How to Change Your Mind

condition I could pull it off, but I’d found that even in the middle of the
journey it was possible to summon oneself to a semblance of normality
for a few moments at a time.
Loaded on my laptop was a brief video of a rotating face mask, used in
a psychological test called the binocular depth inversion illusion. As the
mask rotates in space, its convex side turning to reveal its concave back,
something remarkable happens: the hollow mask appears to pop out to
become convex again. This is a trick performed by the mind, which
assumes all faces to be convex, and so automatically corrects for the
seeming error—unless, as a neuroscientist had told me, one was under
the influence of a psychedelic.
This auto-correct feature is a hallmark of our perception, which in the
sane, adult mind is based as much on educated guesswork as the raw data
of the senses. By adulthood, the mind has gotten very good at observing
and testing reality and developing confident predictions about it that
optimize our investments of energy (mental and otherwise) and therefore
our survival. So rather than starting from scratch to build a new
perception from every batch of raw data delivered by the senses, the mind
jumps to the most sensible conclusion based on past experience
combined with a tiny sample of that data. Our brains are prediction
machines optimized by experience, and when it comes to faces, they have
boatloads of experience: faces are always convex, so this hollow mask
must be a prediction error to be corrected.
These so-called Bayesian inferences (named for Thomas Bayes, the
eighteenth-century English philosopher who developed the mathematics
of probability, on which these mental predictions are based) serve us well
most of the time, speeding perception while saving effort and energy, but
they can also trap us in literally preconceived images of reality that are
simply false, as in the case of the rotating mask.
Yet it turns out that Bayesian inference breaks down in some people:
schizophrenics and, according to some neuroscientists, people on high
doses of psychedelics drugs, neither of whom “see” in this predictive or
conventionalized manner. (Nor do young children, who have yet to build
the sort of database necessary for confident predictions.) This raises an
interesting question: Is it possible that the perceptions of schizophrenics,
people tripping on psychedelics, and young children are, at least in