> This means it's very different from programming where we have a dumb machine that is, at it's best ability, able to look at two numbers and tell you which one

> This means it's very different from programming where we have a dumb machine that is, at it's best ability, able to look at two numbers and tell you which one is bigger. And even that can get wonky. Do a search for "What Every Computer Scientist Should Know About Floating-Point Arithmetic" by David Goldberg. * https://en.wikipedia.org/wiki/Floating-point_error_mitigation https://en.wikipedia.org/wiki/Floating-point_error_mitigatio... * https://en.wikipedia.org/wiki/Machine_epsilon https://en.wikipedia.org/wiki/Machine_epsilon