The universal machine

In 1936 Alan Turing publishes a paper on computable numbers. In it he describes a simple machine that reads and writes marks on a tape according to rules. He shows that one such machine, given the right rules, can do anything any particular machine of that kind can do. That is the universal machine.

The contribution is the idea of software. You do not build a new physical machine for each job. You write instructions and the same hardware carries them out. Every computer after this, from the room-sized machines of the 1940s to a phone, is an attempt to build Turing's paper in switches.

Turing did not build the first computer. Konrad Zuse, Tommy Flowers with Colossus, and the ENIAC team build working electronic machines in the following decade. John von Neumann's stored-program design is the architecture they settle on. The paper is the clean statement of what those projects are for.

The machine in the paper is a thought experiment. It has no screen and no speed. The contribution is the proof that instructions are enough.