There will probably be a good deal of work of this kind to be done, for every known process has got to be translated into instruction table form at some stage.

@Turing's ideas about the proper approach to computer design stressed the need to build computing capabilities into the program, not the hardware.

Delay there must be, due to the virtually invisible snags, for up to a point it is better to let the snags be there than to spend such time in design that there are none (how many decades would this course take?).

It was during this American visit that @Turing picked up practical knowledge of electronics. Turing had first become acquainted with what were then called "electronic valves" when he investigated the possibility of using the exotic vacuum-tube devices coming out of radar research to speed up the massive information-processing tasks needed by the Bletchley code-breakers.

@Turing learned his electronics from some of the best in the business -- the engineers at Bell Laboratories in New York (including one named @Shannon, a prodigy of a different kind, who will enter the story again).

His role at Bletchley wasn't Turing's only wartime contribution. He was sent over to America, at a time when it was indeed dangerous to take a North Atlantic cruise, to share crucial aspects of British cryptanalytic progress with American intelligence and to lend his intelligence to several American war-related scientific projects.

@Turing was involved in the writing of instruction tables that automatically converted human-written decimals to machine-readable binary digits. If basic operations like addition, multiplication, and decimal-to-binary conversion could be fed to the machine in terms of instruction tables, Turing saw that it would be possible to build up hierarchies of such tables. The programmer would no longer have to worry about writing each and every operational instruction, step by repetitive step, and would thus be freed to write programs for more complex operations.

That's what programmers do. They think of machines people might want to use, and figure out ways to describe those machines to general machines -- computers, that is.

The way the universal @Turing machine imitates other Turing machines is as automatic as the way our doubling machine multiplies the input by two. Assuming that the control unit of the device is capable of interpreting simple instructions -- something that had been a matter for toolmakers, not mathematicians since Babbage's time -- it is possible to encode a more complex list of instructions describing various Turing machines and put them onto the input tape, along with the starting position.

@Turing not only anticipated the fact that software engineering would end up more difficult and time-consuming than hardware engineering, but anticipated the importance of what came to be known as "debugging":

In America, @Turing was involved in another hypersecret project, this time involving voice encryption — what the spy novels call "scramblers."

@turing's public talks and private conversations indicated a strong belief that the cost of electronic technology would drop while its power as a medium for computation would increase in the coming decades.

@turing's reports on the hardware and software design for ACE were ambitious, and if the machine he originally envisioned had been constructed as soon as it was designed, it would have put ENIAC to shame.