Since the 1950s, it has been the goal of workers in the field of artificial intelligence to create an autonomous thinking computer. This aim has always been ten years in the future, its attainment retreating as fast as we approached it. Many gave up hope of ever seeing it; indeed the very term artificial intelligence has become a joke in some circles. More recent projects, such as the Japanese drive to develop a Fifth Generation computer, have also failed to meet their ultimate aims. But the idea of a machine that can match or surpass the human brain in its ability to reason has recently resurfaced, along with a debate on the ethics of actually building one. Part of the resurgence in interest can be attributed to Sony’s toy dog Aibo, shortly to be joined by Poo-Chi from Sega.
Artilect has started to be used as a term for devices that exhibit autonomous learning behaviour, a blend from artificial intellect. It was apparently coined by Professor Hugo de Garis, head of the Brain Builder Group at the Advanced Telecommunications Research Institute in Kyoto, Japan. Prof de Garis, who calls himself an intelligist (another word he seems to have invented), argues that by 2050 we shall indeed have computers of superhuman intelligence. At the moment, he’s working on Robokitten, a device with the intelligence level of a kitten, a big step in computer terms, but hardly threatening to humanity’s dominance as yet — well, not till it gets hung up on the curtains ...
Asteroid-sized, self-assembling, nanoteched, one bit per atom, reversible, heatless, 3D, quantum-computing artilects could have intelligences literally trillions of trillions of trillions of times the human level.
Independent, Nov. 1999
“Humanity will have to make a choice about whether we want to build these artilects or not,” says De Garis, who sees two factions arising — one for and one against the “artilects”.
Daily Telegraph, Dec. 1999