Key Moments
Constructing Self and World: A Conversation with Shamil Chandaria (Episode #320)
Sam HarrisKey Moments
Brain constructs reality via predictive models, minimizing error. Psychedelics, meditation, and non-duality offer insights into self and experience.
Key Insights
The brain functions as a predictive machine, constantly constructing a model of the world rather than passively receiving sensory data.
Bayesian inference, specifically approximate Bayesian inference, is central to how the brain makes predictions and updates its model of reality.
Hierarchical predictive processing suggests a layered model where higher levels generate expectations and lower levels report prediction errors.
Subjective experience, including vision, arises from this internal generative model, calibrated by sensory input and prior probabilities.
Psychedelics and meditation can alter the brain's predictive mechanisms, offering profound shifts in the perception of self and reality.
The concept of a 'phenomenal self model' is key to understanding our subjective experience of being an agent in the world.
THE BRAIN AS A PREDICTIVE MACHINE
The prevailing view of the brain is being overturned; instead of processing incoming sensory data bottom-up, the brain actively constructs a model of the world. From a first-principles perspective, the brain, enclosed in the skull, only receives noisy, time-series data from the nervous system. It must infer the causes of this data to interact with the environment. This process is fundamentally a statistical inference problem, akin to Bayesian inference, aiming to determine the probability of what's causing the sensory input.
BAYESIAN INFERENCE AND APPROXIMATE SOLUTIONS
Bayesian inference provides a framework for calculating the probability of a hypothesis given the observed data, incorporating both the likelihood of the data under the hypothesis and prior probabilities. For instance, determining the probability of seeing a tree involves the likelihood of specific sensory data if it were a tree, multiplied by the prior probability of trees existing. Since formally solving Bayesian inference is computationally explosive, the brain employs approximate Bayesian inference, using a generative model to simulate expected sensory data and minimize prediction errors.
HIERARCHICAL PREDICTIVE PROCESSING
The brain likely operates through hierarchical predictive processing. Higher cortical areas generate predictions (priors) about what to expect, which flow down to lower levels. The feed-forward pathways primarily transmit prediction errors—the discrepancies between predictions and actual sensory input. These errors then flow upwards, prompting adjustments in the predictive models at higher levels to minimize future errors. This creates a dynamic, multi-layered simulation calibrated by incoming data.
THE CONSTRUCTION OF VISION AND SUBJECTIVE REALITY
Our subjective experience, such as vision, is not a direct perception of reality but the output of this internal generative model. For example, we experience a clear, colorful visual scene despite our eyes' limited foveal vision and the blind spots created by saccades. The brain 'fills in' the gaps and presents a coherent, stable simulation. This suggests that our perception is a 'controlled hallucination,' heavily influenced by our prior expectations and the constant recalibration of our models.
THE ROLE OF MEDITATION AND PSYCHEDELICS
Practices like meditation and experiences with psychedelics can profoundly alter our subjective reality by influencing the brain's predictive mechanisms. For instance, psychedelics are thought to promote neuroplasticity and may disrupt or alter these predictive models, leading to novel perceptions. Meditation, on the other hand, can involve attention and mindfulness training, potentially affecting the weighting of prediction errors versus priors, leading to a deconstruction of rigid self-models and a shift in perspective.
THE PHENOMENAL SELF MODEL
At the higher levels of the predictive processing hierarchy, a 'phenomenal self model' is hypothesized to emerge. This model represents our sense of being a conscious agent in the world. The flexibility of our predictive models, influenced by practices like meditation or psychedelics, can lead to a loosening or deconstruction of this self model. This shift has implications for understanding non-duality, love, gratitude, and ultimately, human flourishing, by potentially reducing suffering rooted in fixed self-perceptions.
Mentioned in This Episode
●Supplements
●Software & Apps
●Companies
●Organizations
●Concepts
●People Referenced
Common Questions
The Free Energy Principle, proposed by Karl Friston, suggests the brain functions as a predictive machine. It constantly constructs internal simulations of the world to minimize prediction errors between its expectations and incoming sensory data, effectively calibrating its model through approximate Bayesian inference.
Topics
Mentioned in this video
Center founded by Shamil Chandaria in 2019 at Imperial College London, focusing on psychedelic research.
University where Shamil Chandaria is funding research on the neuroscience of meditation.
A retreat center in the UK where Shamil Chandaria spent significant time, influenced by the teachings of Rob Burbea.
Institute at Oxford University that Shamil Chandaria helped to endow in 2018.
Institution where Shamil Chandaria is a research fellow and helped found the Center for Psychedelic Research.
Institution where Shamil Chandaria earned his PhD in mathematical modeling of economic systems.
A large language model developed by Google AI, mentioned as an example of sophisticated AI with many layers.
A large language model developed by OpenAI, mentioned as an example of sophisticated AI with many layers.
A DeepMind AI system that achieved superhuman performance in games like chess, shogi, and Go using self-play.
A DeepMind AI program that defeated the world champion in the game of Go, cited as a significant AI breakthrough.
Philosopher who has done significant work on the concept of the phenomenal self model and its implications for consciousness.
Neuroscientist who previously led the Center for Psychedelic Research at Imperial College London and collaborated on predictive models of psychedelic action.
Neuroscientist known for developing the Free Energy Principle and his work on computational neuroscience and predictive processing.
Enlightenment philosopher whose work on epistemology and the categories of understanding supports the idea that we only know appearances, not reality itself.
Resident teacher at Gaia House and author of 'Emptiness, the Waking Eye,' whose teachings on emptiness greatly influenced Shamil Chandaria's meditation practice.
Cognitive psychologist and collaborator with Daniel Kahneman, known for his work on heuristics and biases.
Nobel laureate psychologist known for his work on decision-making and cognitive biases, particularly 'Thinking, Fast and Slow'.
Host of the Making Sense podcast, discussing neuroscience, AI, and philosophy with Shamil Chandaria.
Philosopher with whom Shamil Chandaria has had conversations, sparking Sam Harris's interest.
Research fellow at Imperial College London, related to Shamil Chandaria's work on psychedelics and computational models.
Institution where Shamil Chandaria completed a Master's in Philosophy, developing interests in philosophy of science, biology, neuroscience, and ethics.
Non-dual meditation teacher with whom Shamil Chandaria has been working, appreciating his broad experience and ability to integrate different styles.
Serotonin receptors involved in mediating the effects of psychedelics like LSD and psilocybin, with potential intracellular actions contributing to neuroplasticity.
A fundamental theorem in probability theory that describes how to update beliefs in light of new evidence. It's central to Bayesian inference.
A machine learning technique whose mathematics shares roots with stochastic optimal control, used in AI and understanding the brain.
A computational model developed by Shamil Chandaria, Robin Carhart-Harris, and Karl Friston that uses a predictive processing framework to explain the effects of psychedelics.
A field of study concerned with the quantification, storage, and communication of information, relevant to understanding the brain's computational processes.
A mantra-based meditation technique that Shamil Chandaria initially practiced when he started meditating 35 years ago.
A philosophy and social movement that uses evidence and reason to determine the most effective ways to improve the world, influencing Shamil Chandaria's philanthropic approach.
A core concept in Buddhism referring to the lack of inherent existence or self-nature of phenomena, discussed in relation to meditation and Shamil Chandaria's influences.
A common cognitive bias identified by Kahneman and Tversky, where individuals tend to overlook the prior probability (base rate) of an event when making judgments.
A model used in machine learning and neuroscience that learns the underlying probability distribution of data, enabling it to generate new data samples or predict future observations.
A theoretical concept, associated with Thomas Metzinger, suggesting that the subjective experience of 'self' is a model constructed by the brain.
A theoretical framework proposed by Karl Friston that describes how biological systems maintain their structure and function by minimizing surprise (or free energy). It is considered the closest thing to a general algorithm for brain function.
A statistical method for updating probability estimates based on new evidence, proposed as a core mechanism for how the brain makes sense of sensory data.
A theory proposing that the brain processes information in a hierarchical manner, with higher levels making predictions about the outputs of lower levels and adjusting based on prediction errors.
The brain's ability to reorganize itself by forming new neural connections throughout life, a process influenced by psychedelics.
A DeepMind AI system for predicting protein structures, mentioned as a breakthrough in AI.
University where Shamil Chandaria is funding research on the neuroscience of meditation.
A naturally occurring psychedelic compound found in certain mushrooms, discussed for its impact on perception and neuroplasticity.
Mentioned by Sam Harris as a substance he has recently had an experience with, related to the discussion on psychedelics.
Lysergic acid diethylamide, a powerful psychedelic substance discussed in the context of its effects on perception and neuroplasticity.
More from Sam Harris
View all 140 summaries
10 minThe War Was Necessary. The Way Trump Did It Wasn’t.
1 minBen Shapiro Knows Better
1 minMost People Know as Much About Politics as They Do Football… Not Much
2 minTrump is Going to Burn it All Down...What Are We Going to Build Instead?
Found this useful? Build your knowledge library
Get AI-powered summaries of any YouTube video, podcast, or article in seconds. Save them to your personal pods and access them anytime.
Try Summify free