OrionWorks - Steven V Johnson <[email protected]> wrote:
Under a 100% free-enterprise system I know of few business enterprises that > could justify to their stock holders a plan to make investments that could > take up to 20 - 50 years to start generating > dividends for their stock holders. If free enterprises was the only game > in town funding the development new unproven technologies like > integrated circuitry, electrification of the grid, building > highways, transistors, etc... could have never gotten off the ground. It is hard to imagine anyone but Uncle Sam building the first computers. Or highways, or the Transcontinental Railroad. Or project Whirlwind. Transistors, on the other hand, would probably have taken off their own. Setting up a transistor production line was cheap. As I said, most transistors were initially more expensive than the vacuum tubes they replaced. Some of them far more expensive. But even at a high cost transistors still had a niche markets in civilian applications, such as hearing aids. Hearing aids were among the first product made with them, although I think it was at no profit, as a public service. They soon began making small portable radios, which sold at a premium. The prices soon fell. Transistors would have taken off, but it would taken many years longer. You cannot run history over again, so it is hard to say how much longer. The point is, cold war weapons and the space race produced hot-house growth. As I said, a whole new type of fast memory chips were developed for the LARC computer. Two LARC computers were constructed. The chips were obsolete and never used for anything else. This is like developing a iPad, selling 100 of them, and then junking the design and starting over again with a different display technology. It is economic lunacy. But it sure pushed the technology along quickly! I have an article written in 1969 with a long list of major computer design innovations developed or advanced for the LARC, such as instruction overlap and hardware modularity. Project like LARC not only advanced the state of art at high speed, they also trained a large fraction of the people who designed and built computers in private industry for the next 20 years. I have heard that just about everyone in the business in the 1960s got started at MIT Whirlwind or SAGE. MIT Whirlwind at the time appeared to be a black hole of expenses. It went on for year after year, swallowing up millions of dollars (in 1950s money) where only a few hundred thousand were initially budgeted. It seemed to produce nothing. It was classic example of an Air Force project gone bonkers. Then, as it wound down, it spun off CPU designs, core memory thousands of times faster, cheaper and more reliable than anything previously developed, orders of magnitude faster processing, the SAGE computer, and host of other things. Just one of those breakthroughs would have paid for the whole thing. As I recall, it never actually achieved the initial goal, which was to make a flight simulator for WWII. It began in 1945 and run until 1953 -- way behind schedule, and WAY over budget. In retrospect, that does not matter. - Jed
