The first time I heard about the predictive brain was during an online training course in 2020, where Lisa Feldman Barrett explained how emotions are generated. This theory captivated my interest, and three years later, I discovered Peter Vermeulen's book “Autism and the Predictive Brain” which explores how this theory can be applied to the context of autism. During a conference organized by Autisme Genève in February 2024, Marie Schaer and Michel Godel presented the link between the predictive brain and autism.

This inspired me to explore the theory of the predictive brain further and write this article. In a future article, I will examine how this theory can shed light on certain characteristics of autism, offering new perspectives on this complex condition.

The theory of the reactive brain

For decades, scientists believed that neurons spent most of their time dormant and only woke up when stimulated by a sight or sound from the outside world. Some people have suggested that we only use 10% of our brain. Today, we know that such claims are completely false. In fact, you are constantly using all of your brain's capacity, although its different parts do not always show the same level of activity.

This theory stipulated the idea that the brain is waiting for stimuli. Once a stimulus has been detected, it processes this information and then triggers a response. It was based on the principle that your brain would proceed in three stages:

• Receiving a stimulus from the environment
• Processing this stimulus in the higher areas of the brain
• Acting in response to this stimulus

This is the theory of the “reactive” brain. Incidentally, this is how computers process information. However, we now realize that the human brain works differently.

A reactive brain would be too slow

We can measure the speed at which signals are transmitted within the nervous system and in the brain. In doing so, we find that the “stimulus-processing-response” process should take longer than what we observe in reality. Here are a few examples:

When your cortex perceives that it is your turn to speak, it takes between 500 and 700 milliseconds for your mouth to start producing the necessary words. However, in a natural conversation, the interval between the moment one person stops speaking and the moment the other person starts speaking is around 0 to 200 milliseconds.

Lisa Feldman Barrett, professor of psychology and neuroscience at Northeastern University (United States), uses the example of thirst. Think about the last time you were thirsty and drank a glass of water. A few seconds after drinking the last few drops, you probably felt less thirsty. This is curious, because it takes about 20 minutes for water to reach your bloodstream. If the brain had waited for this moment, the water you drank would not have quenched your thirst so quickly.

These examples show you that you react faster than would be technically feasible based on the theory of the reactive brain. In my article on the two systems of the brain, I have already demonstrated that your prefrontal cortex, your rational mind, has a relatively long reaction time. Nobel laureate Daniel Kahneman calls it the slow system in his book “Thinking, Fast and Slow.” To explain these phenomena, a new concept of the brain has been developed.

The predictive brain

Paying attention to every single signal coming from your body and your senses would be a huge task that would far exceed the brain's resources, which already consume 20% of the body's energy budget. How does your brain manage to be as energy efficient as possible?

Since the 2000s, neuroscientists (led by Karl Friston) have proposed a new concept of the brain, that of the “predictive brain.” We now know that all our neurons are constantly active, stimulating each other at different rates. This activity within the brain is one of the major recent discoveries in neuroscience. The brain is therefore constantly creating a representation of the world.

The Bayesian approach

Keep in mind that your brain, enclosed within your skull, is bombarded with countless signals from your body and senses via your nervous system. The model of the world it has constructed based on past experiences serves as a reference point, since it has no other.

Your brain does not simply receive sensory signals and then process them. Based on its own model of the world, it constantly tries to predict what will happen next.

If, in certain circumstances, you have often experienced something specific, your brain assumes that it will happen again in the same way in the future. Each of your experiences is thus integrated into its model of the world, which allows it to predict what will happen next. This is known as the Bayesian approach, named after the English mathematician Thomas Bayes (1701–1761). His approach is based on the plausibility of a hypothesis to predict – with a certain probability – what will happen next in a given situation. Some researchers therefore use the term “Bayesian brain” instead of “predictive brain.”

The predictions of the brain

Based on data from the past, your brain creates its own internal model of the world and uses it to anticipate sensory input. At every moment, your brain asks itself: When I was previously confronted with a similar situation, when I received the same nerve signals, what did I do, see, hear, or feel in the next moment? This activity represents millions of predictions of what you are likely to encounter next, based on your previous experiences.

Many predictions occur at the micro level, predicting the meaning of fragments of light, sounds, and other information coming from your senses. Every time you hear speech, your brain breaks down the continuous stream of sounds into phonemes, syllables, words, and ideas through prediction. Other predictions occur at the macro level.

Picture taken by Jopwell (pexel.com)

You interact with a friend and, given the context, your brain predicts that she will smile. This prediction prompts your motor neurons to move your mouth to smile back, and your movement causes your friend's brain to make new predictions and actions, in a back-and-forth movement, in a dance of predictions and actions. If the predictions are wrong, your brain has mechanisms to correct them and produce new ones.

Your brain is constantly generating such anticipations and activating the corresponding neural circuits in advance. In this sense, you are watching the movie of your immediate future experience – with all the associated sensations – before things actually happen. Your brain is so effective at hallucinating that you believe you see the world objectively, and so quick to predict that you confuse your movements with reactions. Isn't that a strange idea?

However, brain research confirm this concept. This also means that sensory input and sensations felt in your body are entirely constructed in your mind, even before the sensory input can be perceived in reality.

Prediction errors

Of course, reality does not always match predictions, and your brain takes that into account. It ignores all the data it already thinks it knows and pays attention only to signals that contradict its established model at that precise moment. In other words, it looks for prediction errors. This method is more energy-efficient than having to consider the immense amount of input available at any given moment.

A similar method is used in IT to compress multimedia data. In a movie file, images are not stored pixel by pixel, but only the differences between one image and the next. This results in much smaller files and saves a significant amount of storage space.

It is only when sensory signals do not match predictions that a surprise or error signal is generated in your brain. This ensures that it does not miss unexpected information and can react immediately in case of danger. It is an efficient and energy-saving method, which was a significant evolutionary advantage, enabling the survival of our human species.

Let's look at an example. If a coworker has changed her hair color, you are surprised when you see her. However, if this coworker often changes her hair color, the fact that her hair has changed color no longer surprises you at all. The same is true when a prediction error occurs several times: the brain adapts its model of the world and no longer takes certain variations in a given situation into account. For example, you ignore usual background noise, such as music in a store, and as a result, when you leave the store, you can no longer tell which song was playing.

How to deal with unclear input?

Predicting is easy when you live in a world that is clearly defined and stable. But the information coming from your world is often ambiguous and can be interpreted in different ways, like faces for example.

The theory established in the 1960s, according to which there are universal facial expressions that are interpreted in the same way all over the world, is no longer valid. Hundreds of studies have proven that exactly the same emotion can be expressed by different facial movements. The reverse is also true. A facial expression can be interpreted in many different ways. Take a look at the photo below.

Picture taken by Alena Darmel (pexels.com), modified by the author

Is this man happy, angry, worried, sad, or surprised? Or is he perhaps afraid? His emotions are indeed difficult to read.

According to Lisa Feldman-Barrett, facial expressions alone are not enough to correctly interpret a person's emotions. But how do you guess what someone is feeling? The context provides you with additional information to interpret the other person's emotions. If you knew the context in which this man finds himself, you would be able to better interpret his facial expressions.

These conclusions rely on the results of scientific experiments. But in reality, at least in a given social and cultural context, a specific facial expression will appear regularly in certain situations, so that you will end up interpreting it even without context. In this case too, the brain relies on a predictive process.

Discarding irrelevant information

Thanks to its predictions, your brain omits what seems irrelevant when it feels that it was not important in the past. This allows you to listen to what a person is saying despite background noise and conversations around you, even though your ears are picking up all the sounds in your environment. This is a top-down process, as the brain's model of the world defines which signals should be taken into consideration.

This reminds me of an incident that happened two or three times in my life. The person I was talking to started speaking in my native language, but I couldn't understand them right away because I thought they would be speaking English. I found it very strange. But today, thanks to my knowledge of the predictive brain, I can explain it. In these situations, my brain was expecting sentences in English and was not at all prepared to hear words in my own language.

Without the brain's ability to filter sensory input in advance based on predictions, you would be completely overwhelmed by the enormous amount of information coming in from your senses. You may not be aware of it, but this filtering is a fundamental ability that allows you to manage stimuli from the outside world. Not only do you need to discern what is important in a given situation, but you also need to be able to ignore what is superfluous. Fortunately, your brain does this reliably and efficiently.

Your brain determines in advance what you will perceive

The theory of the predictive brain is counterintuitive and may have surprised you. However, it explains many phenomena. Your brain's role is to ensure your survival in the most efficient way possible. To do this, your brain creates its own model of the world based on your past experiences. Using this model, it tries to predict what is likely to happen next – with a certain degree of probability of course – and triggers the relevant neural circuits. It's a top-down cascade in your brain. Sensory signals from the outside world, which flow into your brain from the bottom up, are only considered if they contradict preconceived sensations. In these cases, they are prediction errors, which lead to an adjustment of the predictions.

The subjective experience – for example “seeing a cup of coffee” – is therefore determined by the content of predictions (descending in the brain) and not by sensory signals (ascending). You never directly experience the sensory signals themselves, but only perceive their interpretations. Neuroscientist Chris Frith uses the term “controlled hallucination” for this phenomenon. In fact, the predictive brain theory provides a better understanding of the unusual hallucinations experienced by psychotic individuals. Furthermore, artificial intelligence is based on the concept of prediction.

Some would argue that every action a person takes is predetermined by the brain and that there is no such thing as free will. Even if one can argue this way, we must not forget that it is how you live today that influences your brain's predictions in the future, and therefore your behavior in the future. In this sense, you are still responsible for your actions, because you can decide at any time what experiences you want to “feed” your brain.

We know that autistic brains process sensory signals differently than non-autistic brains. In a future article, I will discuss how the concept of the predictive brain could help us better understand certain aspects of autism.

• Psychomédia : Le cerveau prédictif : avancée récente la plus intéressante en neurosciences. Interview avec Lisa Feldman Barrett (article in French published on 20 juin 2016) (www.psychomedia.qc.ca/psychologie/2016-06-20/avancee-neurosciences-cerveau-predictif)
  
• Lisa Feldman Barrett : The Predictive Brain. (www.edge.org/response-detail/26707, 29.02.2024)
  
• Lisa Feldman Barrett : Your Brain Predicts (Almost) Everything You Do. (www.mindful.org/your-brain-predicts-almost-everything-you-do, 29.02.2024)
  
• Feldman Barrett, Lisa, 2021. Seven and a Half Lessons About the Brain. Pan Macmillan. ISBN 978-1-5290-1864-6.
  
  

This article was first published in French in March 2024 and has now been translated into English by the author himself.

Author of the article: Beát Edelmann, expert in neurodiversity (autism, ADHD and high IQ) and personality exploration. He is the founder of the Abundana Institute for Self-Management in Geneva, which offers coaching, training, and consulting services in French, English, and German.

Articles about similar topic: Brain