Just what you wanted! And just in time for National Cherry Popsicle Day! A summary of my 2024 book, The Weirdness of the World, by the author himself. You probably haven't had time to read the whole thing (I don't blame you). So here's a synthesis of what I see as the most central ideas.
(I wrote it for a book symposium forthcoming in Journal of Consciousness Studies with critiques by Edouard Machery, Luke Roelofs, Susan Schneider, and a reply by me [with Jacob Barandes, co-author of one of the chapters that Machery targeted in his critique]. I'll share links to the critiques when they're available, along with the final manuscript of my reply.)
----------------------------------------------------------------------------------
1. Bizarreness and Dubiety.
On fundamental issues of consciousness and cosmology, something both bizarre and dubious must be true, and we’re in no good position to know which of the various bizarre and dubious possibilities in fact holds. That is the guiding idea of The Weirdness of the World.
Interpretation of quantum mechanics isn’t a main theme, but it nicely illustrates the thought. According to many-worlds interpretations, the world is constantly splitting into myriad near-duplicates. According to some versions of the Copenhagen interpretation, observation by humans somehow causes the collapse of the wave function. Other possibilities include Bohmian mechanics and objective collapse theory. All the possibilities are bizarre: They starkly violate the ordinary common sense of nonspecialists. All are also dubious: We have no compelling epistemic grounds to believe any one over all the others. We see several theoretical options, A, B, C, and D. Each option is bizarre and dubious, and it’s a good guess that any viable new options that we add to the mix will also be bizarre and dubious. Quantum mechanics will never fit neatly into common sense, unless common sense changes radically. And there’s no near-term or even medium-term hope of a decisive theoretical resolution.
Across many fundamental issues, bizarreness and dubiety reign. Common sense fails us. So also does scientific and philosophical theorizing. We are adrift in weirdness, unable to anchor.
2. Mind and Body: Universal Bizarreness.
Consider the general relationship of the mind to the apparently physical world. The live options appear to be (1.) materialism, (2.) substance dualism, (3.) metaphysical idealism, or (4.) one of a grab-bag of compromise or alternative views such as panpsychism, property dualism, or transcendental idealism. Each option is both bizarre and dubious, either on its face or when theoretical choices are made and consequences explored.
The bizarreness of materialism is sometimes underappreciated. Weirdness explores materialist bizarreness in two ways (but not only these two):
The pain trilemma. Horn 1: If pain depends on having a brain of a certain physical type, then alien species built very differently from us can never feel pain, no matter how behaviorally and functionally similar. Horn 2: If pain depends instead on having a certain functional profile of inputs and outputs, then strangely constructed entities (e.g., huge assemblies of beer cans and wire) can feel pain if they have the right functional profile; and, conversely, no one can feel pain if their pain plays an unusual causal role. Horn 3: The standard ways to escape these two simple options appeal to typicality or historical function. This entails denying the local supervenience of pain: Two entities can be molecule-for-molecule identical, writhing and evading equally, emitting exactly the same sounds of protest, while one is in pain and the other is not. One of these three options might be correct, but none comports with common sense. A fourth might be discovered, but it’s a good bet that it will not preserve common sense appreciably better.
Group consciousness. If materialism is true, it ought to be possible to arrange groups of people into configurations that give rise to consciousness at the group level, over and above the consciousness of each individual considered separately. Arguably, the United States, conceived as a concrete entity with people as its parts, is such a group. It possesses the main features characteristic of brains in virtue of which materialists ordinarily think brains give rise to consciousness: massive internal information transfer, sophisticated representation and self-representation, sophisticated action in its environment, complex learning, preservation of its boundaries, a long history of interaction with entities of its type in a natural world, and so on. There are various ways in which materialists can avoid concluding that the United States is literally conscious. For example, they can insist that no conscious entity can have a smaller conscious entity as a part. Or they can insist that consciousness requires a highly specific type of internal processing or self-representation, which is not easily achievable in the United States. But each escape brings other counterintuitive consequences in its train. It doesn’t follow that the United States is conscious. But we can’t comfortably retain all of our common sense assumptions about such matters. Something has to give. One or another bizarre thing must be true.
The bizarreness of substance dualism is also sometimes underappreciated. One way to appreciate the bizarreness of dualism is the scope of mentality quadrilemma. Which organisms have immaterial souls? Horn 1: Only human beings. But then, if the soul is the locus of consciousness, even dogs and apes must be entirely nonconscious. Horn 2: Every organism has an immaterial soul. But then even bacteria and slime molds are conscious. Horn 3: There’s a sharp line in the sand between ensouled and unensouled animals. But phylogeny reveals a smooth gradation of capacities and physiologies, with no plausible place to draw a sharp line. Horn 4: Having a soul is not an all-or-nothing matter. Some animals (frogs? ants?) have kind-of-souls or 60% of a soul. But no dualist has ever made good metaphysical sense of this idea.
I think I need not argue that metaphysical idealism, panpsychism, transcendental idealism, and other well-known alternatives to materialism and substance dualism also have points of sharp conflict with common sense.
In the entire history of philosophy, East and West, no philosopher has ever developed an account of the mind-body relation that fully conforms to common sense. By his own admission, even “common sense” philosopher Thomas Reid couldn’t preserve common sense on this issue. For example, he held that physical objects require the action of immaterial souls to sustain their shapes. Philosopher after philosopher delivers bizarreness, both by the standards of their own culture and by the standards of mainstream early 21st century Anglophone readers. The most plausible explanation of this historical pattern is that commonsense opinions about mind and body form an incoherent set, so that no coherent philosophy is able to adhere to the entire package.
3. Mind and Body: Universal Dubiety.
The dubiousness of every answer to fundamental questions concerning the relation of mind and body follows from two considerations.
First is the lack of a good method to resolve such disputes. There is, for example, no good material-world-ometer by which metaphysical idealism can be refuted. Nor is there an epiphenomenal-substance-ometer that can refute epiphenomenalism. Deniers of causally inert epiphenomenal substances usually claim that the universe is appealingly simpler without them. But simplicity is a complex business. Arguably the simplest cosmology is solipsism: Only one thing exists, your own mind. (In a co-authored chapter of Weirdness, I attempt to refute solipsism experimentally with the help of an interlocutor who convinces me to admit the existence of something in the universe that surpasses me at chess.)
Even if we commit to materialism, no general consciousometer can be built that doesn’t beg the question against rival views. To develop a general theory of animal consciousness, we need to presuppose something about the distribution of consciousness in the animal world. How plausible are conservative views that restrict consciousness only to the more sophisticated mammals and birds? How plausible are liberal views that attribute consciousness to insects, worms, and oysters? A theory of consciousness needs some initial assumptions about approximately how sparse or abundant consciousness is in the animal kingdom, or some assumptions from which sparseness or abundance quickly follows. But those assumptions will rightly be unappealing to people who begin with very different initial assumptions. No decent resolution of this tight epistemic circle is on the horizon, no plausible path by which advocates of sparseness can be rationally compelled to abandon their view in favor of abundance or vice versa.
Second, grand-scale cosmological uncertainty warrants uncertainty about the metaphysics of consciousness. For example, if we are living in a computer simulation, then what we think of as the material world might be simply how the underlying computer presents itself to us. The real grounds of conscious experience might be very different than we suppose. Or if the universe is infinite, its laws might be radically different in other regions (which would help address the “fine tuning” problem of why the cosmological constants seem so well suited for life). Materialism might then be a local contingency rather than a metaphysical truth. If a consciousness-based interpretation of quantum collapse is correct, then consciousness might be among the fundamental properties of the universe rather than something arising from physical matter as we ordinarily understand it. I’m not saying any of these possibilities deserves the bulk of our credence, but if mainstream cosmology is correct, we have seen only a very small, perhaps infinitesimal fraction of reality. We are like fleas on the back of a dog, watching a hair grow, and saying “Ah, so that’s how the universe works!”
4. 1% Skepticism.
A radically skeptical scenario is one in which a huge swath of what you ordinarily take for granted is radically false. If you are now dreaming, or if you are a radically deceived brain in a vat, almost all of your perceptual beliefs are false. If the world was created ten minutes ago, almost all of your beliefs about the past are false. It is not unreasonable, I suggest, to have about a 1% credence that some radically skeptical scenario obtains.
There are no grounds for assigning more than a tiny credence to the possibility that aliens extracted your brain while you were sleeping last night, teleported it to Alpha Centauri, and are now feeding it fake input. However, there are grounds for thinking you might currently be dreaming. Past false awakenings show that people sometimes mistakenly judge themselves to be awake. Some leading scientific theories describe dream experiences as subjectively indistinguishable from waking experiences. It’s not unreasonable to assign perhaps a 20% credence to the possibility that you sometimes have, while asleep, experiences much like the experience you are now having, along with a disposition to falsely judge yourself awake. Once I admit this, I find it hard to sustain absolute 100% confidence that I am now awake, or even 99.9999% confidence. About 0.1% credence, plus or minus an order of magnitude, seems about right.
Similarly, for various reasons I think I’m probably not an Artificial Intelligence inhabiting a short-term computer simulation. I feel highly confident that’s false – 99.9% confident! But again absolute certainty, or even 99.9999% certainty, seems out of reach.
Admitting about a 0.1% to 1% credence that I’m in a skeptical scenario is not behaviorally inert. For example, in a situation of almost exact indifference between a short-term gain and a long-term gain, 1% skepticism ought to tilt the balance in favor of the short-term gain.
5. Kant Meets Cyberpunk.
Transcendental idealism is a historically important metaphysical position, associated with Kant, but now unfortunately neglected among non-historians. I offer a how-possibly argument for transcendental idealism. As I intend the phrase, transcendental idealism conjoins two theses: (1.) Spatial properties depend on our minds. (2.) The fundamental features of things as they are in themselves, independent of us, are unknowable to us.
Here’s how transcendental idealism might possibly be true: We are AIs living inside a computer simulation. But the computer running the simulation isn’t material. Instead, it’s an immaterial soul – the mind of an angel. If we accept Turing’s standard account of digital computation, computation does not require a material substrate, as long as there are the functional equivalents of machine states, symbols, a read-write head, and transition functions. Spatiality might then just be our minds’ way of making sense of the structure of the immaterial soul we inhabit.
That we are sims in a virtual reality instantiated in an immaterial angel is not, I admit, very likely! But once we grasp this possibility, we can see more generally how a version of transcendental idealism might be true – perhaps one closer to Kant’s, in which fundamental reality is even less comprehensible than computation in an immaterial soul.
6. Infinite Puppetry.
If standard physical theory is correct, the universe will eventually cool into heat death. During heat death, occasional chance fluctuations will produce lower entropy states, for example when several particles by chance converge into an organized pattern. There appears to be in principle no limit to the size of these fluctuations, and no future temporal boundary on the universe. Given infinite time, large configurations will eventually fluctuate into existence, including particle-for-particle duplicates, within some error tolerance, of your brain, of your brain and your body, and even of the whole galaxy. This is the well-known Boltzmann brain hypothesis.
In a chapter co-authored with Jacob Barandes, we defend the physical plausibility of this hypothesis and derive some causal consequences. Suppose you raise your hand right now, intending eventually to bring into existence a future duplicate of yourself who thinks “I came into existence because an earlier duplicate of me raised their hand”. This intention will likely be fulfilled. By raising your hand, you set in motion a ripple of causal effects that echo forward into the post-heat-death universe, until eventually it produces an entity of exactly the sort you imagine – one who would not have existed had you not raised your hand in that way at that moment (though duplicates of that entity would have existed elsewhere).
7. Debatable Personhood, the Full Rights Dilemma, and the Design Policy of the Excluded Middle.
We don’t understand the metaphysics of consciousness. Even if we assume materialism, we don’t know the right scientific theory of consciousness. We will likely soon create AI systems that are, according to some plausible theories, as richly conscious as ordinary adult human beings, or even more so, while according to other plausible theories, they are no more conscious than a toaster.
I assume that sufficiently rich conscious cognition is sufficient for personhood – sufficient, that is, to deserve humanlike moral consideration and humanlike rights. I also assume that an entity with as little consciousness as a toaster is far short of deserving humanlike moral consideration and rights. Thus, if we create entities whose consciousness is radically debatable, their rights will be radically debatable too.
How should we treat such debatable persons, if we ever create them?
Here we face an unfortunate dilemma. If we deny debatable persons the rights they might deserve, we risk perpetrating the moral equivalent of slavery and murder against them, perhaps on enormous scale. This possibility has been imagined again and again in science fiction, from Asimov to Star Trek to Westworld. If, to avoid that moral risk, we err on the side of granting humanlike rights to debatable persons, we risk sacrificing real human lives for entities without interests worth the sacrifice. Human rights presumably include citizenship, reproduction, property ownership, and the vote. The consequences could be immense, especially if the AI systems in question have interests different from ours and can reproduce quickly. One compromise might be to grant only limited rights – for example, no right to reproduce, no right to vote, no right to privacy of thought. That might be safer for humans, but it would be seriously oppressive if the entities do in fact deserve full humanlike moral consideration.
The full rights dilemma has only one good solution: Don’t create debatable AI persons. I call this the Design Policy of the Excluded Middle (originally proposed with Mara Garza in 2015). Avoid creating AI entities if it’s unclear whether they would deserve moral consideration similar to that of human beings. Either create entities whose moral standing is known to be far less than ours, and prioritize human welfare; or go all the way to creating entities known to be persons and give them the rights they deserve.
[The consciousness question is the topic of my just-released book AI and Consciousness, and the ethical question is the topic of my book in circulating draft, Humanlike: A Defense of AI Rights.]
8. Weirdness and Wonder.
Imagine a planet on the far side of the galaxy, one we will never interact with, forever blocked from view by the galactic core. What do you hope for this planet? Do you hope that it’s a sterile rock? Or do you hope that it hosts life?
I think you will join me in hoping that it hosts life. Not just microbial life, but even better, plants and animals. Not just plants and animals, but even better, intelligent creatures capable of abstract thought and long-term social cooperation, capable of love and art and science and philosophy. That would be an amazing, wonderful, awesome planet!
Earth, for the same reasons, is an amazing, wonderful, awesome planet. Among the most awesome things about Earth is this: Certain bags of mostly water can contemplate profound and difficult questions about the fundamental nature of things, their position in the cosmos, the grounds of their values, the limits of their own knowledge. A world in which no one ever did this would be sadly impoverished. The ability to reflect seriously on these questions is cause for celebration even if we can’t find our way to the answers.
Earth is a bright spot of value in the cosmos in part because it contains human beings who, driven by curiosity and a non-instrumental desire to address hard questions, attempt some philosophy. Philosophical inquiry makes the entire planet better. Assume for a minute the worst about philosophy: that it has no instrumental value for other ends, that philosophers will never agree on the correct answers to anything, that studying ethics makes us no ethically better (as my own empirical research on the behavior of ethicists tends to suggest), that philosophy often only serves to rationalize our vices, that our philosophical inclinations reveal more about our fixations and pathologies than about the world we seek to understand. Fine and (for the sake of argument) granted! Maybe it really is that bad. And yet, simply to pose a philosophical question is already a magnificent cognitive achievement. Our capacity for philosophical inquiry separates us from all other species. It makes Earth a very different and more interesting planet than it would otherwise be.
By reading this blog post, you have made this region of the cosmos more wondrous. Now go enjoy a cherry popsicle.
