Have been more preoccupied with technology than I expected over the last two weeks. Innovation Watch has now moved to a faster server, and the site was down for a week as we worked out the kinks. Thanks for your patience. Things should get back on track this week as we pick up the threads.

I've been fascinated with the idea of artificial intelligence since I first read Edward Feigenbaum's book Computers and Thought many years ago. Of course, in the early days, AI was all about clever programming. Computers applied heuristics to solving problems, exhaustively explored a 'mental map' of possible options, and made a choice after evaluating all of the consequences. The strategy of a chess-playing machine was to look ahead at all of the possible moves, but it was only as good as the heuristics created by a human programmer.

Artificial intelligence was soon embodied in robots that faced the challenge of vision and movement. Early robots moved by creating detailed models of the space they inhabited, and used heuristics to get around. Then along came Rodney Brooks. A robotics pioneer at MIT, Brooks challenged the prevailing wisdom with a contrary view. Instead of building elaborate mental models, he said, robots needed only simple reflexes to get around. Insects did just fine at that, without complex intelligence.

Perhaps five years ago -- I can't remember exactly -- I read a Globe and Mail article about Stephen Thaler. A former scientist with McDonnell Douglas, he had developed and patented a machine -- he called it the Creativity Machine -- that was capable of real invention. The machine itself held a number of US patents. It composed music, helped with automobile and aircraft design, and 'brainstormed' new consumer products. The cross-action toothbrush was just one example. When the Creativity Machine was connected to a robot, the robot learned in just seconds how to move.

Thaler's approach to artificial intelligence closely parallels the behavior of the human brain -- with profound consequences. The Creativity Machine can take on problems in almost any field, Thaler says, and could even be at the heart of an intelligent Internet. It's clear that artificial intelligence has moved far beyond its origins, and it's much more intriguing now.

Stephen Thaler recently made a video about his invention, and I've posted a link to it here.