In the 1830s Charles Babbage designed a calculating machine directed by punched cards, able to store intermediate results and to branch depending on a condition: the Analytical Engine. It was to run on steam and consist of tonnes of brass gearing. The design was finished; the funding was not, and the machine went unbuilt in his lifetime.

In 1843 Ada Lovelace translated into English a French paper on the machine by the Italian mathematician Luigi Menabrea, and appended notes three times the length of the original. Her Note G sets out the step-by-step sequence the engine would follow to generate Bernoulli numbers — with loops and variables, and it is generally counted as the first published computer program. The more lasting contribution is in Note A: she writes that what the machine operates on is not numbers but symbols, so that anything representable in symbols — sound, and therefore music — could be operated on too. That is the conceptual leap from calculator to general-purpose computer.

In the same notes Lovelace remarks that the engine 'has no pretensions to originate anything' and can only do what we know how to order it to do. In 1950 Turing cited that sentence by name and argued against it, so the oldest objection in the artificial-intelligence debate and its oldest answer come from the same text. A section of the Analytical Engine was built at the Science Museum in London in 1991, faithful to Babbage's drawings, and it worked — the design proved correct a hundred and fifty years after it was drawn.