Book of the Week: Surfing Uncertainty: Prediction, Action, and the Embodied Mind
How do we humans think, feel, and dream? In this book, cognitive scientist Andy Clarks taps into the predictive processing theory of mind, which may provide a unified framework that explains thoroughly how minds work. The central idea is that our brains are just like prediction machines: based on prior experiences, they constantly generate internal models of the world (e.g. the sun rises from the east every morning) and make predictions of the future, and oftentimes these predictions themselves shape how we perceive the world. Clark takes us through a journey to discover the flow of human predictions that structuring the environment surrounding us and may form the basis of our own existence. Enjoy the read!
Neuro-Talk:
https://www.youtube.com/watch?v=RgWgfdLZciM
Remember the epic dress from a couple of years ago, which some people saw as black and blue while others saw as white and gold? That is your brain’s prediction at play. In this TED Talk, neuroscientist Georg Keller proves that, if you place the same snippet of cloth from that dress, and put it under different contexts, your brain automatically perceives it differently. Our brains are predictive by nature, and they constantly run internal models of the world we live in. As you will see in this video, those predictions are in fact so powerful that they can directly influence our own perceptions -- what we hear and what we see are not reflections of objective reality but tainted with expectations.
https://www.youtube.com/watch?v=pLcIT3kjsW0
How, exactly, do our brains make predictions? Say I had never seen a raven before. Just seconds ago, I saw a raven; it was black. Now I see another raven, and it is also black. While it may be too early for me to conclude that all ravens are black, my belief in whether many ravens are black will certainly change based on the two observations I’ve made. But how? Scholars have proposed that our brains constantly update predictions following Bayes’ theorem, a mathematical rule that provides a specific of revising existing predictions given new evidence. The video gives an intuitive, non-mathy understanding of what Bayes’ theorem is (don’t be scared!) and how our brains may function in a Bayesian fashion.
Neuro-Quest:
To Make Sense of the Present, Brains May Predict the Future / The Bayesian Brain: An Introduction to Predictive Processing
The Predictive Processing theory is one that seeks to unify the understanding of all aspects of our brains and minds. According to the theory:
1. At each level of a cognitive process, the brain generates models/beliefs about what information it should be receiving from the level below it.
2. As these predictions run from the top to lower-level sensory regions of the brain, the brain compares its predictions with the actual sensory input it receives from bottom-up.
3. When prediction errors -- discrepancies between your internal models and actual inputs -- arise, your brain would seek to minimize prediction errors in three ways: updating the internal models, engaging in actions that make your predictions better match the actual states of things, or deeming the errors unimportant and ignoring them.
4. The first article is a magazine-style account of the theory, whereas the second one gives a more technical introduction to this concept, with the highlight that it explains some everyday topics, from action and emotion to learning and memory under the Predictive Processing framework. Take a look.
Does autism arise because the brain is continually surprised?
(We previously linked this article in an issue on autism. Read the whole issue it here).
The theory of Predictive Processing, with its broad implication, has given a new explanation for autism. In the predictive-coding model, the brain decides whether to ignore the discrepancies between its predictions and real cases by assigning every expectation a precision. High precision, on the one hand, means: “This is important. Pay attention!” Low precision, on the other hand, says: “Just a fluke, never mind.” For neurotypicals, their brains start with high precisions but could gradually turn them down, but for autistic patients, the precisions may be stuck at a high level. So loud sounds on a busy street that may be seen as low-precision noises and easily ignored by a typical brain would constantly “surprise” an autistic brain. For a more in-depth account, read this article to find out.
Hallucinations are surprisingly common in our lives -- when is the last time you thought you heard your phone ‘ding’? Over the past few decades, scientists have developed many theories to explain the phenomenon, one of which related to the predictive nature of the brain: we constantly make predictions about the world, but sometimes our expectations can overplay, hence hallucinations. Read this article for more details about this explanation and relevant experiments.
