Matsushita is developing something called Brain-on-a-Chip. It sounds quite interesting, even if the name is a bit spooky.

I'm getting a bit worried about the fact that American companies aren't willing to take risks like this. We abandoned fuzzy logic and robotics because they didn't get a significant return on investment in 18 months. The Japanese, on the other hand, persisted and are now starting to come up with some quite innovative products (Aibo, etc.) which we simply lack the expertise to produce.

This is also the same timeframe which saw the demise of the research houses of Bell Labs, PARC, HP, and -- just a couple weeks ago -- a good chunk of IBM (which is selling their storage systems division to Hitachi -- a field they invented, for crying out loud).