Vol. 67 No. 2 2000 - page 188

188
PARTISAN REVIEW
was used in the 1890 census (the first time the American census was
processed with electrical data processing equipment) and the computer
that Turing came up with that broke the German Enigma code.
A Polish spy, I think his name was Lewinsky, had stolen this Enigma
coding machine that had three coding wheels-and they could see how
the machine coded information. He was an unheralded hero of World
War
II.
All they needed to do then to decode the messages was to figure
out every combination of the coding wheel, so they could simply use a
computer to look at every single combination and apply that to the mes–
sages and get a decoding. The only problem was that the computer had
not been invented yet. (Actually, Charles Babbage invented a computer
in the nineteenth century, but never got it to work.) There had never
been a fully functioning computer; there had only been calculators. So
Turing and his team invented the first special purpose computer with
telephone relays which provided an uninterrupted decoding of the Nazi
messages . Churchill got every key message from the Germans from his
intelligence office, decoded by Turing's machines, and refused to use
most of it. He refused to warn the English cities that they were going to
be bombed because he was concerned, correctly, that the Germans, if
they saw this preparation, would realize that their code had been
cracked. But in the Battle of Britain, he did use this information, so that
the English planes knew exactly where the German planes would be,
and were able to prevail even though they were outnumbered. Other–
wise history might have been a little different.
The computer that CBS used to predict the election of Eisenhower in
1952 is on the chart, as is the computer you bought for your daughter
last December. I put the forty-nine computers on an exponential graph,
in which a straight line would mean exponential growth. Every line you
go up on the chart means multiplying computations by a factor of one
hundred. So a straight line on an exponential graph means exponential
growth. And one of the first things I noticed is that the exponential
growth of computing goes back one hundred years, but Moore's Law
didn't kick in until 196o-semiconductors were invented in 1957-s0
the paradigm of shrinking transistors on an integrated circuit didn't
really come in to play until the 1960s. Moore's Law is not the first, but
actually the fifth paradigm to provide exponential growth of computing.
We had electromechanical calculators with relay-based computers, like
Turing's; we had vacuum tube-based computers, which included the CBS
machine; we had transistor-based computers which we used in the first
space launches, and then we had integrated circuits. And each new para–
digm came along just at the time when the other paradigm was running out
175...,178,179,180,181,182,183,184,185,186,187 189,190,191,192,193,194,195,196,197,198,...339
Powered by FlippingBook