Silicon Valley startup Sabi unveils brain-reading beanie to decode human thoughts |
Imagine you are wearing a cap that can decode your thoughts. At first, the idea sounds absurd and not closer to reality.
To one’s surprise, Silicon Valley startup Sabi is developing a brain wearable that can decode a person’s internal speech into words on a computer screen.
This brain-reading beanie is based on technology known as a brain-computer interface, providing a direct communication pathway between the brain and an external device.
By utilizing a dense array of 70,000 to 100,000 miniature EEG sensors woven directly into fabric, Sabi is developing a non-invasive thinking cap that would translate imagined words into text.
“Given that high-density sensing, it pinpoints exactly what and where neural activity is happening. We use that information to get much more reliable data to decode what a person is thinking,” CEO Rahul Chhabra said.
The initial winter-hat design is slated for release by late 2026. According to Chhabra, the company is also planning to design a more streamlined baseball cap.
Khosla, founder of Khosla Ventures, one of Sabi’s investors, said, “The biggest and baddest application of BCI is if you can talk to your computer by thinking about it. If you're going to have a billion people use BCI for access to their computers every day, it can't be invasive.”
Initially, the company is working on an initial typing speed of 30 or so words per minute. This speed is often slower than most people type, but this speed will eventually improve with time depending on the time users spend with the cap.
Here is the catch! The brain-ready wearable also has some drawbacks. For instance, it is possible that sensors might not catch strong neural signals coming from the brain due to layers of skin and bone. On the contrary, the surgically implanted devices have an edge by picking up much stronger signals.
Moreover, BCI models also depend on AI models to translate neural signals into real-time commands. To power the interface, Sabi has trained a "brain foundation model" using 100,000 hours of neural data collected from 100 volunteers.