Slashdot Mirror

The Japanese are wise in their attempt to deal with an aging population via automation rather than immigration but they are foolish to do it in another Apollo-style project. They tried this already and it was destructive to their technical base and to ours.

Around 1980 the man who had told me I could work on getting PLATO to the mass market (thus drawing me away from working on an 8086 operating system on CDC Cyber emulation before the first silicon had been etched) ran into the system programmers' room in Arden Hills, MN and excitedly told us that the big new challenge was to beat the Japanese at their recently announced "Fifth Generation Computing" initiative -- which was supposed to do things like give us AI, robots, etc... etc... I was even invited to join the newly formed Microelectronics & Computing Consortium which would have required that I abandon the work to bring PLATO to the mass market. I refused the offer and ended up being disappointed that CDC's management didn't follow through when presented with the scale and economy of systems/networking (cable/phone/etc.) required to meet mass market demands of the era.

So the mass marketing of networking was delayed a decade or two while Cyc (a spin-off of MCC) attempted to solve the 'common sense' computing problem. What they did was create a cool logic programming system with persistence but not the 'common sense' solution predicted by the Japanese or the supporters of industrial policy in the US.

I've said it before and I'll say it again -- the future of all this stuff is solving the rational programming problem deriving from Russell and Whitehead's vision of a relation arithmetic.

PS: Ask yourself why there has yet to be a programming language that uses dimensions and their units as an integral aspect of types? Statistics apply to anything.