So he went ahead and voted for it. Makes total sense.

I just don't believe Windows is a suitable OS for embedded devices.

I'm all for bashing Microsoft for its numerous failings, but we are talking about a mobile OS, completely rebuilt from the ground up, which is yet to be released. It completely breaks backward compatibility, has a completely new programming model and APIs (as far as Windows Mobile is concerned) and imposes strict requirements on the hardware, in stark contrast with past WM versions. In this case, any previous experience we've had with Windows Mobile is irrelevant, with the only possible exception being Zune, as it's said that WM7 borrows parts of the UI (or maybe much more than just the UI?) from it.

In the end, you may turn out to be right, but until we've had the chance to play with it, such dismissals on general grounds seem unjustified.

A potential issue I didn't see addressed in the article - crosstalk. What happens when an outbound signal, neuronal activity triggering a light pulse, is produced at the same wavelength (color) as another is tweaked to "listen" for? Would the brain be able to compensate and filter out such signals, as this essentially creates a form of an artificial permanent link between the two? Or maybe this isn't an issue beyond, say a few hundred microns, because the energy of the outgoing photons is below the sensitivity threshold of the listening ones? Otherwise, it might not be very practical to have to deal with your left knee bending every time you think of the letter "P".

Also, in theory, how many wavelengths could different molecules be produced to fluoresce (if this is even the right term) at? Same thing for the light-sensitive ones. How wide is the sensitivity waveband? IOW, how large a bandwidth can we expect to command, in each direction? Or put in even simpler terms, how many different things can this technology be employed for, simultaneously?