--- > According to that overarching theory (the topic of my talk) phantom vibrations are just one vivid demonstration of the way all human experience is built. According to the new theory (called “predictive processing”), reality as we experience it is built from our own predictions. - [Location 68](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=68) --- > Contrary to the standard belief that our senses are a kind of passive window onto the world, what is emerging is a picture of an ever-active brain that is always striving to predict what the world might currently have to offer. ... Nothing we do or experience—if the theory is on track—is untouched by our own expectations. Instead, there is a constant give-and-take in which what we experience reflects not just what the world is currently telling us, but what we—consciously or nonconsciously—were expecting it to be telling us. - [Location 75](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=75) --- > The illusion occurred because predictive brains are guessing machines, proactively anticipating signals from the body and the surrounding world. That guessing is only as good as the assumptions it makes, and even a well-informed best guess will frequently miss the mark. - [Location 87](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=87) --- > But experience still reflects the brain’s current best guessing. It is just that each new round of guessing is a little bit better informed. - [Location 92](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=92) --- > According to this new picture, experience—of the world, ourselves, and even our own bodies—is never a simple reflection of external or internal facts. Instead, all human experience arises at the meeting point of informed predictions and sensory stimulations. - [Location 96](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=96) --- > Incoming sensory signals help correct errors in prediction, but the predictions are in the driver’s seat now. This means that what we perceive today is deeply rooted in what we experienced yesterday, and all the days before that. Every aspect of our daily experience comes to us filtered by hidden webs of prediction—the brain’s best expectations rooted in our own past histories. - [Location 105](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=105) - _Note_: This is, in a way, similar to how fhe stock market forecasts alter the stock market themselved, as some sort of "self-fulfilling prophecies". --- > Emotion, mood, and even planning are all based in predictions too. Depression, anxiety, and fatigue all reflect alterations to the hidden predictions that shape our experience. Alter those predictions (for example, by “reframing” a situation using different words) and our experience itself alters. - [Location 115](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=115) --- ### UNBOXING THE PREDICTION MACHINE > The idea (the main topic of this book) is that human brains are prediction machines. They are evolved organs that build and rebuild experiences from shifting mixtures of expectation and actual sensory evidence. - [Location 151](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=151) --- > That new information (signifying the lack of birdsong) generated “prediction error signals” and these—on this occasion at least—were all it took to bring my experience back into line with reality. ... in other cases, as we’ll see, mistaken predictions can become entrenched and contact with reality (itself a complex and vexed notion) harder to achieve. - [Location 155](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=155) --- > With the prediction machinery up and running, perception becomes a process structured not simply by incoming sensory information but by difference—the difference between the actual sensory signals and the ones the brain was expecting to encounter. - [Location 168](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=168) --- #### The Smart Camera Model of Seeing > Versions of the smart camera (feedforward) view have been influential in philosophy, neuroscience, and AI. Such a view is intuitive because we typically think of perception as all about the flow of information from the world to the mind. ... As impressions from the outside world (and from within the body) flowed forward into the brain, they were said to be preserved in our minds much the way pushing your fingers into wax preserves information about their shape. ... That idea was pretty much standard in late-twentieth-century cognitive neuroscience. This was probably because it appeared as a governing principle of David Marr’s hugely influential computer model of vision. - [Location 194](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=194) --- > Notably absent from Marr’s model, however, was another direction of influence—one running backward, from deep within the brain down toward the eyes and other sensory organs. The number of neuronal connections carrying signals backward in this way is estimated to exceed the number of connections carrying signals forward by a very substantial margin, in some places by as much as four to one. - [Location 215](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=215) --- > It was puzzling enough to lead the artificial intelligence pioneer Patrick Winston to comment, even as recently as 2012, that with so much information apparently flowing in the other (downward) direction, we confront “a strange architecture about which we are nearly clueless.” - [Location 224](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=224) --- #### Flipping the Flow > A predictive brain is a kind of constantly running simulation of the world around us—or at least, the world as it matters to us. Incoming sensory information is used to keep the model honest—by comparing the prediction to the sensory evidence and generating an error signal when the two don’t match up. - [Location 233](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=233) --- > As a brain encounters new sensory information its job is to determine if there is anything in that incoming signal that looks like important “news”—unpredicted sensory information that matters to whatever it is that we are trying to see or do. - [Location 243](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=243) --- #### Bad Radios and Controlled Hallucinations > The contemporary picture of the predictive brain has historical roots in the nineteenth-century ideas of a German physicist and polymath named Hermann von Helmholtz. Helmholtz was the inventor of the ophthalmoscope used by opticians to examine the eye and formulated the law of conservation of energy. - [Location 257](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=257) --- > we perceive the world only thanks to a kind of unconscious reasoning or inference in which the brain is asking itself, “Given everything I know, how must the world be for me to be receiving the pattern of signals currently present?” - [Location 260](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=260) --- #### The Frugal Brain > Making perception turn on prediction has another important benefit too. It enables the brain to process incoming sensory information in a way that is quite remarkably efficient. - [Location 288](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=288) --- > Linear predictive coding has its roots in a paper published in 1948 by Claude Shannon, a mathematician and cryptographer who was working for Bell Laboratories. ... If a certain letter is almost always followed by another, then an efficient coding scheme can simply assume this to be so unless the case is marked as an exception. Marking only those occasional exceptions is far more efficient, using less bandwidth than would be needed to encode every letter. ... This shows that active but nonconscious predictions bias response and judgment too. - [Location 296](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=296) --- > The trick is trading intelligence and foreknowledge on the part of the receiver against the costs of encoding and transmitting all the information. - [Location 302](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=302) --- > The beauty of this procedure is that by transmitting just a few bits of error, a rich content (such as an image or message) can be reconstructed. The rich content is built mostly out of the predictions but gets anchored to reality by the residual errors. - [Location 307](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=307) --- > Human brains seem to benefit from intelligent prediction strategies of just that kind, and they do so in an especially powerful way, thanks to the use of multiple “levels of processing.” In these multilevel contexts, simple predictions are nested under less simple, more abstract ones—much - [Location 327](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=327) --- > in this kind of multilevel arrangement all that flows forward (from the sensory edges ever deeper into the brain) is news—deviations from what is expected. This is efficient. Valuable bandwidth is not used sending well-predicted stuff upward. - [Location 337](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=337) --- #### Sine-Wave Speech, and the “Green Needle” Effect > Hearing ordinary speech in your native language involves the very same trick—the better your predictions (perhaps because you know the speaker or share the accent) the clearer the sounds. In every case, perception is improved by the presence of a good predictive model. - [Location 395](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=395) --- #### Hallucinating a White Christmas > Back in 2001, researchers at Maastricht University in the Netherlands gave this task to a set of undergraduate students, who were asked to press a button if at any point they believed they were hearing the song. The trick, though, was that nowhere in the tape was there any hint of the supposedly hidden song—the tape was 100 percent white noise, and 0 percent “White Christmas.” About a third of the students in that study pressed the button at least once—a significant result. - [Location 411](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=411) --- #### That Dress, and Other Illusions > This shows—just as Helmholtz thought—that what we see is not simply how things are: rather, we see whatever our brain infers (guesses) as the most likely cause of the evidence coming in from the senses. - [Location 438](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=438) --- > Pascal Wallisch and a team based at New York University conducted an online survey of 13,000 subjects, who were asked not just about how they saw the dress in the photo, but also about how they believed the lighting to be for the photo—did they think it was shot in artificial light, natural daylight, or were they unsure? Sure enough, there was a strong correlation such that those who said that they assumed natural light tended to see the dress as white and gold, while those who thought the lighting was artificial tended to see blue and black, with those who were unsure displaying a more varied mixture of responses. ... The respondents were also asked whether or not they self-identified as “larks” or “owls.” Larks are those who tend to get up early, go to bed early, and feel best in the morning, while owls have the opposite profile, preferring to sleep in, staying up later, and feeling best at night. Remarkably, these self-identified “circadian profiles” correlated strongly with how the dress was perceived. - [Location 449](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=449) --- > This is our first encounter with something that will loom large in our discussions—the impact of our own daily actions upon our brain’s predictive models. - [Location 462](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=462) --- #### Learning to Predict > To some degree, we are obviously not starting from scratch. Millions of years of evolution have determined the bedrock configuration of the machinery we command at birth: the early wiring of the brain, the structure of our sense organs, and the shape of our bodies. Courtesy of all that, we start our journey already armed with plenty of hard-won knowledge. You might even say that (in a slightly strained sense) creatures with lungs are already structurally “expecting to breathe.” - [Location 468](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=468) --- > just by attempting to predict the world we can acquire the knowledge that later enables us to predict the world better. - [Location 474](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=474) --- > In the absence of a good-enough predictive model, we will not succeed at turning the raw evidence into anything approximating a coherent understanding of the world. It will be like viewing those Mooney images, or worse. But even so, the brain can still manage to learn. It does so by looking for better and better ways to predict that unruly sensory barrage. Very young infants seem to spend most of their time doing just this, trying to find useful patterns in the sensory stream. - [Location 478](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=478) --- #### Perceiving as Predicting > It’s important not to think of what I will often be calling the “sensory evidence” or the “raw sensory signal” as itself something that is experienced. Instead, experience is what happens as sensory evidence (for example, patterns of reflected light impacting cells in the retina) gets matched by better and better predictions of that evidence. These predictions are the distilled fruits of previous experience and learning. The initial predictions act like a rough draft of experience. - [Location 525](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=525) --- ### PSYCHIATRY AND NEUROLOGY: CLOSING THE GAP > What we feel is in every case a construct (just as Helmholtz suggested). It is a construct that reflects a process of unconscious inference—informed guessing—about the nature of the events causing our sensory stimulations. - [Location 579](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=579) --- #### Beyond Tissue Damage > Pain has generally been assumed to fall into two different categories—nociceptive and neuropathic. Nociceptive pain is pain that is performing its adaptive function, indicating actual or threatened bodily damage. Neuropathic pain, by contrast, is defined as pain that is caused by damage or disease affecting the sensory systems that deliver the experience of pain, or the processing of pain information. Nociceptive pain is telling us that something is wrong in the body. But neuropathic pain is more like a malfunction in the pain-signaling system itself. - [Location 596](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=596) --- > In 2016 a third category was added, known as “nociplastic pain.” This was defined as pain arising from abnormal processing of pain signals without any clear evidence of either tissue damage or any other recognized systemic pathology. - [Location 604](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=604) --- > In one striking fMRI study, they showed that religious beliefs could regulate the experience of physical suffering, arguing that a kind of high-level reframing of the sensory signals mediated their actual experience and exerted an analgesic effect. When shown religious images, religious subjects rated a sharp pain as less intense than atheists shown the same image. - [Location 610](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=610) --- > We have seen that human experience arises at the meeting point of predictions and sensory evidence. But exactly how those two potent forces meet and balance is flexibly determined by a further factor: the brain’s best estimate of their relative reliability and significance. Predictive processing accounts refer to this as their estimated “precision” and incorporate it as a varying weighting on predictions and on sensory stimulations. - [Location 622](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=622) --- #### Self-Confirming Cycles of Pain > Such studies suggest a complex dynamic in which false expectations, once they get a grip on us, become increasingly resistant to change. This phenomenon of spuriously self-confirming expectations is probably more common than we realize, as when a patient, expecting dentistry to hurt, experiences greater pain than they otherwise would—which then in turn appears to confirm, and thereby cements, their own prior belief. - [Location 698](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=698) --- > Functional Disorders - [Location 701](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=701) --- > In traditional psychiatry, diagnoses are made, and treatments are given, on the basis of loosely correlated sets of symptoms and associated chemical changes in the brain. But an emerging multidisciplinary approach known as “computational psychiatry” approaches mental health and illness in a more fundamental way. Arising at the crossroads of neuroscience and computer models of the mind, computational psychiatry aims to develop a more insightful and systematic alternative to the standard symptom-based approach. It seeks to understand psychiatric conditions (and psychological diversity more generally) as a reflection of differing balances in the ways our brains process information. - [Location 704](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=704) --- > Functional disorders are cases where symptoms such as motor problems, paralysis, or even blindness are present, but no standard physiological cause can be identified. - [Location 714](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=714) --- > What could reasonably lead medical practitioners to diagnose a functional neurological disorder? The most striking form of evidence is that the contours of functional problems often follow our intuitive notions of disease or anatomy rather than medically or physiologically sound ones. - [Location 743](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=743) --- #### Disordered Attention > In brains, precision-weighting involves the action of (among other things) complex neurotransmitter systems centered on dopamine and other chemical messengers. Their coordinated action amplifies some aspects of neuronal activity at the expense of others. - [Location 788](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=788) --- > precision is thought to be estimated at all times and for all neuronal populations. Varying estimates of precision alter patterns of post-synaptic influence and so determine what (right here, right now) to rely on and what to ignore. - [Location 791](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=791) --- > That’s what attention, if these accounts are correct, really is—attention is the brain adjusting its precision-weightings as we go about our daily tasks, using knowledge and sensing to their best effect. By attending correctly, I become better able to spot and respond to whatever matters most for the task I am trying to perform ... Precision estimation is thus the heart and soul of flexible, fluid intelligence. - [Location 799](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=799) --- > unwilled misallocations of precision act as self-fulfilling prophecies. Predictions of pain or impairment become highly overweighted, and those predictions overwhelm the actual sensory evidence, forcing experience to conform to our own hidden but misplaced expectations. - [Location 805](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=805) --- > There is good evidence that misfiring precision assignments (unusual patterns of attention) play a role in many, perhaps all, functional neurological disorders. - [Location 811](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=811) --- > simply distracting the sufferer by making them direct their attention elsewhere often makes functional (but not structural) tremors vanish. - [Location 812](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=812) --- #### Hoover’s Sign > A classic demonstration of the role of aberrant attention in functional disorders involves the phenomenon known as “Hoover’s sign.” Named after the American physician Charles Franklin Hoover, this is present when a patient with unexplained weakness in one leg proves able, when their attention is directed elsewhere, to exert normal amounts of pressure with that leg. - [Location 822](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=822) --- > What Hoover’s sign shows is that the problem is not really with the power of the leg, nor even with the ability of their brain, when distracted, to deploy that power. Rather, it reflects what happens when attention is directed toward using the afflicted leg. - [Location 833](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=833) --- > what Hoover’s sign suggests is that a certain unwilled pattern of expectation and attention—caused, of course, by very real changes somewhere within the brain—may be the hidden cause of the apparent weakness. Another way to think about this is that the absence of disease or gross physical damage does not mean that there is no pathological change at all. - [Location 837](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=837) --- > To make sense of these self-constructed feelings of pain, numbness, weakness, or paralysis, sufferers may start to suspect deep hidden causes—such as persistent hidden illness. These new beliefs then further reinforce the expectations of those symptoms, reinforcing the cycles of aberrant attention. A similar circular pattern of false confirmation may occur in some cases of psychosis, as we’ll later see. - [Location 851](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=851) --- #### Expectancy and Its Role in Chronic Pain > what all this suggests is that there is no such thing as a raw or “correct” experience of a medical symptom anyway. Since all human experience is constructed from mixtures of expectation, attention, and sensory stimulation, it will never be possible to experience either the world or your own body “as it really is.” - [Location 857](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=857) --- > symptoms across a wide range of conditions strongly match their bodily causes only for early, acute, and localized dysfunction or pain—for example, the temporary sharp pains caused by surgery, cuts, and broken bones. Move to chronic conditions and the picture looks very different. - [Location 863](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=863) --- > we need to move away from thinking of pain as a simple sensation, a direct signal of damage or potential damage, to a view of pain as a perception. Like all perceptions, it takes shape only thanks to the precision-weighted interaction of predictions and current bodily signals. - [Location 889](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=889) --- #### Altered Balances in Autism Spectrum Condition > there are two very broad ways for such processing to go wrong. The first is for the brain to underweight predictions and expectations. This will make it hard to detect faint but predictable patterns in a noisy or ambiguous environment. But the second general way to go wrong is for the brain to overweight expectations. In extreme cases, overweighting results in hallucinations. - [Location 896](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=896) --- > Autism spectrum condition was initially thought to reflect a specific imbalance of the first kind—a systematic underweighting of prior expectations. This was the “weakened prior” theory of autism. ... Recent evidence casts subtle doubt, however, on this bald initial hypothesis. Rather than weakened predictions, intriguing evidence is emerging that suggests that the core issue involves (not underweighting knowledge-based predictions but) actively overweighting the incoming sensory evidence. - [Location 902](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=902) --- ### Enhanced Sensory Worlds > George Musser reports a conversation with Satsuki Ayaya, a PhD student in Tokyo with a diagnosis of autism. Ayaya reports that her experience presents her with excessive amounts of detail that often do not serve her daily needs. ... She doesn’t just feel “hunger,” instead the more fine-grained specifics of the bodily signals dominate. You are feeling a whole lot of something—but what is it? - [Location 922](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=922) --- > an individual with autism spectrum condition might start to self-select more predictable environments, becoming increasingly wary of complex social encounters. Repetitive and stereotyped behaviors such as rocking or hand-flapping might also emerge, as these would offer a clever way to ensure (by self-generating) a predictable stream of sensory input. Yet another way to reduce sensory surprises is to develop extreme expertise in a restricted domain. - [Location 938](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=938) --- #### The McGurk Effect > In the McGurk effect, subjects are shown a video clip where the sound “ba-ba” is played, but the person’s lips are actually moving in the ways they would if they were saying “ga-ga.” Faced with this apparent contradiction, neurotypical subjects tend to merge the two sources of information, and clearly hear “da-da.” ... The McGurk effect is diminished—and sometimes entirely absent—in those with autism spectrum condition. - [Location 953](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=953) --- > Altered Balances in Schizophrenia - [Location 964](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=964) --- > The psychologist Peter Chadwick describes his own experience of the onset of schizophrenia as involving what he called a “step-ladder to the impossible.” As he ascended the rungs of the ladder, he ascribed increasing significance to an array of patterns and coincidences, slowly forcing him to shift his fundamental understanding of how the world works. - [Location 966](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=966) --- > Falsely generated, highly weighted prediction error forces the brain to seek a new predictive model. The resulting hypotheses (such as the Organization, telepathy, alien control, and nowadays all manner of strange beliefs involving the internet) may appear bizarre to the external observer. Yet from within they constitute the best—indeed the only—explanation available. - [Location 981](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=981) --- > This also helps explain what may be going on during the distinctive early stages of the onset of psychosis. There, although the person is otherwise functioning much as usual, the world begins to look somehow different or strange. That strangeness, the authors suggested, reflected the presence of persistent, unresolved prediction errors. Those errors then slowly drive the system to form increasingly radical hypotheses in an effort to accommodate them. - [Location 985](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=985) --- > Aberrant prediction errors, even if they play a role in the development of psychosis, are probably seldom (if ever) the whole story. - [Location 992](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=992) --- #### Post-Traumatic Stress Disorder > PTSD severity was extremely well correlated with unusually large increases in precision-weighting on the prediction error signal in response to unexpectedly negative outcomes (unexpected mild shocks). In the most severely affected individuals, the response to failing to predict the shock was to radically overweight the missed cue (the specific face) and thereby become hypersensitive to its occurrence in future. - [Location 1011](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1011) --- > some individuals will be more susceptible to PTSD and other debilitating conditions. If this proves correct, tests like these could one day be used to identify those people who are most at risk, adjusting their roles during military service accordingly. That would be an instance of what has been called “perceptual phenotyping”—the use of psychological tests to help build cognitive profiles that identify individuals at risk. - [Location 1018](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1018) --- #### So Which Balances Are Best? > But what constitutes a good or optimal setting for these various balances—one that will ensure that perceptual experience reveals things “as they really are”? Unfortunately, there is no way even to address that kind of question without making many assumptions about the nature of the actual environment and its relative stability or tendency to change (its volatility). ... Predictive balances that might save your life in one kind of environment may do you all kinds of harm in another. - [Location 1027](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1027) --- > What we see, hear, and feel—even when everything is working exactly as it should—is never a direct reflection of the state of our own body or the wider world. Instead, the world and body we experience is always part construct: a product of our own conscious and nonconscious predictions. - [Location 1051](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1051) --- ### ACTION AS SELF-FULFILLING PREDICTION > prediction is also the engine of action. This is because ordinary daily actions (according to predictive processing) are caused by predictions of bodily sensation. They are caused, more precisely, by predictions of the flow of bodily sensations that would occur if that very action were to be performed. - [Location 1069](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1069) --- #### Ideomotor Theory > The core idea was that actions come about because we mentally represent the completed effects of the action. In other words, the idea of the completed action is what brings the actual action about. ... This became known as the “ideomotor theory of action,” since the idea (or mental image) of the completed motor action is what brings the actual movements about. - [Location 1083](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1083) --- > This is also known as the passive motion paradigm. This states that the task of the brain, when controlling action, is to determine how each bodily joint would have to move if some external force somehow pulled the body toward the goal. It is then the brain’s careful simulation of all the bodily effects expected under that scenario that causes the bodily parts to move in exactly the ways required. - [Location 1096](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1096) --- #### Seeing Seagulls > How did my brain find just the right signals to send to my body to enable all that to unfold? The answer is again by means of a learned predictive model. ... Courtesy of that distilled wisdom, a predicted effect such as “now seeing the gulls,” leads to multiple further predictions of much more specific sensory effects—for example, the ones that would be occurring if my neck muscles were moving in a certain way, the way they would have to move if my head were turning toward the sound. Since my head is not yet turning in those ways, those sensory consequences are not occurring. That delivers a rich stream of prediction errors that are then eliminated by moving my neck muscles so as to make the sensory consequences occur. - [Location 1126](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1126) --- > At the bottom rung of the ladder lie predictions that engage spinal reflexes that move the body. So the whole process is one in which abstract predictions cause ever-more-concrete predictions, enabling my top-level representation of a desired consequence (seeing the gulls) to cause the bodily actions that bring it about. - [Location 1136](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1136) --- > There are two different, but equally effective, ways to minimize prediction errors during our encounters with the world. The first is by using prediction errors to help us discover the best guess about how things are out there in the world. But the second is to act so as to make the world fit some of our predictions. - [Location 1139](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1139) --- #### One Wiring Diagram to Rule Them All > Perception is the inward flow of sensory information, while action unfolds in the opposite direction. But if perception and action were really constructed by the brain in radically different ways, one might expect a corresponding difference in the directionality and flow of information processing in the brain. Yet surprisingly enough, the wiring diagram of “motor cortex,” and the flow of information in that cortex, turns out to be very much like that in sensory (perceptual) regions of the brain. ... Predictive processing resolves this anomaly by showing how action and perception can each involve the same kind of wiring and flow of information. - [Location 1148](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1148) --- > in order to move my body at all, my brain needs to downplay some perfectly accurate information about my own bodily state. ... To move my arm, I must give high weighting to the desired future state (arm moving) rather than the undesired present state (arm not moving). - [Location 1161](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1161) --- #### What Tickling (Really) Teaches > Back in 1950 the great German behavioral physiologist Erich von Holst proposed that every motor command given by the brain is accompanied by a second copy of that command, the “efference copy.” That second copy, Holst believed, was sent internally to a kind of onboard simulator able to predict (in advance) the sensory consequences of the action. - [Location 1175](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1175) --- #### Lessons from the Outfield > “active sensing.” The idea is that sensing itself is an intelligent action, aimed at delivering just the right information, just in time for use. - [Location 1212](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1212) --- > well-timed bodily movements improve our states of information. Perception is no longer a passive phenomenon. Instead, perception and action constantly engage in a kind of coupled unfolding—movements serve up perceptions that enable more motor movements that deliver further perceptions. - [Location 1218](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1218) --- #### Embodied Expertise > the conscious awareness of the skilled driver is freed to be dominated not by thoughts about the unfolding details of their motor action but simply by a kind of “expert seeing”—seeing just where and how the vehicle must move. The prediction-error-correcting brain then does the rest. In this way the car behaves as a kind of extension of the driver’s body—a selected trajectory for the car recruits the cascade of predictions of gear-stick motion and foot-pressure sensations needed to bring that trajectory about. - [Location 1252](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1252) --- > The hidden task of all that training, we can now appreciate, is to enable our brains to predict (via a cascade that often starts with a very high-level goal or aim) the many subtle sensory consequences of an unfolding successful action. ... we learn how things will look and feel if we are getting the action right—we learn its sensory consequences, highlighting those most necessary for success. - [Location 1270](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1270) --- #### The Long Game (and the Role of Optimistic Predictions) > goal-directed behavior involves using predicted outcomes to help structure the actions that will best serve to make those outcomes real. - [Location 1278](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1278) --- > What about even longer-term goals and projects? These are not so very different, except that we must now minimize prediction errors in a yet more abstract and temporally extended space. If I plan to become a better surfer, my brain needs to make the “realistic-yet-optimistic” prediction that I will indeed later be such a surfer. With that goal (long-term prediction) active, I can use what I know about how things work in the world to identify important steps along the way, generating a policy that might—according to my current skill set and personal circumstances—include moving to the coast, taking classes, or vacationing in Tarifa. ... Predictions like that are somewhat belief-like (being about what is predicted to occur) and somewhat desire-like (being nicely poised to bring those very things about). - [Location 1292](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1292) --- ### PREDICTING THE BODY #### Escaping the Dark Room > Homeostasis (deriving from Greek words meaning “same” and “steady”) implies a tendency to return to a state, or to return to a place within a range. It is to maintain a viable set of internal conditions despite external fluctuations. - [Location 1351](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1351) --- > Where homeostasis implies constantly returning to some fixed point, allostasis highlights the need to alter the fixed points themselves so as to adapt to changing needs and environments. This is exactly what we humans do when we respond to worries and threat by increasing cortisol levels. This so-called stress hormone floods the bloodstream with glucose, powering up our cells so we can take effective high-cost actions such as running fast. - [Location 1364](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1364) --- > Rather than waiting for something to go awry, then using negative feedback to bring things back into line again, predictive models allow us to make early, preemptive responses. - [Location 1369](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1369) --- > These states need to be proactively maintained just to keep the living being in existence. This is the “interoceptive” (inward-looking) sensory system and it is distinct from, but interacts with, the “exteroceptive” (outward-looking) sensory system that includes vision, touch, and hearing. - [Location 1375](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1375) --- #### Curiosity and Prediction Error > Positive and negative moods and feelings are thought to be reporting these important error dynamics, bringing them forcefully to our attention by making some events and situations simply strike us as way more pleasant and fulfilling than others—a good day on the tennis court, or in the office, when you are really “in the zone” for whatever you are trying to achieve, versus a day when every minor task seems like an uphill struggle. - [Location 1394](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1394) --- > Being natively attracted to environments in which greater than expected amounts of prediction error are resolved is a neat way of ensuring adaptively beneficial tendencies toward play, learning, and exploration. Such creatures cannot help but seek out and prefer those parts of their world in which useful learning is currently possible. - [Location 1411](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1411) --- #### Predictive Body Budgeting > In her groundbreaking book How Emotions Are Made, the psychologist and neuroscientist Lisa Feldman Barrett captures the bedrock role of predictions in maintaining a viable bodily state using the compelling metaphor of “body budgeting.” Just as a financial budget tracks income and expenditure, a body budget tracks and anticipates the use and replenishment of key resources for maintaining bodily life and functioning. These resources include water, salt, and glucose. To renew them, we engage in familiar activities such as finding and consuming food and sleeping. Allostatic mechanisms are vital to this process. - [Location 1421](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1421) --- #### Embodying Emotion > The central idea is that a single kind of process combines inner and outer sources of information, generating a context-reflecting amalgam that is experienced as emotion. - [Location 1463](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1463) - _Note_: Of the interoceptive predictive processing theory --- > According to interoceptive predictive processing, feelings and emotions are what result when we integrate basic information about bodily states and general arousal with higher-level predictions of their most probable causes—for example, heart attack versus exercise. - [Location 1466](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1466) --- > According to interoceptive predictive processing, emotions and feelings reflect a process that combines interoceptive (inward-looking), proprioceptive (action-guiding), and exteroceptive (outward-looking) information with model-based predictions of all those signals as they are occurring. - [Location 1476](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1476) --- #### Wiring the Mesh > But contrary to that neat, incremental view, the cortex is not a newcomer to human brain evolution. ... There is growing evidence that cortical and subcortical areas evolved in a highly coordinated fashion that produced rich interdependencies. What resulted is a complex looping arrangement linking cortical and subcortical structures in a web of continuous two-way influence. - [Location 1508](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1508) --- #### Seeing from the Heart > consider the bizarre case of Capgras delusion. This is the belief that your loved one has been replaced by an impostor. The delusion seems to be triggered when—for whatever reason—your own body ceases to respond in the usual way to the loved one’s presence. These missing bodily responses, such as a slightly raised heartbeat or increased galvanic skin response when in the loved one’s presence, are not consciously registered. But their sudden absence again acts as evidence that the ever-whirring predictive brain needs to explain. - [Location 1546](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1546) --- #### Depression, Anxiety, and Bodily Prediction > There is, at the very least, a complex two-way street—often mediated by predictions of energetic need—linking the mental and the more standardly physiological. - [Location 1563](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1563) - _Tags_: `c1` --- > This unifying perspective also makes sense of the finding that chronic depression involves abnormalities not simply of “mood” but also of sleeping-waking cycles, and of metabolic and immunological response. Tying all these together, Barrett suggests, may be a “central problem with inefficient energy regulation.” - [Location 1566](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1566) --- > among the most notable and devastating characteristics of chronic depression, anxiety, and many other psychiatric conditions is their surprising resistance to new information. This suggests that where such conditions really take hold, there is also a problem with either generating or learning from the prediction error signal. - [Location 1571](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1571) --- > Overweighted expectations and underweighted new information would result in a kind of permanent or semipermanent lock-in of the existing model, leading us to continue with depressive behaviors that actually serve to reinforce the bad model, and that lend false justification to our prior expectations. - [Location 1576](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1576) --- #### Immunizing Ourselves to Positive Information > One important and consistent finding in this area is that chronic depression involves a resistance to updating our negative expectations when confronted with what ought to be good evidence of positive outcomes. This failure to update in the face of good evidence (Barrett’s “locked-in brain”) most likely involves abnormally high precision on prior negative beliefs, which in turn robs unanticipated positive information of the power to alter the inner model that is delivering negative anticipations. - [Location 1597](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1597) --- > Therapists and clinicians should explicitly target not just negative expectations but also the complex patterns of downgrading or rejecting new positive evidence that result from the abnormally high weighting placed on prior negative beliefs. - [Location 1618](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1618) --- > Also known as psychogenic chills, these are the distinctive shivery feelings (often paired with piloerection or goose bumps) that many people experience at moments of sudden high emotion. Aesthetic chills occur in many contexts, including as a response to art, film, poetry, scientific insight, social ritual, or even when watching a skilled sports performance. ... these chills occur when we encounter something that our brain identifies as critical new information that resolves important uncertainties. - [Location 1638](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1638) --- > This is because they are—to use University of Oslo Professor of Comparative Literature Karin Kukkonen’s memorable phrase—“probability designs”: artifacts engineered to interact in reliable ways with our own predictive brains. Books, novels, plays, and movies are all probability designs. - [Location 1654](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1654) --- > Aesthetic chills are a physiological marker of this sudden increase in estimated importance (precision). - [Location 1657](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1657) --- ### The Hard Problem—Predicting the Predictors? > What, you may ask, does all this tell us about the origins of what philosophers call “qualia”—the distinctive qualitative feel of “seeing red,” “feeling angry,” or “tasting like raspberry Kombucha”? Are all those exhilarating and nauseating qualitative experiences that populate our day-to-day mental life nothing over and above that multifaceted precision-weighted prediction machine in action? ... This, of course, is nothing other than the “hard problem of consciousness”—the ... The hard problem we seem to confront is, in some ways, a trick of the mind. - [Location 1701](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1701) --- > what David Chalmers has dubbed the “meta-problem of consciousness”—the problem of explaining why it is that we are drawn toward mind-body dualisms and tempted to posit an unbridgeable “explanatory gap” between the best science can offer and the facts about our own experience. - [Location 1715](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1715) --- > The various phenomena that underlie the hard problem are widely felt to be among the most important yet scientifically ill-understood features of the universe. Philosophers and scientists disagree, both with each other and among themselves, about what the problem really is, how best to solve it, or even if it is possible to solve it. - [Location 1720](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1720) --- #### Simple Sentience > Philosophers and psychologists talk here of “affordances,” where these are the opportunities for action that arise when a certain type of creature encounters a certain kind of situation—a - [Location 1734](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1734) --- > The term sentience was used by the political and social philosopher Jeremy Bentham as long ago as 1789 (in his Introduction to the Principles of Morals and Legislation) to mark a distinction between the capacity to feel and the capacity to think and reason. ... We can now think of sentient beings as those whose neural model of the world is in constant two-way communication with a model of their own changing physiological state. Basic sentience emerges in creatures whose sensitivities to states of the external world are subtly but pervasively responsive to the likely future states of their own bodies and metabolisms. ... They live, we might say, in a world that is temporally extended and perceptually meaningful. - [Location 1750](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1750) --- #### Expecting Ourselves > we find ourselves able fluently to predict our own and others’ responses by modeling ourselves using some simple shorthands that suffice for our daily purposes. That simple shorthand says we are home to “qualitative experiences,” many of which we either like or dislike. Knowing that you like sweet things I can select a Drambuie Scotch liqueur for you this holiday season, rather than a peaty malt. - [Location 1791](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1791) --- > spotting ordinary things like instances of dogs and cats (which seem to just be “things in the world”) and stranger things, like the “taste of honey,” are all actually on a par—they are all just inferred causes that the brain conjures up to help us predict our own sensory flows. But in the latter case, a major part of what we are predicting is ourselves: our own matrix of hidden tendencies to act and respond—to - [Location 1797](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1797) --- #### Simple Self-Models > The point to note is that every property and feature here reported has—according to predictive processing—been extracted in exactly the same way. Each feature and property has simply been inferred as part of the current best attempt to predict the current waves of sensory stimulation. ... They all serve to organize and predict whole swaths of interacting inward- and outward-looking sensory information. - [Location 1828](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1828) --- > All that nature cared about when building us was that we be able to detect what’s out there (is that a predator or a friend?) and how things are physiologically, whether we need food, water, or rest, or to avoid injury. Adding machinery to enable us to appreciate the full complexities of our own inner processing would have been costly and quite possibly counterproductive, diverting attention from what really matters in body and world. - [Location 1842](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1842) --- #### Questioning the Philosophical Zombie > Back in 1996, Chalmers asked himself how the process of perception would strike an advanced intelligence that has access only to the end products (the best guesses, as we’d say) resulting from their own complex inner processing. His reply was that such an agent, when asked how they know that the honey tastes sweet, may be forced to say they know this directly, in some brute but puzzling manner. They might then start to judge that they are home to mysterious “qualia.” ... Chalmers himself has frequently invoked the possibility of full “philosophical zombies” in arguments concerning consciousness, where a full zombie would be a creature all of whose behaviors (including everything they say and do) perfectly match our own, but one that is lacking any form of inner mental life or subjectivity. - [Location 1854](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1854) --- > a clever algorithm can caption an image, spotting that it contains a dog (say) without experiencing anything at all. ... But the more detail we plug in here, the less plausible the “full zombie” picture really seems. Our imagined being will know a lot more than simply “that there is a dog (say) out there.” They will know the shape, color, texture, and behavioral tendencies of the animal. ... Most importantly perhaps, they may know that they generally “like” dogs—they tend to respond positively to their looks and seek out opportunities to pet and play with them. Could all that happen “in the dark” experientially speaking? The more you examine such a claim, the harder it becomes to really imagine it. The being just described sounds a whole lot like us. Such a being does not know how their knowledge about the world actually comes about. ... They are thus poised to find their own subjective experience strange and puzzling in just the ways Chalmers described. I believe that we are those creatures. Self-opacity and simplified self-models lead us—clever but limited beings—to infer that we are home to a mass of extremely puzzling “qualitative states.” ... But there is an opportunity here too. We can act to remedy this, simply by proactively structuring more of our fictional worlds in ways that are more realistic, or that are helpfully aspirational. - [Location 1867](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1867) --- ### EXPECTING BETTER #### Perceiving What You Feel > A recent paper looking at shooter bias noted that between 2007 and 2014 a full 49 percent of officer-involved shootings of unarmed victims were linked to what have become known as “threat perception failures.” In such cases, officers mistook an innocent object such as a cell phone, or a nonthreatening movement, as an armed threat. Moreover—as has become increasingly appreciated in recent years—these threat perception failures predominantly involved Black victims. - [Location 1918](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1918) --- #### Responding to Predictive Bias > Understanding how misplaced stereotypes and unjust practices interact with our own changing bodily signals should add to the growing realization that what is most urgently needed is deep and abiding change in bedrock societal practices and institutions. These include media depictions, police culture, and press coverage, all of which play a role in encouraging the conscious and unconscious racist beliefs that lead to misguided predictions: predictions that distort perceptions and result in inflammatory and sometimes fatal interactions. - [Location 1944](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1944) --- > In all these interlocking ways, the authors of the study conclude, “feeling significantly distressed or threatened can predictively contribute to perceiving the world as more stressful or threatening in a very literal sense.” - [Location 1958](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1958) --- #### Helpful Fictions > Fictional worlds provide one small, much more easily manipulated, lever for change. We humans do much of our learning in media, advertising, and entertainment. We read books and see movies, we play video games, some of which may involve immersive virtual realities combining passive perception and real-world action. This is both a vulnerability and an opportunity. It is a vulnerability because many of our constructed fictional worlds either fail to reflect the true nature of our societies or reflect aspects that we would not wish to promote. - [Location 1966](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1966) --- > By retraining our unconscious prediction machinery, fictional worlds can act as powerful tools for pushing back here too. They can reduce estimated certainty regarding existing pernicious stereotypes and help install new and better ones - [Location 1976](https://readwise.io/to_kindle?action=open&asin=B0B5CY7Q9K&location=1976) --- #### Improving Interoception