About this video

• Video Title: How the brain shapes reality - with Andy Clark
• Channel: The Royal Institution
• Speakers: Andy Clark
• Duration: 59:51

Overview

This video explores the concept of the brain as a prediction machine, arguing that our perception and experience of reality are actively constructed rather than passively received. Andy Clark discusses how our expectations and internal models shape our sensory input, leading to phenomena like illusions, phantom sensations, and even conditions like chronic pain and functional neurological disorders. The talk also touches upon the implications of this predictive processing framework for understanding consciousness and developing new therapeutic interventions.

Key takeaways

• Perception as Prediction: Our brains constantly make predictions about the world based on past experiences and internal models. Sensory information is used to update these predictions, and often, what we perceive is the result of this predictive process rather than direct sensory input.
• The Brain as a Prediction Machine: This is the central theme, positing that the brain's primary function is to predict future states and the present sensory input, using a vast internal model of the world.
• Prediction Error: When sensory information contradicts a prediction, a "prediction error" signal is generated. These errors are crucial for updating the brain's model and learning, but their misinterpretation or misweighting can lead to altered perceptions.
• Controlled Hallucination: Perception is described as a "controlled hallucination," where the brain generates a model of reality and then uses sensory input to keep this model in check. This explains why illusions like the hollow mask effect occur.
• Precision Weighting: The brain constantly balances the influence of predictions versus sensory evidence. This "precision weighting" determines how much we rely on our internal models versus incoming data, and imbalances can lead to various perceptual and experiential differences.
• Applications in Psychiatry and Medicine: The predictive processing framework offers insights into conditions like chronic pain, autism spectrum disorder, and functional neurological disorders, suggesting that imbalances in precision weighting or persistent prediction errors play a significant role.
• Therapeutic Interventions: Understanding the brain as a prediction machine opens avenues for interventions like placebo effects, pain reprocessing therapy, and virtual reality therapy, which aim to alter maladaptive predictions and improve subjective experience.

The hollow mask illusion demonstrates how the brain's predictions can override sensory information. When viewing a mask that is concave (curving inward), our brain, based on the strong prior prediction that faces are convex (curve outward), interprets the concave shape as convex. Even though the visual information clearly shows a concave surface, the brain's prediction that it's seeing a normal, convex face is so powerful that it enforces that perception. This happens because the brain prioritizes its established models of the world over contradictory sensory input in certain situations, essentially creating a "controlled hallucination" where our expectation shapes what we perceive.

Phantom phone vibrations are the sensation of a phone vibrating in one's pocket when it is not actually vibrating, or even when it's turned off or not present.

The speaker suggests that these arise from the predictive brain in the following way: The brain has a strong, often unconscious, prediction that a phone will vibrate when a call or message comes through, especially if the person is expecting one. This prediction is so powerful that it can generate the sensation of vibration, even in the absence of actual physical stimuli. This is a form of prediction error where the prediction of a vibration is generated internally by the brain, overriding the sensory information that there is no actual vibration occurring. Factors like stress and caffeine can exacerbate this by influencing the neurochemical systems involved in prediction.

The speaker suggests that in autism spectrum disorder, there might be an "overweighting" of sensory evidence. This means that the brain assigns too much importance or confidence to the incoming sensory information and not enough to its internal predictions or models of the world.

Here's how precision weighting relates to this:

• Normal Function: In typical brains, precision weighting helps balance predictions with sensory input. If predictions are highly confident and sensory data is noisy, the brain relies more on predictions. If sensory data is clear and novel, it updates predictions accordingly.
• Autism Spectrum Disorder: In ASD, the theory proposes that sensory information is given excessively high precision. This means that the brain struggles to filter or down-weight sensory input, leading to a more overwhelming and potentially chaotic experience of the world.
• Consequences: This can make it difficult to detect faint patterns in noisy environments (as everything is given equal weight) and may lead to challenges in social situations where subtle cues need to be interpreted within a broader context. The speaker also notes that this is not due to a weakened influence of models, but rather a heightened influence of sensory information.

Pain Reprocessing Therapy (PRT) is a therapeutic approach suggested to work within the predictive processing framework by reframing the experience of pain.

Here's how it's described:

• The Problem: In chronic pain, the brain's "warning light" (the sensation of pain) can become a malfunctioning alarm. The prediction of imminent bodily danger associated with an action gets locked in, even when there's no longer a structural or peripheral cause that warrants such intense and persistent alarm signals. The pain itself becomes the problem, not necessarily the underlying physical damage.
• How PRT Works: PRT aims to push back against these locked-in, maladaptive predictions. It involves:
  1. Reframing: Educating individuals about how their brains construct pain and how expectations can influence pain perception. This helps to reframe the pain not as a sign of ongoing physical damage, but as a potentially false alarm.
  2. Installing New Predictions: Encouraging patients to gradually engage in activities they might have been avoiding due to pain. The idea is that by doing more, the brain begins to form new, less threatening predictions about bodily states and actions.
• The Mechanism: By engaging in activities despite the pain, the brain receives feedback that contradicts the prediction of severe danger. This can lead to a reduction in the perceived intensity of pain, creating a positive feedback loop where doing more makes the pain feel less and the new, less alarming predictions become more entrenched. Essentially, it helps to re-balance the precision weighting, allowing the brain to rely less on the dire prediction of harm and more on the evidence that engaging in activity is safe.

The brain isn't just receiving info, it's predicting reality! 🤯 Andy Clark explains how our brains are prediction machines, constructing our experience through expectations & models. This sheds light on everything from illusions & phantom sensations to chronic pain & mental health. Perception = controlled hallucination! #Neuroscience #CognitiveScience #BrainScience #Prediction

[Link to video if applicable]