An Argument for Vastly Many Observers Smaller than a Proton

Here's an argument that I'm sure will give my philosopher of physics collaborator Jacob Barandes the fits. The conclusion will be that there are vastly many complex structures smaller than a proton, right here in our region of spacetime, populated by vastly many scientific and philosophical observers. Look at that drop of water running from the faucet to the drain. Within it are many tiny people, living many tiny lives.

[Cancer can go hang. I'll keeping fly my weird philosophical speculations on this blog as long as I can. Naturally, I'm hoping to land in the long statistical tail, so that the field won't be rid of me anytime soon.]

Assumption 1: The universe is infinite.

The observable portion of the universe, of course, is not infinite. It extends a only minuscule 46 billion light-years in every direction from us. But there's no reason to think the universe ends exactly at the observational horizon, with us at the center! How inelegant and un-Copernican that would be.

An infinite universe is slightly more elegant than one with an edge. (Why would there be an edge? What would happen there?) And current physical estimates also tend to suggest that at the largest scales, the universe is spatially flat, rather than looping back upon itself in a closed but edgeless geometry. Lots of theoretical physicists think infinitude is a good guess. In any case, I will assume it.

Assumption 2: If the universe is infinite, complexity recurs at large scales.

This assumption is less standard and harder to justify, but not, I think, wholly implausible. Imagine zooming out and out and out so that our galaxy becomes a speck, less than a speck, less less less than less than less of a speck; keep zooming.

It might be that once you've zoomed out far enough, every scale looks smooth and boring -- that things simplify at large scales and zooming further out yields only more featureless simplicity. But it might instead be that complexity reappears at some larger scales, generating structures as intricate as those at our familiar scale. Given infinitude, we can keep zooming out until we find those scales. My second assumption is that however far we zoom out, there will always be larger scales at which complexity reappears.

Such complexity could arise from large-scale interactions among processes governed by the laws we already know. Or it could arise from phenomena we're familiar with but don't yet understand (such as dark matter or dark energy). Or it could arise from unfamiliar processes with measurable effects only at enormous scales. Compare: If we were proton-sized, we'd have difficulty detecting gravity, which accumulates noticeably only at macroscopic scales. There is no guarantee -- nor indeed any particular reason -- to think that all physical laws must operate at scales we can readily detect.

Assumption 3: If the universe is infinite and complexity recurs at large scales, so do life and intelligence.

I see no reason to suppose that the evolutionary processes that produced life and intelligence on Earth could occur only at our scale. Evolution by natural selection requires patterns that reproduce with stable heritable variation and differential reproduction rates depending on that heritable variation. If complexity recurs at large scales, some of these scales will plausibly host complex processes capable of sustaining evolution over long time periods, producing entities approximately as intelligent as we are.

Since we have infinitude to play with, if one otherwise complex level for whatever reason can't host stable evolutionary processes like we see on Earth, forget that level and keep going up.

Now let this evolution run until we have scientific and philosophical observers.

Putting Assumptions 1-3 together suggests that there are infinitely many cosmic observers, capable of philosophy and science, vastly larger than we are, whose worlds contain us as tiny parts -- far tinier relative to them than a proton is relative to us. There's no particular reason to think their sensory organs or scientific equipment could detect events at our scale, which would be to them sub-sub-sub-sub-microscopic.

Assumption 4: Given Assumptions 1-3, it would be an un-Copernican coincidence if we happened to occupy the smallest scale hosting intelligent observers.

According to the Copernican Principle in cosmology we should assume as a default -- pending contrary evidence -- that we don't occupy any particularly special position in the universe, such as its exact center. To dwell at the very lowest, most fundamental level would make us special in an un-Copernican way. Why should we, of all the infinitely many observers at infinitely many scales, be so lucky as to occupy the foundation?

Copernican reasoning thus gives us grounds to think that if there are larger philosophical and scientific observers there are also smaller ones.

Now if we had good reason to rule out life-supporting complexity at smaller scales, we'd know that we are at the lowest level. But our failure to detect complexity at smaller scales doesn't establish its absence. As I mentioned above, there's no particular reason to think all scientific laws operate at scales we can measure or perceive.

Thus, Assumptions 1-3, plus a plausible application of the Copernican Principle and the absence of a reason to think we do dwell at the lowest level, invite the conclusion that within that tiny droplet falling toward the drain are realms of complexity far too small for us to register, populated by vastly many tiny tiny lives and tiny tiny philosophers and scientists, gazing out across their portion of the droplet, wondering what exists beyond the reach of their senses and equipment.

[droplets from a faucet: image source]

AI and Consciousness: A Skeptical Overview Available Free Until Oct 19

My recent book AI and Consciousness: A Skeptical Overview is free online today and tomorrow (through Oct 19, presumably UK time). I don't think the hard copy is available yet. Get it while it's free! Here: https://www.cambridge.org/core/elements/ai-and-consciousness/E77C92088DA3C9F89E7FE7C75CBB1896

I Have Cancer

Friends and readers: I have cancer.

The type of cancer is not yet fully confirmed, but all signs point toward kidney cancer that has spread to my hips, back, and pleura. On the bright side, all my blood tests so far are normal, suggesting that the cancer is moving slowly, which is typical of kidney cancers. There is no evidence yet that the cancer has spread to any vital organs, and I am not in imminent danger of death. With new treatments, I am hopeful that I have several good years, perhaps many good years, still in front of me.

I do not plan to crawl into a hole. I will continue to do what is important to me, and one of those important things is discussing philosophy through my blog and social media. I have been posting at least weekly on The Splintered Mind for more than twenty years, and I hope to continue to do so, though it is likely that I will start to miss some weeks.

I am inspired in part by the example of my dear departed friend, collaborator, and fellow blogger Helen De Cruz, who was so brave, open, and positive-spirited through her slow death by cancer -- at a younger age and with a worse prognosis than I.

In case any of you are worried in light of last week's post about self-sacrifice with terminal diagnoses, I assure you that although my reflections at the time were more immediately personally relevant than they might have seemed, while I was awaiting results, those thoughts remain entirely hypothetical. I remain hopeful and will fight this thing like a cornered cat.

Giving One's Life

[Content warning: suicide]

This will be a morbid post; forgive me. It will be hard to write this well. I will write it badly.

You are young or middle aged -- healthy you think. You get a fatal diagnosis out of the blue, let's suppose. You have a few months to live and no cure.

You will seek whatever pleasure, peace, and companionship feels right in those last months.  You will complete a few important projects maybe, and prepare your loved ones as best you can for a world without you.

But you also have an unusual opportunity. You could give your life. You could die in a good cause, if you knew a good cause that would be advanced by your death. Perhaps this is worth contemplating. In the past, you have given money and you have given time, but now you could, if it seemed right, give something very different from money or time.

Socrates drank the hemlock. Would Plato have been as inspired if Socrates had fled Athens and died quietly ten years later in Corinth? Thich Quang Duc immolated himself in the streets of Saigon in 1963, shocking the world and drawing attention to the oppression of Buddhists in South Vietnam. Many, many soldiers have hopelessly charged the enemy, or stayed behind to cover a retreat, accepting the likelihood of death in what they believed to be a good cause.

Need volunteers to sprint into the intense radiation of a nuclear meltdown and contain the damage?

Is there some great scientific advance that could be achieved only by exposing an experimental subject to some fatal intervention? Of course we would never want to do that -- but now, maybe, if one could volunteer in the right kind of way for the right kinds of reasons, the "of course" needn't be quite so certain..

Is there some terrible but ethically compelling crime one could commit, a crime that would result in one's death or a life sentence, yet leave the world substantially better?

In the spirit of "effective altruism", one might wonder whether there is some way to use those final months to accomplish something that cannot be accomplished by gifts of money.

I have touched on four categories: symbolic martyrdom, dying to save others, lethal research, and ethical crime. I will not explore them systematically. Something in me rebels against doing so; perhaps it's too close to recommending suicide. This isn't a domain where I want to apply the cold equations of effective altruism. Nor do I want to breathe even the whisper of a suggestion that anyone who does not think this way in their final months is being selfish. As Kant would say, with considerable plausibility, one's life -- even the last little bit of it -- should not be treated merely as a means for the benefit of others, but valued for its own sake.

Still, if I received a terrible diagnosis, thoughts of this sort would arise in me, I think. They might even constitute some kind of comfort -- some new applied philosophy of direct practical importance to vividly consider, and a way to think of my situation as an opportunity -- before I, presumably, just died in the ordinary, sensible way that most people with fatal diagnoses do.

The Appeal of (Semi-?) Biological Naturalism about Consciousness

Biological naturalism about consciousness will, I predict, soon be hot. As AI systems start to look more and more like conscious or sentient entities, at least superficially, doubters will be increasingly drawn to arguments against their consciousness. Especially appealing will be arguments that characterize consciousness specifically as a biological phenomenon, one that can't occur in ordinary, non-biological computer systems. Accept biological naturalism, and it looks like you're safe from having to take the possibility of AI consciousness seriously for at least a couple of decades.

(Non-naturalist views could also serve this purpose. One might hold, for example, that God installs conscious souls by miracle. But most mainstream philosophers and scientists prefer explanations in terms of purely natural phenomena.)

Searching for a Biological Property That Is Necessary for Consciousness

Biological naturalist arguments against AI consciousness can be expressed in two steps:

(1.) Property X is a biological property, of the sort that cannot be instantiated in standard (non-living) AI systems.

(2.) Property X is necessary for consciousness.

Different versions of biological naturalism propose different candidates for X.

Peter Godfrey-Smith for example, appeals to metabolism. For this appeal to work as an argument against AI consciousness, two things have to be true. The relevant sort of metabolism must not be instantiable in standard AI systems, and a system without metabolism in this sense must necessarily lack consciousness. Given the first condition, drawing power from a solar panel can't count as "metabolic". Godfrey-Smith suggests that molecular-level spontaneous motion is crucial to metabolism in the relevant sense. The second condition requires that metabolism in this specific sense is necessary for consciousness. Godfrey-Smith suggests it might be necessary, but he offers little positive argument. The suggestion remains an intriguing possibility rather than a fully developed position.

In an already-influential article published last month in Behavioral and Brain Sciences, Anil Seth proposes autopoiesis as the X. Autopoiesis -- a concept first developed in a 1972 book by Humberto Maturana and Francisco Varela -- occurs when a system continuously regenerates its own material components, actively maintaining a boundary between itself and its surroundings (Seth 2026, sec 4.1). For autopoiesis to serve as X, standard AI systems must be incapable of it, and consciousness must require it.

Neither claim is obvious. Regarding the first, it seems that we can, with a bit of creative engineering, design an AI system that regenerates its own material components and enforces a boundary between itself and its surroundings. Imagine a solar-powered robot that seeks out solar charge. It has redundant parts that it monitors for defects and which it can replace and rebuild by ordering shipments of snap-together components. For more details, see this post and my commentary on Seth's article.

Regarding the second claim, even if we grant that autopoiesis is limited to biological systems, it's not clear why consciousness should require it. Seth appeals to predictive processing and Friston's free energy principle, but these appeals are highly speculative. They do not establish the necessity of autopoiesis for consciousness, as several of the other commentaries emphasize.

There's substantial plausibility in the thought that consciousness requires using energy and some kind of stability over time. But even granting this, it remains unclear why consciousness should require metabolism or autopoiesis understood narrowly enough that no ordinary AI system could instantiate them.

Some of the replies to Seth's article suggest other candidates for X. For example, Cao, Gottlieb, and Moore propose evaluative states that are internalized and integrated. Again, the view is potentially attractive, but both premises are highly debatable.

First, it's not clear that a sufficiently sophisticated AI system couldn't meet Cao and colleagues' condition. It might have an optimization process that corrects deviations from a goal state, plus explicit representations of these processes, plus the capacity to integrate many factors in achieving one or more goals. This seems especially plausible for autonomous embodied robots who must negotiate complex situations in the wild. Second, it's not clear why consciousness requires valuing rather than simply registering. It might! But this is the type of proposed requirement for consciousness that I criticize in my recent book AI and Consciousness: broadly plausible, but difficult to defend rigorously.

Biologicity as a Matter of Degree?

In his reply to my commentary (section 5.2 here), Seth notes that my hypothetical robot depends on the environment for ready-made components. But this is arguably only a difference of degree from the human case, not a difference in kind. We humans don't manufacture all of our components from the atomic level up. We obtain some amino acids and vitamins from the environment because we can't produce them ourselves. And my robot does a little self-manufacturing: It snaps together components of a replacement leg.

Seth writes:

On the other hand, if a robot could indeed manufacture its own components from generic substrates under energetic constraints, then it might be truly said to be autopoietic, but it would also be considerably closer to a living system than any existing computer or robot.

So I think we should imagine a spectrum of biologicity, not a sharp divide between the biological and non-biological. If autopoiesis marks the biological, we can conceive of systems ranging from those that self-maintain in only the most limited sense, to those that do somewhat more, to those that self-construct at many levels at once, including their chemical components. AI systems might advance partway along this spectrum with no radical change in their underlying technology. But I'd agree with Seth that systems that manufacture even their own chemical components might merit the label "biological" -- perhaps as instances of artificial biology.

Maybe we can call systems in the middle of this range semi-biological. One might imagine semi-biologicity for other processes too: metabolism that somewhat but only somewhat resembles the kind highlighted by Godfrey-Smith, embodied goal integration of the kind highlighted by Cao and colleagues. Rather than thinking of the "biological" in a sharp-edged way, we might imagine ranges of artificial systems that have relatively more or less of various biology-like properties.

Of course, none of this addresses the second premise. Why should autopoiesis, or any other such biology-like property, be essential to consciousness? If we're inclined to reject near-term AI consciousness, requiring at least mid-grade autopoiesis or biologicity might yield intuitively the right results. But a rigorous defense of that requirement is still lacking.

ETA, Oct 3: Godfrey-Smith replies on X that his aim in his 2016 paper was more like "here is what it is, in us and other animals, that might enable us to get a grip on it". So it would be an interpretative reach to rely on that paper for a version of Premise 2 above, but of course one might also be inspired to consider whether Premise 2 would be true. He also points to a more recent paper of his that emphasizes the importance of large scale dynamical patterns in nervous systems (which would be difficult, but perhaps not impossible, to implement on current hardware).

[A Mark Tobey painting; image source]