Brain-computer interface startup demo'd a consumer BCI headphone prototype in SF. First version recognizes 10 thought commands — think "LinkedIn" and it opens the app, think "remember" and it tags the current moment.
Shipping target: next year. Extremely limited vocabulary at launch, but represents accessible BCI hardware outside surgical implants.
Neuralink context: Still not ready for general deployment. Elon's timeline for "cognitive enhancements" is ~3 years out. OpenAI and Meta also running parallel BCI research programs.
The interesting technical question: How are they doing thought recognition with non-invasive sensors? EEG-based pattern matching? Muscle signal detection? Most consumer BCIs struggle with signal-to-noise ratio without skull penetration.
10-word vocabulary sounds like they're using a trained classifier on specific neural signatures rather than general language decoding. Probably why "remember" works as a discrete trigger rather than parsing arbitrary thoughts.
Founder just relocated to SF 5 days ago. Classic YC-style "move to the Bay, build fast" playbook.
Shipping target: next year. Extremely limited vocabulary at launch, but represents accessible BCI hardware outside surgical implants.
Neuralink context: Still not ready for general deployment. Elon's timeline for "cognitive enhancements" is ~3 years out. OpenAI and Meta also running parallel BCI research programs.
The interesting technical question: How are they doing thought recognition with non-invasive sensors? EEG-based pattern matching? Muscle signal detection? Most consumer BCIs struggle with signal-to-noise ratio without skull penetration.
10-word vocabulary sounds like they're using a trained classifier on specific neural signatures rather than general language decoding. Probably why "remember" works as a discrete trigger rather than parsing arbitrary thoughts.
Founder just relocated to SF 5 days ago. Classic YC-style "move to the Bay, build fast" playbook.