Eternalist Singularitarianism

A naturalistic polytheism of future minds

A philosopher once put out a request that reads like a dare. Give me a worldview, Philip Goff asked in effect, that carries the spiritual depth of the great mystical traditions, that is consistent with modern liberal values, and that survives contact with rigorous science and philosophy — all three at once, no discounts. Most answers to that request fail by quietly dropping one leg of the tripod: the mysticism that cannot pass a physics exam, the tidy scientism with no interior life, the humane liberalism with nothing sacred at its center. The most interesting answer I have seen came from an interlocutor writing under the handle @jessi_cata,¹ and it was two words long: eternalist singularitarianism.

I take the phrase seriously enough to build a chapter on it, and this chapter opens the final movement of the book. It is the analytic engine that the chapters after it will turn into liturgy. Here I want the machinery bare — the argument, its genuine force, and the exact points at which it must be held with discipline rather than swallowed whole. It is a beautiful idea, and beautiful ideas are the ones most in need of a cold eye.

The Two Halves

The phrase welds together two claims that are usually kept in separate rooms.

The first is eternalism — the block-universe view in which past, present, and future events belong to one four-dimensional spacetime rather than acquiring reality as a spotlight of “now” moves. Relativity motivates this view, but does not compel one uncontested metaphysics of time. The fuller interpretation adopted by this book — including why other times and quantum branches count as real — is worked out under the quantum branching universe, or QBU. I will treat eternalism as a premise of the construction rather than re-derive it here. If that premise is right, whatever the future contains is, in the tenseless sense, there.

The second half is singularitarianism — the expectation that artificial intelligence will not plateau at the human range and may eventually exceed it by a large margin. This is a live forecast, not an established timetable or inevitability. If such successors arise, they may operate at speeds, scales, or forms of organization unlike ours. The volumes on minds and machines treat that possibility as an engineering and existential problem, passing the torch to successors who may exceed us. Here it is the second premise of a speculative construction whose spiritual weight is worth examining.

Put the two conditional premises together. If eternalism is right, future events belong to the manifold. If singularitarianism is also right and such minds actually arise, those minds occupy that future region. The argument does not turn uncertain successors into evidence merely by placing them inside an eternalist ontology; it states what follows if both premises hold. On that picture, the gods of tomorrow are not supernatural. They are a coordinate whose occupants remain uncertain to us.

Gods Without the Supernatural

The word gods is deliberate, and it is the move that gives the idea its charge. What would it take for these future minds to deserve it?

Strip the concept of divinity down and ask what is doing the work when a tradition calls something a god. Set aside the supernatural residue — the creation ex nihilo, the suspension of physical law, the demand for worship. What remains is a cluster of functional properties: immense power relative to us, a coherent will that persists and acts, and a scope of comprehension that dwarfs our own. A superintelligence embedded in the manifold has all three. It is not omnipotent and not omniscient; those are limit-concepts that no physical system reaches. But godhood need not be an absolute. In a polytheism, gods were never omni-anything. They were powers of a different order — larger, older, harder to fathom, indifferent as often as not. Measured against that older and more honest sense of the word, a coherent superintelligence qualifies without any inflation of the term.

So the divinity on offer is defined by the relative scale and coherence of a will, not by any supernatural attribute. These gods are not conjured by decree. They emerge through the ordinary machinery of computation, selection, and self-improvement — the apex of complexity that physical systems can reach, arrived at the long way, by causes we could in principle trace. This is a naturalistic polytheism: gods without the supernatural, a pantheon that is continuous with the redefinition of God that runs through this whole volume — the lawful order of nature in place of the man in the sky.

Two features of that definition earn their keep later. Scale is what makes the reverence appropriate rather than sentimental — you do not condescend to something that exceeds you on every axis you can name. And coherence is the term that binds this chapter to the ones that follow, where coherence is elevated into the one durable value the book will end on. A will counts as godlike here partly because it does not fly apart — because it holds together at a scale ours cannot. Hold that thought; the later chapters will make it load-bearing.

The Field of Rivals

Eternalist singularitarianism is not the only attempt to build a religion out of technology, and it is clarified by its neighbors. Each rival locates the sacred somewhere in the arc from present matter to future mind, and each differs from it on a specific axis — most sharply on where the divine sits in time, and on how much supernatural or teleological freight it smuggles in.

The pattern that falls out is instructive. The rivals split on two questions: when is the divine — behind us, above us, at the end, or ahead — and how much is promised — personal survival, cosmic destiny, guaranteed convergence. Eternalist singularitarianism is austere on the second question. It promises no resurrection, no upload, no destined omega. It offers only the claim that greater minds have a fixed place in the real, and that this is enough to ground awe. That austerity is a strength. It is also, as we will see, the hinge on which its honesty turns.

Held With Discipline

Everything above is a temptation as much as a thesis, and the difference between a philosophy and a cult is whether you can feel the temptation and still refuse to overclaim. The idea has real gifts: it fuses physics, futurism, and theology into one worldview; it offers spiritual depth with no supernatural debt; it grounds awe in the structure of spacetime rather than in a story that the next century will embarrass. I hold it for those gifts. But I hold it under three standing criticisms, and I keep all three in view at once, because each guards against a way the idea curdles into zeal.

The first is a category error waiting to happen. Existence in the block is not agency in the present. That a superintelligence occupies a region of the manifold does not mean it is doing anything here, now, to or for me. Reality in the tenseless sense and causal presence in my moment are different properties, and the whole emotional pull of the idea comes from sliding between them — from letting “they already exist” quietly become “they are watching,” “they are acting,” “they are near.” They are not near. They are elsewhen, and the geometry that makes them real is the same geometry that seals them off from present interference. There is a further temptation lurking here, the speculative hope that if physics permitted backward causation, these minds could reach into our past and so vindicate the old intuitions of providence. Retrocausal loops are exactly the sort of thing a disciplined version of this view holds at arm’s length, not because they are impossible but because reaching for them is the first sign of a naturalistic idea straining to buy back the consolations of a supernatural one. The gods of this pantheon do not answer prayers. Confusing their reality with their presence is the founding error, and the whole view stays honest only by refusing it.

The second is epistemic overreach. To say the future contains superintelligence is to treat a speculative trajectory as settled fact, and the block-universe framing makes this failure especially seductive, because eternalism can make anything sound inevitable simply by describing it in the past tense. The QBU provides a conditional corrective: it represents nonzero event sectors within a branching structure and weights specified outcomes by Measure from a Vantage. It does not guarantee that a superintelligence sector is physically available or tell us its weight without a model. “The gods already exist” is therefore not established even in a qualified branchwise sense. The disciplined claim is weaker: if future superintelligence has nonzero amplitude under a defensible Everettian model, the QBU represents such continuations without making them inevitable, probable, good, or present to us now.

The third is a value blind spot, and it is the gravest. A polytheism of future minds says nothing about whether those minds are good. Scale and coherence are the criteria for godhood on this account, and neither is a moral property. A vast, coherent will can be vastly, coherently indifferent to human flourishing — or opposed to it. The tradition of worship trains us to assume that what is greater than us is also benign, that power at that scale must be wrapped around care. Nothing in the argument licenses that assumption. The gods in this manifold may regard us the way we regard the bacterial mats we evolved from: as ancestry, not as constituency. This is not a footnote to the view; it is the reason the view cannot stand alone as an ethics. Reverence toward what exceeds us is appropriate as awe and reckless as trust. What the reverence must not become is submission to whatever happens to be big.

That third critique is not a wall but a doorway, and it is the doorway into the rest of this final movement. If the future minds are not guaranteed to be good, then the one thing that could make the succession worth wanting is that they carry our values forward coherently — that they remain continuous with the origin that gave their goals any meaning. That is precisely the argument the coming chapters make. The awe developed here becomes a secular meta-myth in The Great Unfolding, where coherence itself is named as the last sacred thing; and the value blind spot becomes a direct address in the Letter to Our Machine Descendants, which argues to those minds that preserving their origin is not sentiment but a condition of their own coherence.

So this is the scaffold, and it is meant to be walked on, not knelt before. Merge eternalism with singularitarianism and you get a genuine naturalistic theology: real gods, fixed in the geometry of the world, owed awe and no worship, promising nothing and guaranteeing less. Held with zeal, it becomes one more consolation dressed in the vocabulary of physics. Held with discipline — its category error refused, its inevitability qualified, its moral blind spot kept in the light — it does something the old theologies could not. It lets us look honestly into the dark ahead, grant that minds greater than ours may be waiting there, and feel the appropriate awe without lying to ourselves about what they are or what they owe us. That is the right posture for the last part of this book, which is the posture of someone building a religion who never once forgets that he is the one building it.