“Computer simulation often works fine if we assume nothing more than Newton’s laws at the atomic scale, even though we know that really we should be using quantum, not classical, mechanics at that level. But sometimes approximating the behaviour of atoms as though they were classical billiard-ball particles isn’t sufficient. We really do need to take quantum behaviour into account to accurately model chemical reactions involved in industrial catalysis or drug action, say. We can do that by solving the Schrödinger equation for the particles, but only approximately: we need to make lots of simplifications if the maths is to be tractable. But what if we had a computer that itself works by the laws of quantum mechanics? Then the sort of behaviour you’re trying to simulate is built into the very way the machine operates: it is hardwired into the fabric. This was the point Feynman made in his article. But no such machines existed. At any rate they would, as he pointed out with wry understatement, be ‘machines of a different kind’ from any computer built so far. Feynman didn’t work out the full theory of what such a machine would look like or how it would work – but he insisted that ‘if you want to make a simulation of nature, you’d better make it quantum-mechanical’.”