Steve Patterson

Category: Metaphysics

  • The Problem of Physical Interaction

    Reductive physicalism undermines itself and cannot explain how interaction happens at all. Let’s start with a syllogism:

    1. All interaction is relational.
    2. All relations are non-physical.
    3. Therefore, all interaction is non-physical.

    This is a valid argument; the conclusion necessarily follows from the premises. But are the premises true?

    (1) should be uncontroversial. “Interaction” is inherently relational—an interaction is the state of A affecting the state of B.

    Therefore, (2) is where the argument lies. I’m going to make the case that relations are, by definition, non-physical.

    Atoms “Bumping Into Each Other”

    Let’s take a simple example. Two hydrogen atoms colliding with each other.

    No matter how you conceive of the atoms—as probability clouds, states of space, fields—the question arises: why don’t the atoms simply “pass through” each other? Why do they interact at all? Why do the states of one field/cloud/structure affect the states of another?

    There are different ways to answer, but the answers look similar in the abstract. There are some physical objects and physical processes that give rise to other physical objects and processes, generating a regress of physical explanation. So at some point, physical interaction always ends up appealing to laws, rules, or principles—inherently abstract things that specify what happens when A relates to B in a particular way. It’s a variation of, “Specific structures behave in specific ways because of the laws of physics.” Those laws are sometimes captured with mathematical expressions.

    Even describing the elementary “behavior” of physical objects requires abstractions—the concept of “behavior” itself is relational and non-physical, as “behavior” is about how patterns are related across time.

    So “atoms bumping into each other” turns out to be a mixture of physical and abstract. Atoms are physical (let’s say), and bumping into each other is a highly complex, abstract, relational thing.

    (Note: if you are tempted to say, “The relation is itself physical; it’s like a string attached to the two atoms”, that does not work, because we’d require an explanation for how the string affects the atoms—the string would be, in effect, another atom, and therefore we’d be left asking, “By what mechanism does the string interact with the atoms connected to it?”)

    “It’s Just a Description!”

    One attempt to avoid this conclusion is to say, “No no, physical laws don’t have any causal power. They are merely descriptions of observed patterns!”

    But this only creates a bigger problem. The descriptionists are in principle rejecting explanations altogether, which is not far from a rejection of rationality. If rationality is about finding explanations for things, the descriptionist insists one should be content with no explanation for causal interaction at all.

    Ironically, this position does give credence to the syllogism at the beginning of this article. If interaction is relational, and relations are non-physical, they conclude interaction must not be happening at all, and we’re left with the inexplicable appearance of interaction.

    This leaves us at a fork in the road. When you look deeply enough at the world, it starts to look back, and the temptation is to close your eyes.

  • Everything in Discrete Space

    I’m working through the implications of discrete space and am starting to build some intriguing intuitions. I am trying to reduce the physical world down to a bunch of geometric atoms changing state—essentially, to a grid of voxels.

    I don’t claim the following is true, only that it’s a coherent way to explain a bunch of concepts, including motion, mass, fields (gravity and electromagnetism), local and non-local interaction, fluid dynamics, and even exotic things like quantum tunneling and time dilation. All of these phenomena follow from a simple model with a simple axiom.

    Model: The physical world is composed of discrete atoms of space (“voxels”) in particular states. These states are called “bits.” In the simplest model, there are only two possible states—off or on, 0 or 1, empty or full.

    Axiom: Two bits cannot occupy the same space.

    That’s the setup. Let’s see how far we can push it.

    The basics:

    “Motion” is a transmission of bits through the grid. The voxels themselves do not move; instead, their bit state transfers.

    Local interaction happens when there is a local “collision” of these bits.

    Non-local interaction happens when bit states are connected at a distance.

    Mass is a type of geometric structure in the grid. It is a pattern of voxels.

    Relative to the individual voxel, objects like atoms are incredibly large, dense structures.

    Fields, like mass, are a type of structure in the grid. The difference between mass and fields is only the configuration of their voxels. These different configurations produce different patterns when interacting with the world—e.g. the gravitational and electromagnetic fields create different “forces” because they are different geometric structures. Your car acts differently whether it’s driving on gravel or asphalt, going uphill or down.

    Objects as Extended Structure:

    Where are the boundaries of objects?

    In this model, the totality of an object includes all its geometric structure—both its mass and fields. It’s helpful to think of mass as “macrostructure” and fields as “microstructure”. A gravitational field, then, could be considered as part of the extended structure of an object—reaching beyond its dense atomic structure to interact with distant objects. So for example, the mass of your body affects the motion of the sun; therefore, your extended body reaches to the sun.

    Fields and Biased Motion:

    Motion is the transmission of bits through the grid. As the bits move, they must interact with the existing structures in the grid. This includes macrostructures (atoms) and microstructures (fields).

    The gravitational field is simply a microstructure that biases motion towards its center of mass.

    Take bowling as an analogy. You throw the bowling ball down the lane, towards a triangular structure of pins. If it lands on the left side of the triangle, it will bounce off to the left. If the ball lands on the right side, it will bounce towards the right. So we could say the triangular structure of the pins is biasing the motion of the ball (preventing it from moving straight).

    With a gravitational field, if it’s actually just a microstructure in a grid, we could say it has an inward bias. The greater the mass, the denser the field, the stronger the bias.

    If this is correct, then we could say gravity has a shape. That is, there’s a micro-structure that we call “gravity” which forms a field that interacts with passing bits and nudges them towards a center mass. Rather than space itself being distorted, gravity is a structure in space that distorts motion.

    Electromagnetic fields are more complex. Instead of simply nudging bits towards a center mass, they bias motion depending on other factors like charge and velocity (which can also be interpreted geometrically). “Field lines” would then map real biases generated by underlying geometric structures.

    Exotic Intuitions

    This model also gives a plausible reason for “time dilation.” Larger masses create denser microstructures in the field, which increases the interactions encountered by propagating bits. These interactions effectively slow the transmission of states, resulting in an observable time dilation effect—akin to moving through a denser medium.

    And the “quantum tunneling” phenomenon has a perfect explanation. In discrete space, objects have jagged edges, which means the “energy barriers” between boundaries are neither uniform nor smooth. This means, for purely geometric reasons, some bits can “hop” across boundaries effortlessly. In fact, in discrete space, lines can “intersect” without even sharing a common point!

    If space is discrete, we would expect to see tunneling. I consider the empirical discovery of tunneling to be evidence of the discreteness of space. (Imagine making such a prediction prior to the discovery of tunneling—people would think it’s ridiculous and impossible!)

    A Conceptual Toolbox

    I am early in my journey through theoretical physics and still playing with ideas. I only have a handful of strong beliefs. However, I’m slowly growing my conceptual toolbox, and I suspect some of these tools will be handy indeed.

  • Information and Spirit

    The spiritual domain heavily overlaps the informational domain—the world of patterns. I don’t think there’s a complete reduction of one domain to the other, but there is considerable overlap. Consider a few spiritual ideas.

    “Telling the truth is of spiritual importance. Lies destroy, while the truth heals.”

    This is not a claim about atoms or specific biological entities. This is a claim about general patterns. Humans act based on their ideas about the world. Lies create intrinsically disharmonious patterns, while the truth creates harmonious patterns—at least in the long run. Relationships based on deception are intrinsically unstable and dangerous to humans, while relationships based on the truth are solid and stable.

    This principle is both abstract and true—the truth is a pattern in the abstract domain which lies above the material.

    “You cannot overcome evil with evil, but you overcome evil with good.”

    This is also a claim about patterns—how some patterns relate to others. Revenge begets revenge; this is true both empirically and theoretically. I often think back to a conversation I had with Dr Hirini Kaa from New Zealand who told me about the never-ending cycles of revenge within indigenous Maori groups. These cycles were only stopped once Christian settlers introduced the concept of forgiveness to them, which was revolutionary. The pattern of forgiveness is logically incompatible with the pattern of revenge; the two cannot exist together. Therefore forgiveness, when manifested in the world, is a kind of destruction of revenge.

    “Spiritual warfare is real.”

    How humans relate to other humans is an objective pattern in the world. Some relationships are objectively harmonious (a loving marriage; a safe and healthy community) and other relationships are discordant (a spiteful marriage; a toxic and dangerous community). The reality of social harmony is dependent on an astronomical number of variables, including physical, economic, political, cultural, and psychological factors. All of these variables intersect with the spiritual.

    The health of your body is connected to your nutrition; your nutrition is connected to your ideas; your ideas are connected to a million other information structures. Controversial cultural subjects like pornography and drug use are fundamentally spiritual subjects—their effects are objective and real in the domain of patterns. Spiritual degradation will eventually manifest in the world as physical degradation.

    “The spiritual world is invisible.”

    We “see” patterns differently than we see material objects. A toxic marriage doesn’t have color, a smell, or a surface area. It’s a pattern which is grasped intellectually and felt intuitively—we “see” it in terms of understanding abstract relations and we can “feel” it in our guts.

    For example, the danger and chaos of physical warfare is easy to observe with your senses. The danger and chaos of spiritual warfare has to be grasped with the intellect or felt by the intuition.

    “The spiritual world is timeless and eternal.”

    Political leaders can be assassinated. Assassination itself, as a pattern, cannot be destroyed. Individuals can stop lying. Lying itself is not going anywhere. You can individually run away from love, but you can’t kill the pattern of love.

    Humans are creatures whose heads and hearts seem to intersect the spiritual world. We have the ability to instantiate or not instantiate patterns. There is no murder in my neighborhood, but at any point, that pattern can be instantiated. The spirit of hatred is always there; the question is whether that spirit will find a host.

    There’s a never-ending battle going on—whether or not specific patterns will be instantiated. It takes spiritual strength and discipline to fight it.

    It is important to re-iterate that the spiritual world does not entirely reduce to the informational world. The metaphysics are complex and confusing. The next article will be on Matter and Spirit.

  • Apriorist Geometry and Curved Space

    Countless thinkers for the past two thousand years have appealed to Euclidean geometry as an example of rock-solid reasoning. The proofs in Euclid’s Elements are beautiful deductive structures. One proof builds on the next, and by accepting the starting axioms, you are compelled to agree with the final conclusions.

    The geometric objects within Euclid have properties which can be grasped by logical reasoning alone—e.g. the reason we don’t believe that parallel lines touch is because of understanding the concept of parallel lines, not by observing and measuring them in the world. Kant famously considered this an example of synthetic a priori reasoning and wondered how it could be possible.

    Moderns have since rejected the idea of the geometric a priori, thanks to the discovery of non-Euclidean geometries in the 19th century. And thanks to Einstein in the 20th, physicists claim that our own universe is non-Euclidean—space is curved by mass, they say. That’s what gravity is all about.

    Geometry Meet Economics

    When I was researching the fundamentals of Austrian Economics and their distinctive methodology, there were often analogies drawn to Euclidean geometry—i.e. the unshakeable axioms of human action in economics are akin to the unshakeable axioms of Euclid in geometry. You don’t go out and measure whether the interior angles of a triangle add up to 180 degrees! You just know it by understanding what ‘triangle’ means!

    When I first heard these analogies, I rather liked them. All of my geometric intuitions were Euclidean, and I hadn’t heard about non-Euclidean geometries before.

    As I learned more, I discovered the history of non-Euclidean geometries. I’d speak with Austrian-types about it, but they tended to be skeptical and would ask questions like: What does it really mean for space to be curved?

    Clever mathematicians and physicists say odd things like, “You see, straight lines can sometimes be circular! It just requires space to be curved!” And my skeptical Austrian friends would roll their eyes. Physicists are just playing word games; straight lines cannot be curved by definition!

    I like this criticism. Since that time, I’ve spent a lot more time thinking about these topics and have become critical of the extreme apriorists in economics. And I don’t believe the axiomatic appeal to Euclid works. But not for the reason you might think.

    No Curved Space—>No Curves At All

    The problem is not with the notion of a priori geometry. From what I can tell, mathematicians and physicists are playing word game when talking about “curved space.” I don’t think that concept makes sense. It’s useful for building practical models, but that doesn’t make it true or even coherent.

    The problem isn’t just with the traditional non-Euclidean notion of “curved space.”

    The problem is with the Euclidean conception of curves.

    Ultimately, curved space doesn’t make sense for the same reason that curved lines don’t make sense: curves don’t make sense at all!

    The reason is actually quite simple: the traditional understanding of curves presupposes the infinite divisibility of space. Euclid, along with 99.9% of everyone else, presupposes that space is a continuum—that between any two points of space, there are an infinite number of additional points. Infinities within infinities.

    If you reject the notion of infinite totalities and do not believe space is a continuum, then that means curves don’t exist. Or at least, it means that the curves that obviously do exist work different than we’ve been told.

    I would like to hereby resurrect the notion of the geometric a priori and claim, by appealing to pure conceptual analysis, that geometry does have an underlying logical framework that we all presuppose in order to make sense of things. That framework is non-Euclidean, not because space is curved, but because space is actually discrete.

    Therefore, the traditional conceptions of “points” and “lines” and “spheres” will have to be reformulated accordingly. If you don’t yet have the finitist intuition, examine the following image of “curved space” closely enough until you see its underlying discreteness and total lack of smooth curves.

    (If necessary, move your physical eye closer to the screen until you see the pixels.)

  • Why Language Machines do not have Souls

    It’s been nine months since GPT4 was released. I’m still trying to make sense of things. There’s a dearth of level-headed analysis out there. Most people’s analysis seems to be framed by science fiction novels, or they are still using frameworks inherited from the pre-GPT world, which did not anticipate the success of LLMs. Even the engineers involved didn’t expect LLMs to be as powerful as they are. That’s a sign we need a fresh perspective.

    I categorically reject the hysterical arguments coming from sci-fi metaphysicians. I suspect even the concepts of AGI vs ASI are too grandiose and sloppy. But here’s what I can say with confidence:

    1. We are going into a world where you’ll be able to interact with machines using natural language. In most cases, it will not require computer programming skills to get the machine to do your bidding.
    2. These machines will be able to do things previously thought impossible. They will be able to effectively reason about concepts. This reasoning will be imperfect, but for many complex tasks, intellectual and otherwise, they will outperform humans.
    3. These machines will eventually be embodied and able to navigate the physical world. This navigation will also be imperfect; in some environments, they will be incompetent, and in others, they will outperform humans.

    Language, not Souls

    There’s no shortage of people ascribing souls to these machines. I think I’m starting to understand why.

    I believe LLMs are a genuine technological breakthrough that will result in numerous other breakthroughs. As usually happens, the engineers have created something that nobody understands yet, and it will take time to figure things out. Empirical breakthroughs come first, theoreticians come afterwards.

    In my mind, the breakthrough comes from a revelation about the philosophy of language. This revelation is so incredible and counter-intuitive, that people find it easier to claim the machine has a soul than to update their philosophy of language. In fact, I expect the only people who are nonplussed by the power of LLMs are those with a soft spot for occultism of some sort—those who think words are magical. Let me explain.

    Fill in the _____

    The core of the LLM is pattern-finding and filling, the basics of which already exist in word processors. For example, a word processor can easily tell us how many specific words are in a text. A computer can count the number of times the word “machine” comes up in this article. It can also tell us what words come before and after “machine.”

    We can imagine coming up with all kinds of complicated relationships for the computer to track. For example, we could say, “Find all the sentences which contain the word ‘machine,’ then record the first and last words of those sentences and dump them into an Excel spreadsheet.”

    We could also tell the computer to find probabilities for us. For example, in this article, the word “will” comes up several times and is followed by different words—”will be”, “will not”, etc. We could easily ask the computer to generate a list of words that follow “will” and show their respective probabilities.

    This simple ability allows the computer to complete sentences. Imagine we load a massive document into the word processor—say, every forum thread and social media post on the entire internet that discusses the Bible. We might tell the computer to find a bunch of patterns within this massive pile of words and ask it to provide the most likely word that finishes the following sentence:

    “In the beginning, God created the heavens and the _____.”

    If the computer spits out the correct answer “earth,” nobody would be surprised. It’s not magic.

    Generalized Patterns

    LLMs are not like Microsoft Word, however. They are “trained” on unbelievable amounts of information—large percentages of the entire internet, which costs millions of dollars. With this much text input, LLMs become more magical when we ask them to produce novel sentences for us. Instead of asking it to complete a well-known sentence, we ask it to create something new. Poetry is one popular example. I know it’s overused, but it really does demonstrate something incredible about LLMs. So for illustration, I’ve asked GPT4 to create a ten-sentence limerick that summarizes the entire Biblical story. It responds:

    1. In Eden’s garden, all was fair and bright,
    2. Till man chose wrong, and lost the path of right.
    3. Prophets foretold a Savior pure and true,
    4. To guide the lost and make all things anew.
    5. Born in a manger, under star’s soft light,
    6. Jesus taught love, to conquer darkest night.
    7. On the cross, His life He did resign,
    8. To save all souls, a sacrifice divine.
    9. The tomb found empty, He rose to the sky,
    10. A promise of hope, for you and for I.

    How can the LLM do this, when it’s never been trained on “Biblical limericks” before? How does it understand what “rhyming” is? How does it know the general rhythmic style of a limerick? How did it know the core concepts of the Biblical story to choose? These things were not directly programmed into the machine.

    If we put metaphysics aside for a moment, we can say this is a demonstration of some form of a high-level understanding of concepts. Call it a practical understanding, or an effective understanding. The machine understood my request and gave me a highly creative and correct output.

    Poetry is one thing; scientific discussion is another. I’ve had lots of high-level philosophical and scientific discussion with GPT4, and it does a great job. It can effectively follow along with better-than-average conceptual nuance. It can effectively reason and even recognize logical fallacies. It’s not perfect, but it’s already shockingly good.

    This shouldn’t work. I have oversimplified the way these machines are trained, but the essential mechanism is correct. They really are just probability machines trying to create strings of symbols that you’ll find acceptable.

    Understanding and Ensoulment

    It is natural to think that a mere word-machine cannot truly understand the meaning of our concepts—at least, not in the way humans understand them. Understanding is very abstract, very personal. Grasping and comprehending abstractions feels like something we do within our souls. I think this is the reason people are attributing souls to LLMs—the computer seems to understand what we are saying, in a way we have only encountered with other humans before.

    Humans are, of course, people who have goals, plans, ambitions, personalities, and consciousness. So, LLMs are instinctively categorized as the same kind of entity—we’ve never encountered this ability separate from a mind before.

    But alas, LLMs are not people. We have not created Frankenstein. We don’t have any reason to believe the machine is conscious or in possession of a mind that is similar to a human’s. It appears to be exactly as designed—a word machine and nothing more.

    Instead of speculating about ensoulment, I think we need to update our philosophy of language.

    Words and Logical Structures

    It turns out, a whole bunch of stuff is encoded in our language. So much, in fact, that by tracking patterns in our language, a mindless machine can effectively reason through concepts. That is, by mimicking our language, the machine can mimic our reason.

    This ability did not come from understanding the meaning of individual words in isolation. It didn’t come from—as with a human—receiving extra-linguistic training from a parent who can point at objects to give them ostensive definitions.

    No, it came from analyzing a massive amount of words on the internet. The patterns within our language, across our sentences—the mathematical patterns of the meaningless symbols themselves—are so strong and definite, a machine is able to imbibe them and gain the ability to use natural language.

    Let me repeat: there is so much abstract structure in our language—the patterns are so overwhelmingly clear, consistent, and objective—that by mindlessly figuring out the probability of one symbol following another, a machine can effectively reason better than the average person for a large number of cases. Extraordinary.

    There’s an analogy to atoms and molecules here. Imagine each word is an atom, and each phrase a molecule. Those molecules can combine with others to form sentences, paragraphs, and other word-forms. Now imagine the computer has the ability to store, say, a trillion such word-forms for reference—a trillion identifiable and repeatable patterns and connections between words.

    That turns out to be roughly analogous to what GPT4 does, and a trillion+ patterns is apparently enough to find high-level embedded patterns of reasoning among the words.

    I can hardly imagine a more mind-blowing idea in the philosophy of language, which is why I claim only the occultists might be unsurprised. Language seems to be a sort of intermediary between the abstract world and the physical world.

    There are the underlying physical words—the ink on paper, or bits on a hard drive. Then, there are sentences—higher level patterns of words. These sentences encode higher-level patterns, concepts, forms, abstractions, and logical structures, all of which are equally objective and real—they are so real, a mindless computer can detect them and even use them for “understanding” the world.

    It’s perhaps the clearest demonstration of Platonism ever.

    There’s a lot more to say about the subject, but this article is long enough. I hope this contributes to more sane discussion about AI.

    Before LLMs, our language could only be understood by other sentient beings. We need to recognize this is not because in principle understanding language requires sentience; we now have an empirical demonstration that a sufficiently advanced calculator can do it, too.

  • Is Everything Quantifiable?

    There’s an ongoing quest to reduce everything to mathematics. It’s part of the reason we’re in a dark age. Seems like a good time to ask the question: is everything quantifiable? Were the Pythagoreans correct in saying that “All is number”?

    Consider three statements:

    1 Alice is taller than Bob.

    This claim is easy to quantify. Alice and Bob have bodies that are composed of atoms which take up space—they exist in the geometric realm. To figure out whether Alice is taller than Bob, count the number of geometric units from the bottom of the feet to the top of the head. If A > B, then Alice is taller than Bob. Simple.

    2. Alice is happier than Bob.

    This claim is harder to quantify. “Happiness” does not seem to exist in the geometric realm—it’s more mental than physical, and the mental is notoriously hard to quantify. There aren’t happiness-units that we know of. But let’s try.

    Suppose that we create a new type of scanner. By pointing it at someone, you can “see” some underlying brain and neurological activity and detect chemicals like dopamine and serotonin. Perhaps there isn’t a specific happiness-unit that you can count, but there are a bunch of corresponding things you can—good enough to reliably predict that somebody is “happy” or “not happy” depending on the readout of the scanner.

    In that situation, we could say that for practical purposes happiness is quantifiable, because we have a happiness-meter which could reliably measure the quantifiable aspects of happiness.

    3. Alice is more virtuous than Bob.

    This claim is quite difficult to quantify. Virtue does not appear to be geometric nor a description of a body-state. It’s very abstract.

    When we say that Alice is virtuous, what does that mean? What are we describing?

    I don’t know how to describe it precisely. It’s an abstract relationship between a person, their actions, and their environment. It rests in the moral, ethical, spiritual, and intellectual dimensions that intersect with the geometric/physical but is not reducible to it.

    Can we imagine scanning Alice with a virtue-meter, like we scanned her for happiness? Are their enough corresponding physical states? Instead of dopamine, might we find virtuamine?

    Probably not. Perhaps we could some day, but it’s so speculative that it feels silly. I think the reason is this:

    The more abstract something is, the more difficult it becomes to quantify.

    Virtue is more abstract than happiness; happiness is more abstract than height. The further up the abstraction chain, the less quantifiable—perhaps by definition.

    Being abstract and non-quantifiable doesn’t imply being less real. Virtue is just as real as height. Abstract stuff is just as real as concrete stuff. It just isn’t quantifiable.

    It’s one thing to say, “In theory, we might some day be able to quantify virtue.” It’s quite another to say, “Hey guys, here’s my virtue scanner that I developed with the most sophisticated modern technology!” Anybody claiming such a thing in 2023 does not understand the complexity problem.

    Unfortunately, that’s the type of mistake that is currently plaguing modern thought—the premature application of mathematics to domains which are too complex for quantification. Biology, economics, epidemiology—scientists are way too quick to develop mathematical models, and they end up oversimplifying the world to an absurd degree. Statistical reasoning generally is shot throughout with these kind of abstraction/quantification errors.

    Non-mathematical reasoning is far less impressive to the modern mind, but it accounts for more complexity; it sits higher in the abstraction hierarchy. Eventually we’ll figure this out.