
At Google X, the tech giant’s “moonshot factory,” the mantra is #MonkeyFirst. The idea is that if you want to get a monkey to recite Shakespeare on a pedestal, you start by training the monkey, not building the pedestal. Because training the monkey is the hard part. Anyone can build a pedestal.
Putting together the ideas of Markov Chain and Monte Carlo, MCMC is a method that repeatedly draws random values for the parameters of a distribution based on the current values. Each sample of values is random, but the choices for the values are limited by the current state and the assumed prior distribution of the parameters. MCMC can be considered as a random walk that gradually converges to the true distribution.
… that we would all hesitate to acknowledge about ourselves that we can easily pick up on in others: Pretty much everything we desire (beyond the core biological impulses) has nothing to do with what we want, but with what we have been conditioned to think we should want.