What is the Beast Machine Theory of Consciousness?

In the history of science, breakthroughs often stemmed from addressing the questions that were ignored or even ridiculed. Classic examples include Galileo’s scepticism against the common sense of his time that heavier objects fall faster and Newton’s seemingly naive question, "Why does an apple fall?"

The same happens to the study of consciousness. Although subjective experience has been explored by philosophers for thousands of years, it was regarded as not fitting for science until two scientists questioned the assumption. They were the two Nobel laureates, Francis Crick and Gerald Edelman. Ironically, neither was studying the brain — Crick was a molecular biologist and Edelman an immunologist. In the 1970s, they switched their careers to become neuroscientists, firmly believing consciousness can be studied scientifically.

Their bravery opened up a serious field for scientific research. More importantly, it has enabled humans to be considerably prepared and mindful to face today's rapid advances in AI.

One young postdoc working in Edelman's lab was Anil Seth. He is now one of the prominent neuroscientists researching consciousness. This article is about his book, Being You, published in 2021.

The book subtitled A New Science of Consciousness, is about Seth’s so-called Beast Machine Theory, which states that consciousness is the brain’s predictive construction of both the external world (perceptions) and the human’s inner body (emotions and moods).

The theory's components are, in fact, not new. What makes the theory outstanding is how it unifies them into a coherent framework.

Numerous psychological and neuroscientific studies have proven that sensory perception (such as visual and auditory) is not a straightforward capture of sensory information from the external world. Instead, it is constructed by the brain based on the interplay between previously held beliefs and current sensory information. Interestingly, the Bayesian theorem can explain the phenomenon perfectly, with a mathematical formula that updates the probability of a hypothesis (perception) based on new evidence (sensory information; please refer to one of my previous articles for more details).

In the 1990s, neuroscientist Antony Damasio published his phenomenal book Descartes' Error: Emotion, Reason, and the Human Brain to illustrate the essential role of emotions and feelings in human consciousness. He used clinical neurological cases to argue that consciousness arises from the integration of body states (somatic markers) and cognitive processes. Specifically, feelings are the conscious experience of emotional states, and they inform the brain about the state of the body — this internal awareness is a key component of being conscious.

In the book Being You, Seth extends the above views and unifies them to underline his Beast Machine Theory, providing a fresh perspective on exploring, understanding, and investigating consciousness.

As reviewed in the previous article, there are currently over 40 theories on consciousness. Here are the main points that distinguish Seth's theory from the rest.

1. Address consciousness holistically as a real problem, instead of a hard problem.

Seth did not spend much time on the definition of consciousness in the book. He simplifies it to "any kind of subjective experience whatsoever," with its content including "the sights, sounds, smells, emotions, moods, thoughts, and beliefs that make up our inner universe."

Based on the one-stop, easy-to-grasp definition, Seth treats consciousness as a "real problem" of physical and material nature, as opposed to the division of easy vs hard problems defined by the philosopher David Chalmers. While Chalmers uses the "hard problem" to address an explanatory gap that modern cognitive science and neuroscience have yet to fill, Seth thinks he accidentally puts on a mystical, nonmaterial veil that "might elude scientific explanation altogether."

In the same spirit, Seth avoids getting bogged down in functionalism or information-processing theories of consciousness, preferring a more balanced and mid-way approach. In his words:

The real problem is distinct from the hard problem because it is not — at least not in the first instance — about explaining why and how consciousness is part of the universe in the first place.
…
The real problem aligns naturally with a physicalist worldview on the relationship between matter and mind.

2. Conscious levels can be measured clinically.

In 2013, Italian neuroscientist Marcello Massimini and his team published a method of measuring conscious states from the human brain. As Seth mentioned in his book, the results correlated "extremely well" with the levels of consciousness disorders caused by various brain injuries in the large number of patients participating in the experiment.

The measure is called the Perturbational Complexity Index (PCI). It reflects the complexity of the brain's electrical responses(EEG) after patients receive noninvasive transcranial magnetic stimulation (TMS). This complexity arises from the way different regions in the corticothalamic system interact, coordinate, integrate, and differentiate their activities.

In a healthy human brain, the cortex and thalamus communicate with each other in much more complex ways during conscious wakefulness than in deep, non-dreaming, slow-wave sleep. The latter still has relatively higher PCI than the state of general anesthesia, while the most impaired vegetative state shows the lowest PCI magnitude.

This method has provided a clinical method to accurately identify a patient's consciousness level, such as differentiating the vegetative from minimally conscious states. It is particularly useful to prevent misdiagnosis when the patient can't speak or move. On the other hand, PCI can't predict subtle differences in normal brains, for example, between REM sleep and full wakefulness.

Anil argues that the success of Massimini's study is an excellent example of addressing the real problem instead of the hard problem. The high complexity value indicates our consciousness is endowed with considerably rich information from our senses, memories, thoughts, emotions, and moods. An instance of our subjective experience is one out of many possibilities depending on the ever-changing external and internal context, and, therefore, unique at a given moment.

3. The beast machine theory of consciousness unifies perceptions, actions, and the Self.

Whenever discussing perception, we can't ignore Bayesian rules. Seth devotes one chapter to explaining how perception is all about Bayesian inference, therefore, a type of "controlled hallucination". He writes:

Abductive reasoning — the sort formalized by Bayesian inference — is all about finding the best explanation for a set of observations, when these observations are incomplete, uncertain, or otherwise ambiguous. … It turns out that the brain is actually implementing Bayes’ rule. More precisely, it is approximating Bayes’ rules. It is this connection that licenses the idea that perceptual content is a top-down controlled hallucination, rather than a bottom-up “readout” of sensory data.

A step further, Seth believes, "Conscious contents are not merely shaped by perceptual predictions—they are these predictions." The purpose of perception (i.e., the predictions) is not to know the world but to control and regulate the body "through a continual process of prediction error minimization."

A compilation of those perceptions, of both external environment and internal body states, comprises the conscious Self:

The experience of being me, or of being you, is a perception itself — or, better, a collection of perceptions — a tightly woven bundle of neurally encoded predictions geared toward keeping your body alive…We do not perceive ourselves in order to know ourselves, we perceive ourselves in order to control ourselves…It’s about physiological control and regulation — it’s about staying alive.

In other words, the perception of Self signifies the embodiment of consciousness, based on the biological material and condition:

The experience of being you, or of being me, emerge from the way the brain predicts and controls the internal state of the body. The essence of selfhood is neither a rational mind nor an immaterial soul. It is a deeply embodied biological process, a process that underpins the simple feeling of being alive that is the basis for all our experiences of self, indeed for any conscious experience at all. Being you is literally about your body.

Furthermore, Seth argues that action is inseparable from perception — they are "the two sides of the same coin."

The most counterintuitive aspect of active inference is that action itself can be thought of as a form of self-fulfilling perceptual prediction. In this view, actions do not merely participate in perception — actions are perception.

Perceptions make active inferences through actions. Both are driven by the mechanisms to minimize sensory prediction errors.

Every kind of action has the potential to suppress sensory errors through active inference, and so every kind of action directly participates in perception.

Now we arrive at the core of Seth's beast machine theory of consciousness:

Here, experiences of selfhood emerge in the unstructured feeling of just ‘being.’ This is where we reach the core of the beast machine theory: the proposal that conscious experiences of the world around us, and of ourselves within it, happen with, through, and because of our living bodies.

4. Consciousness is not intelligence but is embodied in a biological organism.

In Seth's view, "consciousness has more to do with being alive than with being intelligent." He argues that from an evolutionary perspective, the goal of the brain is for any organism to continue staying alive. The significance of consciousness is in precisely achieving this mission.

Specifically, the goal is to maintain allostasis, a process by which the body achieves a dynamic equilibrium through active changes in response to internal or external stimuli (e.g., a danger, stress from work, or physical exercise). It is much more complex than maintaining a static set point, as in the case of "homeostasis" (e.g., a thermostat).

The purpose of consciousness, including both external (exteroceptive) and internal (interoceptive) perceptions, is to regulate the body's physiological condition so that it stays close to the allostatic equilibrium to meet various existential challenges. Emotions and moods, therefore, reflect the brain's assessment of regulatory success or failure, leading to adaptive responses and actions.

This process of regulating and maintaining allostatic equilibrium gives rise to "the experience of the self-as-really-existing." Feelings and moods, therefore, are also perceptions that are indispensable parts of the conscious Self.

That said, Anil believes consciousness exists in at least some animals. For example, he is confident that mammals like rats, bats, monkeys, dogs, and cats have consciousness. The experience would be unique to each species, given that they all have a similar brain structure (the corticothalamic system) as in humans.

How about machines, particularly AI?

Anil argues that consciousness is not equivalent to intelligence. We often confuse the two, which is why we are afraid that AI may soon have consciousness, since it is becoming super intelligent. Anil writes:

Here, consciousness is not determined by intelligence, and intelligence can exist without consciousness. Both come in many forms and both are expressed along many different dimensions — meaning that there is not one single scale for either consciousness or intelligence.

Between life and intelligence, Anil believes consciousness is closer to the former.

What does the Beast Machine Theory feel short?

Seth's theory unifies perception, emotions, and actions into one coherent framework. It anchors consciousness in the evolutionary reality of survival existence and emphasizes its embodied nature, differentiating it from pure information processing and intelligence.

However, there seems to be a gap when accounting for consciousness with perceptions. How does predictive perception give rise to consciousness? If they are the same thing, can we predict that those species without perception don't have consciousness?

Massinini's PCI measurement of a conscious state's brain signal complexity aligns with Shannon's information theory. It implies that consciousness directly relates to how information is processed and integrated. Since information processing is the cornerstone of intelligence, the statement that consciousness is closer to life than intelligence seems speculative and somewhat self-contradictory.

If consciousness is simply the Bayesian predictions based on many inputs from the external environment and internal body, it is a computation process that can be potentially simulated in computers. What's the uniqueness of the body input that consciousness depends on? If body information is simply a source of the content, can consciousness exist without that content?

In Seth's theory, the goal of consciousness is to regulate and control the body's allosteric equilibrium. The body provides something vital for consciousness through emotions and moods. A key question is, what underlying mechanism makes it necessary for consciousness?

In the end, when discussing AI, the author's opinion on machine consciousness is uncertain — "The unsatisfying but honest answer is that I don't know for sure, but probably not." But the beast machine theory logically predicts that machines can't have consciousness because they are not living organisms. What makes the author hesitate to firm this prediction?

The vagueness may come from the gap between the proposed mechanism and the theory itself. The proposed mechanism for perception is a computation that prescribes mathematical operations (Bayesian theorem and prediction error minimizing), while the core of the theory emphasizes the necessity of the biological organism. However, evidence supporting the latter is still lacking, and the ultimate mystery of consciousness is yet to be discovered.

Conclusion

Seth’s book Being You synthesizes the perception, the Self, and embodiment of consciousness from an evolutionary perspective, weaving a coherent picture throughout the animal kingdom, including humans. His beast machine theory is one of the few that offers deeper insights into the perception of Self and the unison of body, emotion, and mind.

Although evidence supporting the embodiment is still lacking, Seth's theory provides a framework for scientists to investigate and research in a direction different from conventional cognitive and computational approaches.

On April 19th, 2025, 21 human-like robots joined the half-marathon race in China. AI is catching up with not only the human mind but also the human body. If robots can run beside us, will they one day also be us, or something entirely different?