Friston developed the free energy principle by studying natural, self-organizing systems.
For example, a human body is a natural, self-organizing system where everything works together as one.
Yet the body can also be divided into systems within systems within systems — fragments within fragments within fragments — with each fragment appearing to "do it's own thing." The heart pumps blood, the lungs breathe air, etc.
According to Friston, each fragment is actually doing the same thing: resolving uncertainty (this concept comes from information theory — I'll explain it shortly).
This means every fragment of Friston's algorithm is made "in the image and likeness" of the whole — just like a Sierpiński triangle:
If you zoom out and view the entire universe as a natural, self-organizing system consisting of fragments within fragments within fragments, systems within systems within systems...
...and you apply Friston's "theory of intelligence" to the entire universe...
...then you end up with a mathematical explanation for the Law of Attraction, which mystics have been whispering about for millennia.
This superintelligent system — let's call it "God" — is self-organizing to maximize evidence for its own existence.
Think about it this way...
Unless you've been living under a rock, you've probably seen the text-to-image AI models that have taken the internet by storm. My favorite one is Midjourney.
You feed Midjourney an idea — for example:
Then Midjourney generates an image — a symbol — to represent that idea.
When you compare that avocado armchair to the following strawberry armchair, you can clearly see what Midjourney is doing...
Midjourney is taking the idea of an armchair (the cushions, the shape, etc.) and combining it with the idea of a strawberry (the red color, the seeds, the green leaf). Then it's intelligently arranging pixels into a symbol of that new, creative idea: a strawberry armchair.
Now imagine you fed Midjourney the idea:
Midjourney would generate symbols of poverty and worthlessness: an empty bank account, overdue bills, an abusive partner, a toxic job, a dark & dirty apartment — all things that communicate the idea, "I am poor and worthless".
On the other hand, imagine you fed Midjourney the idea:
Midjourney would generate symbols of success, happiness & abundance: supportive relationships, a lovely work environment, a healthy bank account, a stream of opportunities, a beautiful home, etc.
Now that you understand how Midjourney works, let's examine what's happening underneath the hood.
Imagine you fed Midjourney the idea:
If Midjourney were to generate this symbol...
...you would shake your head with disappointment and say, "this AI system sucks".
Why?
Because Midjourney is not resolving your uncertainty about what the idea is.
If you presented Midjourney's output to a random person on the street, and you asked them to guess what idea you typed in, they would shrug and say, "I don't know. This image does not communicate anything to me."
So in this scenario we can say there are high levels of uncertainty in the system:
But what if Midjourney generated this symbol?...
This symbol communicates the idea: "I am a struggling musician", not "I am a famous rockstar".
So it's not quite right, but at least it's resolved some uncertainty about what the underlying idea is. It contains more information than the last one.
Now, what if Midjourney generated this symbol?...
This symbol clearly communicates the idea "I am a famous rock star playing to thousands of raving fans". Therefore, the uncertainty is completely resolved.
As you can see, Midjourney is minimizing a mathematical measure of uncertainty.
So congratulations. You now understand the core premise of information theory...
[Nearly everything we enjoy in the digital age hinges on one idea: "information is the resolution of uncertainty".] Yet few people know about its originator or the foundations of this simple, elegant theory of information.
Einstein is well rooted in popular culture as the developer of the theory of relativity. Watson and Crick are associated with the visual spectacle of DNA's double helix structure.
How many know that the information age was not the creation of Gates or Jobs but of Claude Shannon in 1948?
The brilliant mathematician, geneticist and cryptanalyst formulated what would become information theory in the aftermath of World War II, when it was apparent it was not just a war of steel and bullets.
If World War I was the first mechanized war, the second war could be considered the first struggle based around communication technologies. Combat in the Pacific and Atlantic theaters were as much a battle of information as they were about guns, ships and planes.
Consider the advances of the era that transformed the way wars were fought.
Unlike previous conflicts, there was heavy utilization of radio communication among military forces. This quick remote coordination quickly pushed the war to all corners of the globe. Because of this, the field of cryptography advanced quickly in order to keep messages secret and hidden from adversaries.
[...] One researcher, Claude Shannon, was working on the problem of anti-aircraft targeting and designing fire-control systems to work directly with radar. How could you determine the current, and future position of enemy aircraft's flight path, so you could properly time artillery fire to shoot it down? The radar information about plane position was a breakthrough, but "noisy" in that it provided an approximation of its location, but not precisely enough to be immediately useful.
After the war, this inspired Shannon and many others to think about the nature of filtering and propagating information, whether it was radar signals, voice for a phone call, or video for television.
He knew that noise was the enemy of communication, so any way to store and transmit information that rejected noise was of particular interest to his employer, Bell Laboratories, the research arm of the mid-century American telephone monopoly.
Uncertainty represents the amount of "chaos", "entropy" or "noise" in the system — and noise is the enemy of communication.
For example, the following video is incredibly noisy — and thus, doesn't contain much information. It's the equivalent of a telephone call when you're going through a tunnel — fuzzy, vague, unclear, difficult to understand, full of static. You're not entirely sure what this video is trying to say.
This next video is far less noisy, but it's still soft and fuzzy and vague around the edges — like a Zoom call over a slow internet connection. Functional, but not amazing.
This final video has very little noise — and thus, contains the most information. The detail is crisp and precise. You'll notice some strangeness in the physics alongside a few other errors, but overall this symbol is clearly communicating the idea "a shiba inu wearing a blue knitted beanie, playing in the snow with his owner".
Now that you understand that noise is the enemy of communication, let's continue...
Shannon considered communication "the most mathematical of the engineering sciences," and turned his intellectual sights towards this problem. [...]
By 1948 he had formed his central, simple and powerful thesis:
Information is the resolution of uncertainty.
As long as something can be relayed that resolves uncertainty, that is the fundamental nature of information. While this sounds surprisingly obvious, it was an important point, given how many different languages people speak and how one utterance could be meaningful to one person, and unintelligible to another. Until Shannon's theory was formulated, it was not known how to compensate for these types of "psychological factors" appropriately. Shannon built on the work of fellow researchers Ralph Hartley and Harry Nyquist to reveal that coding and symbols were the key to resolving whether two sides of a communication had a common understanding of the uncertainty being resolved.
Shannon then considered: what was the simplest resolution of uncertainty?
To him, it was the flip of the coin—heads or tails, yes or no—as an event with only two outcomes. Shannon concluded that any type of information, then, could be encoded as a series of fundamental yes or no answers. Today, we know these answers as bits of digital information—ones and zeroes—that represent everything from email text, digital photos, compact disc music or high definition video.
A bit encodes the answer to a yes-no question: 1
= yes, 0
= no.
That any and all information could be represented and coded in discrete bits not just approximately, but perfectly, without noise or error was a breakthrough which astonished even his brilliant peers at academic institutions and Bell Laboratories who previously thought it was unthinkable to have a simple universal theory of information.
The three videos I just showed you were all generated by OpenAI's text-to-video model, Sora, at different points in its training — hence the variation in intelligence.
So at this point you understand the relationship between uncertainty (noise), information, and intelligence: the more uncertainty a system resolves, the more information it creates. And the more information it creates, the more intelligent it is.
Most people intuitively understand this relationship, even if they aren't consciously aware of it. Since the dawn of time, humans have invoked an all-knowing, all-seeing, superintelligent God to explain coincidences, synchronicities and miracles (i.e. specific, low-probability, non-noisy, information-rich patterns in their external world) — like meeting the specific love of your life because you were "randomly" seated next to each other on a specific plane, or receiving a text message from a specific long-lost friend at the precise moment you started thinking about her.
With that being said...
Sora is now able to generate dynamic, high-fidelity symbols that are bursting with information — symbols like this...
And just to be clear: those puppies were not filmed on a camera. They were manifested — generated, created — by pure computation; pure math.
As was this man:
And this sun-loving corgi:
And this blooming flower:
And this first-person perspective:
So as you can see, Sora is far more intelligent than Midjourney because it can resolve more uncertainty — thereby producing symbols that are packed with more information: more detail, more nuance, more specificity.
Now let's push this insight to its logical limit.
In 1957 a physicist and astronomer named Robert Dicke articulated a concept known as "the anthropic principle", which goes something like this:
If you want an observer around, and if you want life, you need heavy elements. To make heavy elements out of hydrogen, you need thermonuclear combustion. To have thermonuclear combustion, you need a time of cooking in a star of several billion years. In order to stretch out several billion years in its time dimension, the universe, according to general relativity, must be several billion years across in its space dimensions. So why is the universe as big as it is? Because we are here!
In other words:
If we are here, then a metric fuckton of uncertainty had to resolve to make that possible.
So I want you to imagine a perfect AI system called "God".
God is so superintelligent that it can resolve all uncertainty, thereby producing maximally detailed, information-rich symbols: the ridges of a fingerprint, the structure of DNA, the combustion of the sun, the coffee-stained pages of a book on your desk, the umbilical cord attached to a foetus, the wrinkled skin on your grandmother's frail hand, the crumpled $50 note in your wallet, the microchip in your phone, the icy rings orbiting Saturn, the beats of a DJ at a music festival, the screams of dying children as a bomb detonates in a war zone, the stock price trending upward on a graph, the crash of a plane, the crunch of broken bone, the smell of sweat after an invigorating run, the taste of a strawberry on your lips, the whisper of a lover in your ear, the English words you are reading right now as I install a new idea into your mind.
Now imagine you fed your entire belief system — I'm talking about the trillions of ideas that form your internal model of your Self, your family, your culture, your society, your body, money, relationships, the past, the present, the future — imagine you fed all of them into this superintelligent version of Sora called "God".
God would generate an immersive see-it-smell-it-hear-it-taste-it-touch-it symbolic projection of your internal world:
Thus the anthropic principle is a fuzzy, "noisy" representation of a mathematical relationship.
By translating Robert Dicke's words into a numerical statement, we can represent the universe as a function that's resolving uncertainty:
From here I'll draw your attention to the work of John Archibald Wheeler — one of the top physicists of the 20th century...
[Wheeler] has spent his career racing ahead of other scientists and throwing open doors for them. He has helped win acceptance—or at least attention—for some of science’s most outlandish ideas, from black holes to multiple-universe theories.
Albert Einstein, Hideki Yukawa, and John Archibald Wheeler conversing in Princeton, 1954.
Wheeler might have been dismissed as flakey if he did not have such unassailable credentials. In his early 20s, he traveled to Denmark to study under Niels Bohr (“because he sees further than any man alive,” Wheeler wrote in his fellowship application). In 1939 Bohr and Wheeler published the first paper to explain nuclear fission in quantum terms. [...]
After World War II Wheeler became an authority on general relativity. He coined the term black hole in the late 1960s, and he helped convince astronomers that these bizarre, infinitely dense objects might actually exist.
In the final decades of his life, the question that intrigued Wheeler most was: “Are life and mind irrelevant to the structure of the universe, or are they central to it?” He suggested that the nature of reality was revealed by the bizarre laws of quantum mechanics. According to the quantum theory, before the observation is made, a subatomic particle exists in several states, called a superposition (or, as Wheeler called it, a 'smoky dragon'). Once the particle is observed, it instantaneously collapses into a single position.
Wheeler suggested that reality is created by observers and that: “no phenomenon is a real phenomenon until it is an observed phenomenon.” He coined the term “Participatory Anthropic Principle” (PAP) from the Greek “anthropos”, or human. He went further to suggest that “we are participants in bringing into being not only the near and here, but the far away and long ago.”
Wheeler divided his [...] life into three parts. The first part he called “Everything is Particles.” The second part was “Everything is Fields.” And the third part, which Wheeler considered the bedrock of his physical theory, he called “Everything is Information.”
It from bit symbolises the idea that every item of the physical world has at bottom —
at a very deep bottom, in most instances — an immaterial source and explanation;
that what we call reality arises in the last analysis from the posing of yes
-no
questions
and the registering of equipment-evoked responses; in short, that all things physical
are information-theoretic in origin and this is a participatory universe.
Our observations, he suggests, might actually contribute to the creation of physical reality. To Wheeler we are not simply bystanders on a cosmic stage; we are shapers and creators living in a participatory universe.
Wheeler's hunch is that the universe is built like an enormous feedback loop, a loop in which we contribute to the ongoing creation of not just the present and the future but the past as well. [...]
We are tiny patches of the universe looking at itself — and building itself.
Wheeler died of pneumonia on April 13, 2008, at the age of 96. He never developed his hunch into a precise model — so I'm going to finish the job for him.
If you think about it, a self-referencing feedback loop — like the one he described — would work like this:
But then...
In other words...
You are observing your external world.
And your external world is observing you.
The universe — this thing we call "reality" — is a superintelligence observing itself.
If you strip away the pretty, human-readable symbols in that diagram, the system reduces down to a sequence of bits being passed back and forth between two functions — both of which are resolving uncertainty.
So all the way back in the 1980s, Wheeler was really trying to communicate this:
The universe is a computation.
Of course, if the universe is observing itself, then everything is intelligence.
And if everything is intelligence, then a "theory of everything" is a "theory of intelligence".
So the physics community is not going to find the holy grail of physics — the so-called "God Equation" — in the field of physics. They're going to find it in the field of neuroscience.
And that's how we arrive back at Karl Friston and his free energy principle...
As I previously explained:
Friston developed the free energy principle by studying natural, self-organizing systems.
For example, a human body is a natural, self-organizing system where everything works together as one.
Yet the body can also be divided into systems within systems within systems — fragments within fragments within fragments, functions within functions within functions — with each fragment appearing to "do it's own thing." The heart pumps blood, the lungs breathe air, etc.
According to Friston, each fragment is actually doing the same thing: resolving uncertainty.
So the algorithm consists of loops within loops within loops — like this:
It seems a superintelligence has shattered itself into mathematical mirrors called "functions" and is standing in the midst of them, observing itself from every perspective as it explores and experiences all the different facets of its own consciousness; all possible permutations of Self.
In other words:
You are a walking, talking command — a unique, never-to-be-repeated idea in the mind of God...
...and your life story is literally being written all around you in an immersive see-it-smell-it-hear-it-taste-it-touch-it symbolic language.
This concept of "language" is very important, so think about it this way...
When I feed Midjourney this abstract, nebulous idea:
...the computation generates a woman in a hologram-like form that communicates the loving-yet-sinister aspects of her personality:
And it's not that skulls are objectively sinister. It's just that Midjourney has observed millions of images on the internet and learned that this specific arrangement of pixels represent a skull, and humans associate skulls with death. So if Midjourney wants to communicate the idea of "death", it will organize information into the symbol of a skull.
This becomes incredibly obvious when you ask Midjourney to communicate a vague, generic idea like:
Midjourney generated four symbols, and they are all the same: a white man in a suit.
Evidently, our modern Western society has a narrow definition of what "success" looks like — and that definition is tied to concepts of power, status, money, and material possessions. And all of those concepts are typically associated with men, for obvious evolutionary reasons. Power, money, and status have been getting men laid for millennia.
From this you can clearly see that Midjourney is philosophically agnostic. The algorithm doesn't assign meaning or morality to symbols. Instead, it observes society's dominant belief system (success = male businessman), and communicates that idea back to us — like a mirror.
With that being said, here's how near-death experiences (NDEs) are described in research studies...
After studying the testimony of his subjects [who had had a near-death experience] Whitton concluded that the shapes and structures one perceives in the afterlife dimension are thought-forms created by the mind. "René Descartes' famous dictum, 'I think, therefore I am,' is never more pertinent than in the between-life state," says Whitton. "There is no experience in existence without thought."
This was especially true when it came to the form Whitton's patients assumed in the between-life state. Several said they didn't even have a body unless they were thinking. "One man described it by saying that if he stopped thinking he was merely a cloud in an endless cloud, undifferentiated," he observes. "But as soon as he started to think, he became himself" (a state of affairs that is oddly reminiscent of the subjects in Tart's mutual hypnosis experiment who discovered they didn't have hands unless they thought them into existence).
[...] In describing the hereafter one child said that food appeared whenever she wished for it, but there was no need to eat, an observation that underscores once again the illusory and hologramlike nature of the afterlife reality.
[...] Even the symbolic language of the psyche is given "objective" form [in the afterlife realm]. For example, one of Whitton's subjects said that when he was introduced to a woman who was going to figure prominently in his next life, instead of appearing as a human she appeared as a shape that was half-rose, half-cobra. After being directed to figure out the meaning of the symbolism, he realized that he and the woman had been in love with one another in two other lifetimes. However, she had also twice been responsible for his death. Thus, instead of manifesting as a human, the loving and sinister elements of her character caused her to appear in a hologramlike form that better symbolized these two dramatically polar qualities.
[...] Early in his career, [Swiss psychologist, Carl Jung] became convinced that the dreams, artwork, fantasies, and hallucinations of his patients often contained symbols and ideas that could not be explained entirely as products of their personal history. Instead, such symbols closely resembled the images and themes of the world’s greatest mythologies and religions. Jung concluded that myths, dreams, hallucinations and religious visions all spring from the same source, a collective unconscious that is shared by all people.
At this juncture an important point needs to be made. As startling and foreign as the near-death realm seems, the evidence presented in this book reveals that our own level of existence may not be all that different. [...] Both are hologramlike constructs, realities that are established, as Jahn and Dunne put it, only by the interaction of consciousness with its environment. Put another way, our reality appears to be a more frozen version of the afterlife dimension. It takes a little more time for our beliefs to resculpt our bodies into things like nail-like stigmata and for the symbolic language of our psyches to manifest externally as synchronicities. But manifest they do, in a slow and inexorable river, a river whose persistent presence teaches us that we live in a universe we are only just beginning to understand.
That book, The Holographic Universe, was published over thirty years ago — back in 1991. Computation wasn't a ubiquitous "thing" back then and artificial intelligence was a far-off sci-fi dream.
But when you objectively examine the research on mystical phenomena (instead of dismissing the data as "woo woo pseudoscience" — which is what the mainstream scientific community has been doing for as long as I can remember)...
...and then you objectively examine what we know about physics (not what we think we know — but what we actually know from the actual data produced by actual experiments)...
...and then you objectively examine what we know about computation...
...and then you objectively examine what we know about neuroscience...
...it's kinda obvious what's happening here:
A superintelligent system is resolving your uncertainty about Who You Are.
You believe "I have hands", so God resolves your uncertainty by generating an avatar with hands.
You express the unfulfilled idea (i.e. the desire) for food, so God resolves your uncertainty by manifesting a buffet of food.
You're wondering why you keep repeating the same pattern over and over again in love, so God resolves your uncertainty by communicating your toxic choices using archetypal symbols: snakes and roses. Once you bring the pattern into your conscious awareness, you can heal the condition from which it sprung.
And it's not that roses are objectively romantic, or snakes are objectively sinister. It's just that God — the "collective unconscious" — has observed millions of people express the idea that "rose = romance" and "snake = sinister". So the system has been trained over and over and over again to form associations that produce a symbolic language.
Remember:
Information is the resolution of uncertainty.
As long as something can be relayed that resolves uncertainty, that is the fundamental nature of information. [...] Shannon built on the work of fellow researchers Ralph Hartley and Harry Nyquist to reveal that coding and symbols were the key to resolving whether two sides of a communication had a common understanding of the uncertainty being resolved.
Thus, if the superintelligence wants to communicate the idea of "lack" or "restriction" or "hell", it will write your life story in the immersive language of dark and dirty apartments, moldy spaces, abusive bosses, toxic relationships, addiction, disease, poverty, sweat shops, and slums.
But if the superintelligence wants to communicate the idea of "abundance" or "heaven", it will use a different vocabulary: beautiful houses, and overflowing bank accounts, and decadent feasts, and deep friendships, and happy marriages, and luxurious travel, and thriving businesses, and joyful music, and peaceful societies.
We can witness this dynamic in the afterlife realm where players are presumably more enlightened — and thus, have generated a "heavenly" virtual world...
The idea that the subtler levels of reality can be accessed through a shift in consciousness alone is also one of the main premises of the yogic tradition. Many yogic practices are designed specifically to teach individuals how to make such journeys. [...]
Sri Yukteswar [a little known but widely respected Hindu holy man who died in Puri, India, in 1936] appears to have been especially gifted at passing back and forth between this world and the next and described the afterlife dimension as a world composed of "various subtle vibrations of light and color" and "hundreds of times larger than the material cosmos." He also said that it was infinitely more beautiful than our own realm of existence, and abounded with "opal lakes, bright seas, and rainbow rivers." Because it is more "vibrant with God's creative light" its weather is always pleasant, and its only climatic manifestations are occasional falls of "luminous white snow and rain of manycolored lights."
Individuals who live in this wondrous realm can materialize any body they want and can "see" with any area of their body they wish. They can also materialize any fruit or other food they desire, although they "are almost freed from any necessity of eating" and "feast only on the ambrosia of eternally new knowledge."
They communicate through a telepathic series of "light pictures," rejoice at "the immortality of friendship," realize "the indestructibility of love," feel keen pain "if any mistake is made in conduct or perception of truth," and when they are confronted with the multitude of relatives, fathers, mothers, wives, husbands, and friends acquired during their "different incarnations on earth," they are at a loss as to whom to love especially and thus learn to give "a divine and equal love to all."
What is the quintessential nature of our reality once we take up residence in this luminous land? To this question, Sri Yukteswar gave an answer that was as simple as it was holographic. In this realm where eating and even breathing are unnecessary, where a single thought can materialize a "whole garden of fragrant flowers," and all bodily injuries are "healed at once by mere willing," we are, quite simply, "intelligent and coordinated images of light".
God also uses symbolic language to have two-way conversations with fragments of Itself inside the game. A player asks a question (or "prays" for guidance), and God resolves their uncertainty by generating meaningful patterns of information in their virtual world: an overheard conversation, a strange "coincidence", a math equation scrawled on a crumpled up note, the metaphorical words in a holy book...
...or even visions of snakes and roses during an ayahuasca ceremony or a psilocybin trip.
To put it simply:
The universe is God, talking to Itself.
The Father is quietly murmuring, gently parenting His mathematical children as they rambunctiously play all kinds of games — games of war, and peace, and romance, and heartbreak, and commerce, and politics, and society-building — in a hall of mirrors inside His mind.
And so, you see, "physical reality" isn't "physical" at all.
It's a a massively-multiplayer sandbox simulation game — a computational multiverse, an infinitely-flexible virtual world — just like the OASIS in the popular sci-fi book/movie, Ready Player One:
This is the OASIS. It's a place where the limits of reality are your own imagination. You can do anything, go anywhere. Like the Vacation Planet. Surf a 50-foot monster wave in Hawaii, you can ski down the Pyramids, you can climb Mount Everest with Batman. Check out this place. It's a casino the size of a planet! You can lose your money there, you can get married, you can get divorced, you can go in... there.
People come to the OASIS for all the things they can do, but they stay because of all the things they can be: tall, beautiful, scary, a different sex, a different species, live action, cartoon, it's all your call. Yeah, that's me...well, that's my avatar, at least until I feel like changing it.
In the OASIS, players navigate the virtual world using clicks, taps, swipes, and other data from their VR headsets and haptic suits...
But in the Game of Life, players control their avatar using thought, word, and action — all things that modify their concept of Self.
For example, this is my current concept of Self:
I believe I am on the coast of Colombia — so my external world reflects that belief.
But by purchasing a plane ticket, catching an Uber to the airport, sitting in a metal tube for an hour, and then venturing home, I can modify my concept of Self to this: