Reality is a Controlled Hallucination

Summary

The video explores the idea that our perception of reality is actually a controlled hallucination created by our brains. It discusses how our brains predict and interpret sensory input, often aligning it with past experiences. The script delves into the implications of this theory for understanding hallucinations, psychology, and the nature of existence. Through examples like the rubber hand illusion and magic mushrooms, it highlights how our brains can be tricked into seeing things that aren't real, ultimately suggesting that reality as we know it is a persistent illusion.

Takeaways

• 🪞 The script starts with a personal account of an unusual experience where the narrator's reflection was missing from the faucet, leading to a realization of having six fingers and a distorted perception of time.
• 🧠 It introduces the concept that our everyday perceptions are controlled hallucinations, suggesting that what we see is not a direct projection of reality but a projection of our own minds.
• 🔮 Neuroscience supports the idea that the brain functions as a prediction machine, constantly generating expectations about what it will perceive based on past experiences and current context.
• 🌳 The script uses the example of walking in a forest to explain how the brain matches sensory input with internal predictions, leading to the perception of objects like trees.
• 🔄 The concept of prediction error is introduced, which occurs when sensory input does not align with the brain's predictions, triggering an update to the brain's internal model of the world.
• 📊 The brain is described as a hierarchical prediction machine with top-down predictions from higher layers and bottom-up signals from lower layers, which can lead to agreement or prediction error.
• 🔄 The process of prediction error minimization (PEM) is explained as a feedback loop where the brain adjusts its model to better align with reality when there's a mismatch between predictions and sensory input.
• 🌐 The script compares the brain's predictive process to video compression, highlighting the efficiency of only analyzing sensory information that isn't already obvious to us.
• 🐞 An evolutionary perspective is provided, explaining that our brains haven't evolved to present an accurate depiction of reality but rather a user-friendly interface optimized for survival and reproduction.
• 🦋 The Australian Jewel Beetle example illustrates how an overly optimized nervous system can lead to perceptions that are not representative of objective reality.
• 🎨 The script discusses various illusions and hallucinations, such as the blind spot illusion, the rubber hand illusion, and the Ganzfeld experiment, to demonstrate how easily our senses can be manipulated.
• 🍄 The effects of magic mushrooms on perception are explained, showing how they can disrupt the balance between top-down predictions and bottom-up senses, leading to vivid and immersive hallucinations.

Q & A

What was the protagonist's initial reaction when they saw their reflection in the faucet?

The protagonist panicked, thinking they were going crazy, because they realized their reflection was missing.

How did the protagonist confirm their suspicion about the unusual situation?

The protagonist counted their fingers and found they had six, which confirmed their suspicion of an anomaly.

What does the script suggest about our everyday perceptions?

The script suggests that our everyday perceptions are controlled hallucinations, influenced by our brain's predictions based on past experiences and current context.

What is the brain's role in the predictive processing model of neuroscience?

In the predictive processing model, the brain functions as a prediction machine, constantly generating expectations about what it will perceive and adjusting these predictions based on sensory input.

How does the brain handle situations where predictions do not align with sensory input?

When predictions do not align with sensory input, it results in a prediction error, which the brain uses to update its internal model of the world through a process called prediction error minimization (PEM).

What is the hierarchical structure of the brain's predictive processing?

The brain is considered a hierarchical prediction machine with higher layers making top-down predictions about the causes of sensory input from lower layers, which provide bottom-up signals.

How does the video compression analogy relate to the brain's predictive processing?

Just as video compression only transmits changes in image frames, the brain is more energy efficient by only analyzing sensory information that isn't already predicted, thus minimizing the information it needs to process.

What evolutionary purpose does the brain's predictive processing serve?

The brain's predictive processing serves an evolutionary purpose by providing a user-friendly interface optimized for survival and reproduction, rather than an accurate depiction of reality.

How does the Australian Jewel Beetle example illustrate the concept of controlled hallucinations?

The Australian Jewel Beetle example shows that the beetles' brains evolved to recognize mates based on specific features, but they were unable to distinguish between these features in beer bottles and females, illustrating that their perceptions were simplified for survival rather than accuracy.

What are some examples of illusions or experiments that demonstrate the brain's susceptibility to hallucinations?

Examples include the blind spot illusion, the rubber hand illusion, the Ganzfeld experiment, and the effects of magic mushrooms, all of which show how easily our perceptions can be manipulated.

How does the use of magic mushrooms affect the brain's predictive processing?

Magic mushrooms contain psilocybin, which, when metabolized into psilocin, disrupts the balance between top-down predictions and bottom-up senses, leading to an overactive top-down network and the perception of patterns and connections that aren't real.

Outlines

00:00

🔍 The Illusion of Self and Reality

The script begins with a personal anecdote about a startling realization of seeing one's reflection in a faucet without recognizing oneself, leading to a six-fingered count and a distorted perception of time. This experience introduces the concept that our everyday perceptions might be 'controlled hallucinations' rather than direct reflections of reality. It delves into the idea that our brain functions as a 'prediction machine,' constantly generating expectations based on past experiences and current context. When predictions align with sensory input, the brain assumes correctness, but discrepancies lead to 'prediction errors,' which are essential for updating our internal model of the world through a process known as prediction error minimization (PEM). The narrative challenges the conventional understanding of perception as a direct reflection of reality, suggesting instead that our brain creates a subjective experience steered by the real world.

05:04

🧠 The Brain's Hierarchical Prediction Model

This paragraph explores the brain as a hierarchical prediction machine with top-down and bottom-up signals. Top-down predictions from higher brain layers anticipate broader causes of sensory input, while bottom-up signals from lower layers provide immediate sensory data. When these predictions and signals align, the brain continues with its model; when they don't, a prediction error occurs, prompting an update to the internal model. The script uses the example of walking in a forest and encountering a blue tree to illustrate how prediction errors can refine our perception. It also draws a parallel between this brain function and video compression technology, which updates only the parts of an image that change, minimizing the need for processing raw data. The paragraph concludes with the evolutionary rationale for this model, suggesting that our brains prioritize survival and reproduction over an accurate depiction of reality, using the Australian Jewel Beetle's mating habits as an example of how perception can be misled by evolutionary shortcuts.

10:07

🌈 The Mutable Nature of Perception

The final paragraph discusses the mutable and easily manipulated nature of human perception. It highlights that our senses can deceive us, as evidenced by illusions and experiments such as the blind spot illusion, the rubber hand illusion, and the Ganzfeld experiment. These examples demonstrate how our brain can be led to perceive things that aren't real, such as feeling pain from a rubber hand or hallucinating in a sensory deprivation state. The paragraph also touches on the effects of psychedelic substances like magic mushrooms, which disrupt the balance between top-down predictions and bottom-up senses, leading to vivid and immersive hallucinations. The script concludes by emphasizing that all sensory experiences are internal and that our perception of reality is a 'controlled hallucination' maintained by sensory input, challenging us to reconsider the nature of our existence and the reality we perceive.

Keywords

💡Hallucination

Hallucination refers to the perception of something that is not present in reality, often resulting from the brain's attempt to make sense of incomplete or ambiguous sensory information. In the context of the video, hallucinations are likened to our everyday perceptions, suggesting that what we see is not a direct reflection of reality but rather a 'controlled hallucination' created by our brain based on predictions and past experiences.

💡Neuroscience

Neuroscience is the scientific study of the nervous system and brain functions. The video discusses a model in neuroscience that describes the brain as a 'prediction machine,' constantly generating expectations about sensory inputs. This concept is central to the video's theme, illustrating how our perceptions are shaped by the brain's predictive processes.

💡Prediction Error

A prediction error occurs when there is a discrepancy between the brain's predictions and the actual sensory input it receives. The video uses this concept to explain how the brain updates its internal model of the world, adjusting its predictions to better align with reality. An example given is the unexpected appearance of a blue tree in a forest, which would trigger a prediction error and subsequent update of the brain's model.

💡Hierarchical Prediction Machine

The hierarchical prediction machine is a model suggesting that the brain processes information through a series of levels, with higher layers making broader predictions and lower layers providing detailed sensory inputs. The video explains how this hierarchical structure allows for top-down predictions and bottom-up signals to be compared, leading to either agreement or prediction errors.

💡Prediction Error Minimization (PEM)

Prediction Error Minimization (PEM) is a process by which the brain adjusts its internal model to minimize the difference between its predictions and actual sensory inputs. The video describes PEM as a feedback loop, where the brain makes predictions, receives sensory input, and adjusts its model accordingly to improve accuracy and efficiency over time.

💡Top-Down Predictions

Top-down predictions refer to the brain's process of making broader, higher-level assumptions about the world based on previous experiences and context. The video explains how these predictions are used to interpret sensory information, such as expecting to see a green, brown tree-shaped object when walking in a forest.

💡Bottom-Up Signals

Bottom-up signals are the sensory inputs received from the environment that the brain uses to compare with its top-down predictions. The video uses the example of seeing a tree in a forest, where the actual visual input from the eyes (the bottom-up signal) is matched against the brain's prediction to confirm the perception of a tree.

💡Evolutionary Perspective

The evolutionary perspective in the video suggests that our brains have evolved to provide a user-friendly interface for survival and reproduction, rather than an accurate depiction of reality. It uses the example of the Australian Jewel Beetle to illustrate how evolutionary pressures have shaped perception to be efficient rather than objectively accurate.

💡Controlled Hallucination

Controlled hallucination is a term used in the video to describe the brain's process of creating perceptions that are not direct projections of reality but are instead influenced by the brain's predictions and past experiences. The video argues that our everyday perceptions are a form of controlled hallucination, maintained by sensory input.

💡Psilocybin

Psilocybin is a psychoactive compound found in certain mushrooms, which the video mentions in the context of inducing hallucinations. It is converted into psilocin in the body, which then interacts with serotonin receptors in the brain, disrupting the balance between top-down predictions and bottom-up senses and leading to the vivid and immersive perceptions characteristic of a psychedelic trip.

💡Ganzfeld Experiment

The Ganzfeld experiment is a sensory deprivation technique mentioned in the video, where participants are exposed to a uniform visual field and white noise to induce hallucinations. The video describes this experiment as revealing the brain's tendency to create uncontrolled perceptions when sensory input is limited, demonstrating the brain's role in shaping our perceptions even in the absence of external stimuli.

Highlights

Our everyday perceptions are controlled hallucinations, supported by modern neuroscience.

The brain functions as a prediction machine, generating expectations about perceptions based on past experiences and current context.

Prediction errors occur when sensory input doesn't align with brain predictions, leading to updates in the brain's internal model.

The brain is a hierarchical prediction machine with top-down and bottom-up signals.

Prediction error minimization (PEM) is a feedback loop used by the brain to align with reality.

Our brains prioritize user-friendly interfaces optimized for survival over accurate depictions of reality.

The Australian Jewel Beetle example illustrates how brains can misinterpret sensory input for survival.

Our perceptions present simplified versions of reality, similar to how computer icons represent complex code.

Hallucinations can occur when sensory input is manipulated or when the balance between top-down predictions and bottom-up senses is disrupted.

The blind spot illusion demonstrates how the brain can fail to perceive certain visual information.

The rubber hand illusion shows how the brain can be tricked into perceiving a fake hand as one's own.

The Ganzfeld experiment induces hallucinations through sensory deprivation.

Magic mushrooms can cause hallucinations by disrupting the balance between top-down and bottom-up sensory processing.

Google DeepDream uses a neural network to find and enhance patterns in images, similar to the brain's top-down network.

Psychedelic experiences can lead to a flood of unfiltered perceptions, altering the usual sensory processing.

Predictive processing theory suggests that all our perceptions are internal and only represent reality.

Transcripts

I was washing my hands in the kitchen. Everything  seemed normal at first. I turned on the faucet  

and as expected water came out. The faucet was  made from a very shiny metal, reflecting the  

entire room. As I washed my hands I looked at the  reflection in the faucet and suddenly realized  

something very important was missing... Myself. I  panicked, thinking I was going crazy. But I had a  

hunch about what was going on, so I counted my  fingers. Six. My suspicions were confirmed, but  

I couldn’t believe it. I looked at the microwave  clock and surely enough the numbers seemed normal,  

but they just made no sense. Everything  pointed to one explanation, but I just couldn’t  

believe it. It felt too real to be a dream. Every time you open your eyes, what you see,  

feel and hear happens entirely within your  own skull. Think about it: how can a bunch of  

electrical signals in your brain feel just as real  as the world around you? It turns out that much  

like optical illusions, psychedelic experiences or  schizophrenic episodes, our everyday perceptions  

are controlled hallucinations. It might sound  weird, but modern neuroscience strongly supports  

this idea. If this theory is as correct as we  think it is, it has massive implications for our  

understanding of hallucinations, psychology,  and even the nature of our own existence. 

We all intuitively think that what we see is  a direct projection of reality. It all seems  

pretty straight forward; your eyes take  an image of the world around you that you  

then see and interpret directly within your  brain. And that is how you see the world,  

right? Modern science predicts that it actually  goes the other way around. It says that the world  

as we see is not at all a projection of reality,  but actually a projection of our own minds,  

a controlled hallucination that is only steered  by the real world whenever it goes of course. 

This claim is backed up by a new model in  neuroscience that has recently been gaining  

more and more traction, although its underlying  principles have been around since the middle ages.  

Essentially this model suggests that our brain  functions as a prediction machine, constantly  

generating expectations about what it will  perceive based on past experiences and current  

context. These predictions include everything  you sense like shapes, sounds and smells. 

When these predictions align with the  sensory input from the environment,  

your brain assumes its predictions are correct  and it continues working with them. For instance,  

if you are walking in a forest  and you see a tree shaped object,  

your brain quickly matches the sensory input  with its internal prediction based on past  

experiences of seeing trees and you perceive it as  a tree without conscious effort and move on. 

However, when the prediction doesn’t  line up with the sensory input,  

this is known as a prediction error. Prediction  errors occur when something unexpected happens,  

such as hearing a sudden loud noise in a quiet  room or seeing an object that doesn’t fit with  

your previous experiences, like a blue tree. More specifically, the brain is considered to be  

a hierarchical prediction machine. This means that  there are levels of processing. Higher layers try  

to predict bigger picture causes of the sensory  input coming from lower layers, these are called  

the top-down predictions. So neurons at higher  levels encode predictions about the upcoming  

signal, which is continuously compared with the  signal received from lower levels, which are  

called the bottom-up signals. The information in  these top-down predictions and bottom-up signals  

could then lead to an agreement if our senses  agree with the prediction or a prediction error  

when they don’t agree. And this can happen  in any of the layers layer of the hierarchy. 

The brain then uses these prediction errors to  update its internal model of the world through a  

process called prediction error minimization  or PEM. It’s a bit like a feedback loop;  

the brain makes a prediction, receives sensory  input, and if there’s a mismatch, it adjusts  

its model to better align with reality. So in  a way prediction errors are whatever was left  

unexplained by our best predictions. This way our  brains get more accurate and more efficient over  

time. If you don’t fully understand the details,  don’t worry. The fundamentals will become clearer  

throughout the rest of the video, but let’s go  through an example to clear things up first. 

Let’s say you’re walking in a forest. From a  top-down perspective, you’re aware that you’re  

walking in a forest, so you’re predicting to see  a tree and so you would predict to see a green,  

brown, tree shaped object, but your eyes don’t  just accurately capture an image like your camera  

does. No, you see two separate images, each with  a big blind spot where your eye attaches to your  

optic nerve. Both of these images would also be  flipped and there are also some visible blood  

vessels. So your brain expects something similar  to this, even though you might see something  

closer to the original in your head. Now, if  this is the input your brain gets, then you’re  

not surprised at all, but If one of the trees  was blue, then somewhere along this chain the  

bottom-up sensations from your eyes would cause a  prediction error. This error would then be used to  

update your internal model of the world, so that  next time you wouldn’t be as surprised to see a  

blue tree in the forest and you would just dismiss  it as a blue tree, saving you time and energy. 

You might already recognize this idea of using  prediction to minimize information from video  

compression, because it works in a very similar  way. When you watch a video online not every  

single pixel is transmitted over the internet.  Instead, large portions of the image that remain  

static from one frame to the next are sent  only once. The video player then updates  

only the parts of the image that change, so it  doesn’t have to take in all the raw data from  

the video whenever there’s a new frame, but only  the data that it couldn’t have predicted based on  

the footage before it. Now, it turns out that  our brains use a very similar principle to be  

more energy efficient by only analyzing sensory  information that isn’t already obvious to us.

And from an evolutionary perspective this  efficiency is all that matters, so this model  

of our brains isn’t just backed by scientific  evidence, but there is also a very solid reason  

behind it based in evolution theory. According  to Hoffman our brains haven’t evolved to present  

us with an accurate depiction of reality. They  only provide us a user-friendly interface that’s  

optimized for survival and reproduction. A great  example of this is the Australian Jewel Beetle,  

which evolved to identify mates by recognizing  the shiny, dimpled texture of female beetles.  

Throughout history, if a male beetle wanted to  reproduce, the only thing he had to understand  

was if something was shiny, dimply and brown,  but clearly this isn’t perfect, because it  

turned out that the males were in love with the  Australian beer bottles, because they were shiny,  

dimply and brown. The beetles were so in love  that Australia had to change its bottles to save  

them from going extinct. It always seemed  like the beetles saw reality for what it was,  

but the beer bottle showed that they only ever  understood the bare minimum to thrive. It’s  

internal representation wasn’t able to distinguish  between beer bottles and females. So it's tiny,  

overly optimized nervous system, wasn’t equipped  with the tools to see reality for what it was. 

Of course we can spot the difference between  a beer bottle and a female beetle, but just  

like the beetle, our brains haven’t evolved to  prioritize understanding objective reality either.  

The real world is overwhelmingly complex, so we  have to make a compromise and make assumptions  

that aren’t completely accurate. Just like  computer icons hide the complex code underlying  

our machines, our perceptions present simplified  versions of reality that help us navigate the  

world. Physical objects and time are like desktop  icons. They’re simplifications. They don’t  

accurately represent the underlying complexity of  our world, but they’re useful for our purposes. 

And because we never really see objective reality,  our senses can be easily manipulated and we can  

even start to see things that aren’t real. There  are some great examples like the blind spot  

illusion, the rubber hand illusion, the disturbing  Ganzfeld experiment and magic mushrooms. 

Take the blind spot illusion for example. If you  cover your left eye and look at the black dot,  

you’ll find that if you move your head towards  or away from the screen, the star will suddenly  

disappear, because it perfectly lines up with  where your optic nerve attaches to your eye. You  

could get the same effect by covering your right  eye and looking at the star instead of the dot. 

Another great example is the rubber hand  illusion, where participants see a rubber  

hand being stroked in sync with their own hidden  hand. This seems to be enough for your brain  

to assume that the rubber hand is your own. So when the rubber hand is suddenly hit with  

a hammer, you can feel the impact even though  the rubber hand has no physical connection to  

your body. Your brain essentially predicts  that the impact will hurt, so you feel pain  

even before it could actually reach your brain. A disturbing and more revealing example is the  

Ganzfeld experiment where participants are seated  in a comfortable chair in a soundproof room. They  

wear headphones playing white noise and halved  ping-pong balls over their eyes with a red light  

source in front of them, so that they only  see diffused red light and only hear nothing  

but noise. This setup creates a state of sensory  deprivation. So nothing you see or hear has any  

connection to the real world anymore, leading  to uncontrolled perceptions that are detached  

from reality or in other words, hallucinations. There are also more direct, less legal ways  

to induce hallucinations. Take magic mushrooms  for example. These mushrooms contain a compound  

called psilocybin, which gets turned into  the active molecule psilocin in the body.  

This molecule then travels through your  blood to the brain, where it passes the  

blood-brain barrier and binds to special serotonin  receptors, which causes the whole balance between  

top-down predictions and bottom-up senses to  go out of balance. This is possible, because  

bottom-up connections are primarily handled  by AMPA receptors, while top-down connections  

are primarily handled by slower NMDA receptors.  The AMPA receptors, or the bottom-up receptors,  

are affected by serotonin. And the serotonin  receptors are affected by magic mushrooms.

So the main thing to take away from  all this, is that magic mushrooms can  

downregulate your bottom-up network, causing  your top-down network to become dominant. 

This can lead your brain to making predictions  way too easily, causing you to see patterns where  

there are none. A good example of this is Google  DeepDream. This was a computer vision program  

published in 2015, that was able to find and  enhance patterns in images using a neural network,  

which is essentially what your top-down network  does to what you see. Because this neural network  

was trained on mostly dog images, it hallucinates  dogs on everything that could even remotely 

resemble a dog, which leads to these really trippy  videos that look a lot like a psychedelic trip. 

However, a real psychedelic trip is far more  complex and immersive. During such an experience,  

your brain inserts all your prior expectations  about the world, not just one specific pattern  

like dogs. This means you might see, hear, feel,  and even smell patterns and connections that you  

couldn't have perceived before. Colors might seem  more vivid, sounds could take on a visual quality,  

and ordinary objects might appear to breathe  or shift. This happens because the usual  

filters and constraints your brain uses to  process sensory information are loosened,  

allowing for a flood of unfiltered perceptions. Of course this theory can’t predict everything;  

it has its limitations. But if predictive  processing is as accurate as we think it is,  

that means that everything you  see, feel, hear, touch and taste,  

all comes from within your pitch black, silent  skull. So everything we perceive is a persistent  

illusion that only represents reality and  that’s only kept in check by our constant  

sensory input. So in a very strange way, reality  as we know it is just a controlled hallucination.