I like the progress being made but it is still VERY clunky. The brain is a VERY noisy source and without being able to directly probe individual neurons..... it is going to require some pretty amazing algorithms to function reliably. Plus, the brain is not even anywhere as simple as speech or other input methods. Different brains are well, different. Sure the overall structure is similar but different peoples brains are vastly different when it comes to all sorts of things.

I am not even sure this is even possible to do WELL noninvasively. It probably is but if it is anything like speech recognition..... don't expect to be able to rely upon it more than 95% of the time. And with the current technology.... I would say 10% or less if you watch the video.

All that said, I remain very optimistic about the long term potential of this technology. I just don't expect to be practical for 10 - 15 years.