Whenever I tell someone I work on brain-computer interfaces, the reply is almost always:
"Wait, isn't that already solved?"
EEG has been around for a century. Neuralink is in people.
So what exactly is left?
A thread on what the BCI problem really is π§΅ (1/n)