Why surprise is good for the brain.


There is a new unified theory of the brain that is reported in NewScientist and introduced to me here (Reverendbayes’s weblog). Karl Friston, a neuroscientist, from University College London has come up with this new theory based on mathematics.

The general idea behind the theory is that the brain is a probability calculator (machine) that is constantly determining the probability of the world that is entering into it via the senses. The name given to this theory is the “Bayesain brain”. For those of you not familiar with the term Bayesian from wiki we learn: Bayesian refers to methods in probability and statistics named after the Reverend Thomas Bayes (ca. 1702–1761), in particular methods related to:

– the degree-of-belief interpretation of probability, as opposed to frequency or proportion or propensity interpretations; or

-Bayes’s theorem on conditional probablity.

This idea is built on earlier work in the field (Geoffery Hinton, Terry Sejnowski) – like everything else in the world.

Now the new twist on the general Bayesian brain theory is the addition of the idea of free energy. Free energy, which you might remember from the first energy law, in wiki is defined as “the amount of mechanical (or other) work that can be extracted from a system,”

Hinton, a colleague of Friston, from the same university, found a link between the prediction made by his artificial neural networks and the actual data that it has taken in (senses) and free energy. He realized if he considered the prediction error as free energy then minimized it the neural network could solve difficult problems.

Friston then took the next leap and hypothesized that the brain’s constant updating of probabilities could be thought of minimizing free energy. So here is quote from the article (with some spelling changes). “It’s about minimizing surprise,” he explains. “Mathematically, free energy is always bigger than surprise, therefore if you can minimize free energy you can avoid surprising encounters with the world.”

So the general concept that Friston is suggesting is that the brain, through all the classic mechanisms known to neuroscientists (firing of neurons, rate of firing, re-wiring of the brain (plasticity), neurotransmitters, gene expression, etc), is always changing to suppress prediction error (by minimizing free energy). But this does not just stop at the molecular/neurochemical level, but also at higher levels, including what you pay attention to, and up all the way to the choices you make in life.

So this is where I am going to make a leap (not that far – hopefully I don’t fall) if you want to keep learning and increase plasticity (which is good for the brain) then you need to constantly subject yourself to surprise.

A very simple idea – keep your life full of surprises – built from a complex theory of the brain (Bayesian Brain; which I only gave a very general overview), might be one more ticket to keeping your brain plastic and healthy.

Now you can not just do a task to be surprised (be the definition of surprise) at best you can do activities that increase the probability of your being surprised. Taking the same bike or running route every day offers no new information, very little potential to surprise. However, if you take a new route, and open yourself up to new sights, new experiences, you increase your chance of novelty and surprise.

What are you going do today to give your brain a chance of being surprised?