That’s still a question. While the idea sounds promising, brain-computer interfaces without implants are harder to perfect because of signal limitations. Still, the non-invasive approach makes it easier to test and scale, especially compared to surgical solutions. 

Sabi says the product could arrive by the end of this year, although pricing hasn’t been revealed yet. There’s already a waitlist open for early users. This isn’t replacing keyboards anytime soon. 

Add Techlusive as a Preferred Source

And if it works even half as well as promised, it could change how we interact with devices over time.