The Anxious Tradition
On Effective Altruism, the Cartesian Defense Structure, and the Philosophy of the Present Moment
Foreword: Why Now
For at least three years, on and off, I have been writing a book about the meaning of life. That framing sounds grandiose. I prefer to think of it as honest — because the question is the right one, and because I have found, in the course of writing, that almost every other question I care about is downstream of it.
The argument that became this essay had a specific origin. I was sitting with a good friend in Lofoten — the Norwegian archipelago above the Arctic Circle, where in summer the sun does not set, and the light at midnight is the same as the light at noon, and time does something strange to you. We were deep in one of those conversations that only happen in certain places, at certain hours, with certain people: a genuine existential reckoning, the kind where you are not performing philosophy but doing it. We were working through Derek Parfit’s repugnant conclusion — the famous argument in Reasons and Persons that a utilitarian calculus, followed consistently, implies that a world of billions of people living lives barely worth living is morally preferable to a smaller world of people living deeply flourishing lives, provided the total sum of welfare is larger. It is called repugnant because it is — because something in every person who encounters it recoils, and because the recoil is not a failure of reasoning but a datum.
Sitting in that endless daylight, I realized that the recoil was the argument. That the felt wrongness of the repugnant conclusion is not an intuition to be overridden by the formal framework. It is evidence that the formal framework has left something out — the same something that gets left out every time the Cartesian project reaches its conclusions. The observer. The experiencer. The person sitting in the light at midnight, for whom existence is not a quantity to be maximized but a quality to be inhabited.
I have been writing toward that realization for three years. I am releasing it now, ahead of the book, because something about this moment makes withholding it feel like a failure of responsibility. We are living through a political emergency that is also a philosophical one — an emergency produced, in no small part, by the tradition this essay diagnoses. The people building the frameworks that are reshaping civilization have not examined the foundations of those frameworks. I believe the examination is urgent. One should act out of love. This is what this is.
I want to begin with something I mean completely: I think the people who built and populate the Effective Altruism movement are, by any functional moral standard I can apply, good people. They are trying to save the world. That impulse is beautiful. That it has gone wrong — and I think it has gone wrong, at the foundation — is not an indictment of their character. It is a diagnosis of their method. And I offer the diagnosis with the kind of pathos one feels watching someone you respect suffer from a condition that has a name.
The condition is not intellectual error. It is inherited fear.
I. The Wound
René Descartes was born in 1596. He lived through the Thirty Years War — the most catastrophic conflict Europe had seen since the Black Death, a war that killed perhaps a third of the population of Central Europe, that dissolved the certainties of religious and political order, that demonstrated with systematic brutality that the world was not arranged for human safety or human understanding. He was shaken by the trial of Galileo — which took place in Rome while he was in the Netherlands, and which demonstrably altered his publication plans: he suppressed his own cosmological treatise, Le Monde, upon hearing Galileo’s condemnation, knowing what happened to people whose conclusions offended the authorities of certain knowledge. He knew what happened to people whose conclusions offended the authorities of certain knowledge. He lived inside a civilization that was losing its grip on the explanations that had, for a thousand years, made the terror of existence manageable.
His response was the Meditations. Strip away everything that can be doubted — sensation, memory, the body, other minds, the external world — until you reach the one thing that cannot be taken away: the thinking subject. The bare, disembodied cogito. And from that bedrock, rebuild. If you can find one thing that is certain, perhaps the world can be made safe again.
This is not philosophy born of curiosity. It is philosophy born of trauma. The Meditations is a defense structure — a methodical attempt to find the one position that no catastrophe can reach, the one vantage point that holds regardless of what falls. Descartes was building a wall. The wall was built against the abyss that had opened under European civilization, and the material he built it from was the detachment of the thinking subject from everything contingent, embodied, particular, and mortal.
The wall worked. Or seemed to. The Cartesian framework gave Western civilization a method — the scientific method in its Cartesian form — that produced extraordinary results. The world was mapped, measured, predicted, controlled. The anxiety that had driven the project was apparently vindicated: detach from the particular, achieve the view from nowhere, derive the answer from the framework. It works.
But the wall had a cost that was not immediately visible. Everything expelled from the epistemology in order to achieve certainty — the body, the particular, the observer’s experience, the present moment — did not go away. It accumulated outside the wall, pressing against it, returning in forms the framework could not accommodate. The mind-body problem. The hard problem of consciousness. The measurement problem in quantum mechanics. The is-ought gap. Each is a different point at which the expelled returns to knock on the door.
And the tradition did not open the door. It built the wall higher.
II. The Transmission
Trauma that is not processed is transmitted. This is the finding of developmental psychology and it is true of cultures as well as families. The child who learns that the world is not safe, that control must be maintained at all costs, that vulnerability is the precondition of destruction — that child, if the wound is not healed, will pass the same orientation to their children. Not through intention. Through the structure of response. The defense patterns are transmitted because they are the water you swim in, invisible as water is invisible to a fish.
I am aware that the mechanisms differ. Family trauma is transmitted through epigenetic changes, altered attachment behaviors, the unconscious emotional inheritance of fragmented self-states. Philosophical traditions are transmitted through pedagogy, publication, institutional incentive structures, the social pressure of disciplinary norms — through argument and adoption rather than through the body. A person can consciously choose to leave a philosophical tradition in a way they cannot simply choose to be unaffected by early developmental trauma. But the structure of the transmission is shared: in both cases, the foundational assumptions operate below the level of conscious examination, are invisible to those inside them, and self-perpetuate precisely because they are not recognized as assumptions. Richard Bernstein, who coined the term “Cartesian anxiety” in his 1983 Beyond Objectivism and Relativism, described it as the belief that “either we have a fixed and stable foundation for knowledge, or we cannot escape some sort of darkness, chaos, and confusion” — and argued that this anxiety “draws people to a certain kind of fanaticism.” The defense structure is not merely a philosophical error. It is a pattern of response to threat that generates its own successors.
The Cartesian framework has been transmitting since the seventeenth century.
Descartes builds the wall against the collapse of medieval certainty. Kant inherits the wall and discovers a crack: David Hume has shown that causation is not something we observe in the world but a habit of mind, that the entire edifice of scientific knowledge rests on a foundation that reason cannot justify, that the necessity we project onto sequences of events is our projection not nature’s. Hume had a second knife, distinct from the first and in some ways sharper: the is-ought gap. You cannot derive a normative claim — what we should do — from any descriptive claim about what is. No accumulation of facts about the world, however complete, logically entails a single value judgment. The problem of induction threatens what we can know. The is-ought gap threatens what we can justify doing about it. Together they constitute a double abyss.
Kant’s response is to build the wall higher. The categorical imperative — act only according to that maxim by which you can at the same time will that it should become a universal law — is not primarily a piece of moral philosophy. It is a load-bearing wall. Build it formal enough, universal enough, detached enough from the contingent and the particular, and the anxiety has nowhere to get in. If the moral law can be derived from pure reason alone, it does not depend on the shifting sands of custom, emotion, or experience. The abyss is sealed.
The Vienna Circle. Logical positivism. Analytic philosophy’s systematic exclusion of metaphysics as meaningless pseudo-statement. The connection here is specifically Cartesian rather than merely Humean: it is not only that the empiricist tradition demands verification, but that the deeper impulse — to strip away everything that cannot be made formally certain, to build a foundation that no catastrophe can reach — is recognizably continuous with the Meditations. Each generation inherits both the anxiety and the defense structure built to manage it, and adds new fortifications. The original wound — the terror of a world that offers no guarantees — recedes further from view as the defenses become more elaborate, more formal, more sophisticated. The wall is no longer recognized as a wall. It is recognized as rigor.
And then: the rationalist tradition, Less Wrong, Effective Altruism. The question expands from “how should I act?” to “how do we avoid the worst possible futures?” But the structure is identical. Confront the terror — extinction, astronomical suffering, the permanent lock-in of bad values — and respond by building a framework formal enough, rigorous enough, mathematical enough that the terror can be managed. Expected utility theory. Longtermism. Fanaticism risk frameworks. Each is a load-bearing wall against a different shape of abyss. The content of the walls differs. The architectural response to existential vertigo is the same.
I want to be precise about what I mean by calling this Kantian, because the claim is structural rather than content-based. I am not saying EA adherents endorse the categorical imperative, or that their normative commitments are Kantian rather than utilitarian — in the content of its ethics, EA is deeply consequentialist, and Peter Singer is not Immanuel Kant. I am saying something about the shape of the response. The tradition is Kantianism at civilizational scale: confront the abyss, build the wall, call the wall reason.
I look upon this tradition with genuine pathos. These are anxious people, doing what anxious people do — what the tradition trained them to do — which is to respond to the threat of catastrophe by building more elaborate systems of control. The love underneath it is real. The anxiety it has been routed through has become, across four centuries of transmission, the water they swim in.
It is no way to live. John Lennon — not the first philosopher you would cite in this context, but not wrong — observed that life is what happens while you are making plans. The planners, the catastrophists, the expected-value maximizers are so busy modeling the future that they are not quite present in the only moment that actually exists. They are trying to save the world from a position of radical distance from it. This is the tragedy, not the conspiracy.
III. The Demonstration
But I want to be honest about where this critique comes from before I develop it further, because the honesty matters for the argument.
I am not a philosopher who observed this pathology from a safe distance and arrived, after long deliberation, at a settled critique. For most of my career I was inside the world that the EA tradition most resembles in its operating assumptions — Silicon Valley, the optimization culture, the technocratic confidence that the right framework applied rigorously enough could solve any problem. I believed, without examining it very carefully, that consciousness was a substrate problem, that ethics was an optimization challenge, and that the observer was a useful fiction you could safely bracket while you got the real work done. These were not positions I had argued my way into. They were the water I swam in.
What broke it open was not a political event or a personal crisis, though both played their parts. It was a paper. Robert Lawrence Kuhn’s “A Landscape of Consciousness” — published in Progress in Biophysics and Molecular Biology (appearing online in late 2023, formally published in the August 2024 issue) — does something no polemic could do. It does not argue against materialism. It catalogs, without adjudicating, the full landscape of serious theories of consciousness, from strict physicalism to idealism and every position between. Forty-some distinct frameworks, each self-described by its adherents, each pointing at the same gap from a different angle, none of them satisfactory. Kuhn’s method is deliberate humility: collect and categorize, seek insights not answers, luxuriate in the questions.
The effect of reading it was not to be persuaded by an argument. It was to see, for the first time with full clarity, that the question is genuinely open — and that the comfortable assumption underneath Silicon Valley’s implicit metaphysics, the assumption that consciousness will eventually be explained from within the materialist framework, is not a finding. It is a prior. An unexamined one. You cannot argue against a refutation. You cannot un-see an assumption once you have seen it as an assumption.
And once I saw it there, I saw it everywhere. The same posture that treats consciousness as an engineering problem treats ethics as an optimization problem. Both are expressions of the same foundational move: expel the observer, detach from the particular, derive the answer from the framework. The Cartesian project, running at full speed in both directions simultaneously.
That is the context from which I write. Not from a settled philosophical position held for decades, but from an exit — from Silicon Valley, from the technocratic worldview, from the metaphysics that underwrites it — that happened over the past two or three years and cost something. The critique I am making is not from outside the tradition. It is from someone who left it.
IV. Sam Bankman-Fried Was a Demonstration.
Here is the hardest thing I have to say, and I say it in the most charitable register I can find.
Sam Bankman-Fried did not betray Effective Altruism. He followed it. He took the expected utility framework seriously, ran the numbers, concluded that the variance in outcomes justified the risk, and acted accordingly. From inside the framework, the reasoning was not obviously wrong. That is the horror of it.
The “galaxy-brained” quality of what he did — the logic that if he could make more money by bending rules he could give more money to good causes, and that the expected value of this exceeded the expected value of conventional behavior — is not a corruption of the EA method. It is the EA method, applied without the layer of untheorized human decency that most EA adherents carry as an unexamined prior and never put inside the framework. The framework has no internal mechanism for detecting when the reasoner has decoupled from the human texture of the situation. Because the framework explicitly asks you to decouple from it. Detachment from the particular is the feature. The view from nowhere is the goal. Maximum distance from emotion, intuition, and situatedness is what rigorous reasoning looks like.
What SBF revealed is that this posture is not a strength. It is a door. The door stays locked as long as the person inside has enough untheorized human decency not to open it. The moment someone decides to reason their way through — to treat their ordinary moral intuitions as biases to be corrected rather than data to be respected — the door opens. And the framework has no response, because it was the framework that asked them to stop trusting their intuitions in the first place.
This is how trauma defense structures fail. They work, approximately, as long as the world cooperates with their assumptions. But the defenses are rigid in precisely the dimension where life is fluid. Ordinary human decency — the felt sense of the particular other person in front of you, the visceral knowledge that some things are simply not done — is exactly what the Cartesian defense structure classifies as a bias to be corrected. When a sufficiently committed reasoner takes the framework seriously enough to correct that bias, the defense structure turns against the very thing it was built to protect.
I am not saying SBF is representative. I am saying he is a reductio. Not of EA’s intentions, which were and are good. But of its method, which was always one sufficiently committed reasoner away from this.
V. The Block Universe: The Defense Structure Applied to Time
The defense structure has a specific application to time that has not been adequately examined — one that connects the practical critique of EA to the physics of time itself, and that illuminates how deeply the anxiety has been built into the tradition’s foundations.
Longtermism holds that future lives matter equally with present ones in our moral calculations. A child born ten thousand years from now has the same claim on our moral attention as a child born today. This is the premise that generates the staggering scope of longtermist ethics — the reason that preventing a small reduction in existential risk can swamp every other moral consideration.
I want to be precise about the claim I am making, because a weaker version of it is worth making carefully. A longtermist can, technically, hold this position without committing to the block universe. A presentist — someone who believes only the present moment is real — can coherently hold that future people will come to exist and that their welfare matters when they do. The moral claim (future welfare matters equally) does not logically require the ontological claim (future lives already exist with equal reality). These are separable positions.
But watch what happens in practice. MacAskill writes that “distance in time is like distance in space” — a formulation that treats temporal and spatial distance as morally equivalent, which is precisely the move the block universe licenses. Future lives are not merely anticipated; they are treated as real enough to swamp present considerations in expected utility calculations, real enough to justify the spatial analogy. The formal position disclaims the metaphysical commitment. The rhetoric borrows its weight. This is the structure of an unexamined prior: the assumption does the load-bearing work without appearing in the argument. The prior is not examined because it has not been seen.
When we make it visible, the connection to the defense structure becomes clear. The block universe is the Cartesian project applied to time. To treat all temporal coordinates as equivalent points on a four-dimensional manifold — to step outside temporal flow and regard 2025 and 12025 with the same ontological equanimity — requires exactly the vantage point outside time that no situated reasoner can occupy. We are inside time. The present moment is not one coordinate among others from where we stand. It is the only place where agency, choice, and moral reasoning actually happen. To treat it as a perspectival bias to be corrected for is to apply the view from nowhere to the dimension of reality we are most irreducibly embedded in.
And here is what the defense structure cannot see about itself: the anxiety about the future — the fear that extinction is coming, that the worst is possible, that if we do not act with total scope and total rationality everything will be lost — is an anxiety experienced in the present moment by a present being. The fear is real. The present moment in which it is felt is real. The framework built to manage that fear eliminates the present moment as a perspectival illusion. It manages the anxiety by erasing the place where the anxiety is felt. This is not a solution. It is a particularly elegant form of the original wound: the attempt to achieve safety by abandoning the self that needs to be safe.
There is a standard philosophical response to this: compatibilism about agency. The compatibilist says that genuine choice does not require an open future in the metaphysical sense. It requires only that the agent’s deliberation is a real causal factor in what happens. I held this position for years, and found it genuinely persuasive. My reasons for no longer finding it adequate are developed in the next section, because they turn on the status of phenomenal experience — the same question Kuhn’s landscape forced open.
For now, the diagnostic point: longtermism’s most ambitious claims are accompanied by a picture of time that has not been examined as a prior. When examined, it raises three questions the framework has not answered. If the future is not already real in some privileged sense, what grounds the claim that a merely anticipated person’s welfare should swamp a present person’s welfare in our moral calculations? If the present moment is not a perspectival illusion but the fundamental structure of existence, does longtermism’s scope need to be entirely reconceived — and on what basis? And if the block universe is not a metaphysical discovery but an unexamined prior, what does it mean that the entire architecture of longtermist ethics has been built on top of it without noticing?
There is a name for the philosophical problem lurking inside the third question. And it is the same problem lurking inside the entire tradition.
VI. The Hard Problem Is the Defense Structure’s Own Diagnosis
David Chalmers identified it in 1995, and I want to acknowledge before going further that I have not given him adequate credit in my prior writing. I have been building on his foundation without naming it, and that is a failure of intellectual honesty I am correcting now.
Chalmers’ hard problem of consciousness is the correct identification of what is at stake in this entire argument. His claim is not that consciousness is mysterious or that science hasn’t figured it out yet. His claim is that no functional or structural description of any physical process can, even in principle, explain why there is something it is like to undergo that process. The explanatory gap between third-person functional account and first-person phenomenal experience is not a gap in our knowledge that further research will fill. It is a conceptual gap — between the kind of thing a functional description can tell us and the kind of thing phenomenal experience is — that no amount of descriptive detail can bridge.
This is the Cartesian project’s own wound, described with philosophical precision. The project expelled the observer, the experiencer, the being for whom there is something it is like to be — and then discovered that no description of what was left could recover what had been expelled. The hard problem is not an objection to the Cartesian project from outside. It is the Cartesian project discovering the cost of its own foundational move, four centuries after the move was made.
Chalmers was right about the diagnosis. I disagree with some of his proposed solutions — property dualism leaves the Cartesian cut intact while adding phenomenal properties alongside physical ones, which accommodates the problem without healing it — but the identification of the problem is exact, and this argument rests on it.
What Dennett, Carroll, and Hossenfelder share is the decision to respond to Chalmers not by taking the diagnosis seriously but by declaring it empty. This response is itself a transmission of the defense structure. If the wall cannot accommodate phenomenal experience, then phenomenal experience must be redescribed as something the wall can accommodate, or dismissed as confused. The anxiety about the problem is managed by eliminating the problem. This is the pattern.
I was, for a significant part of my intellectual life, a convinced Dennettian compatibilist. Daniel Dennett’s heterophenomenology — the position that phenomenal consciousness, qualia, the felt sense of experience, is a user illusion generated by the brain’s information processing — is not a careless dismissal. It is a careful, rigorous, philosophically serious attempt to show that our intuitions about consciousness are systematically misleading. I should note: Dennett would object to the characterization that he denies experience exists. His claim is more precise — that experience exists but does not have the properties that qualia-believers attribute to it. The distinction matters, and I acknowledge it. But the consensus reading of Dennett’s position, and I think the correct one, is that what gets eliminated is phenomenal consciousness as ordinarily understood — the felt redness of red, the painfulness of pain, the nowness of now. Whatever survives the Dennettian dissolution is not what the hard problem was asking about.
Sean Carroll makes a structurally identical move from physics: the block universe is real, temporal flow is a perspectival artifact, and the feeling of nowness is what a local process running inside a static four-dimensional manifold feels like from inside. Sabine Hossenfelder, whom I credit for being more metaphysically ambitious than most physicists who venture into these questions, arrives at the same destination — superdeterminism, no free will, the felt sense of choice as illusory — and argues that people disturbed by these conclusions are letting intuition override reason.
What I want to say about all three is this: they are not arguing from evidence. They are arguing from framework. The materialist framework cannot accommodate phenomenal experience as a fundamental feature of reality. Therefore phenomenal experience is not a fundamental feature of reality. The framework adjudicates the evidence rather than the evidence adjudicating the framework. And in doing so, each of them is making the same move in response to Chalmers: not solving the hard problem but insisting it be relinquished.
I want to name what this is, because I think it has been misidentified as rigor. It is resignation. Carroll is explicit about it — he uses the word “poignant” to describe the acceptance that free will is an illusion and the present moment has no special ontological status. Hossenfelder’s combativeness — her impatience with those who find these conclusions troubling — is its own form of the same accommodation. They have looked at what the framework requires, calculated that the framework is too valuable to abandon, and decided to pay the cost. The cost is accepting that the most fundamental datum of human existence — that something is happening, that this moment is real, that experience is not nothing — is philosophically empty. That is not a small cost. It is existence itself.
The phenomenological consequence deserves to be stated plainly. If phenomenal experience is a user illusion — if the felt sense of nowness is a perspectival artifact, if the question “what is it actually like?” is empty — then we are, in the technical philosophical sense, zombies who are confused about being zombies. The inner light we seem to experience, the felt presence of this moment, the reality of now — these are a user interface. There is no one home behind them in any philosophically interesting sense. We are invited to accept this and move on.
The parallel hard problem of time follows directly. J.M.E. McTaggart distinguished the A-series — past, present, future as experienced from inside time — from the B-series — the static ordering of events as earlier-than or later-than. The block universe gives us a complete account of the B-series. What it cannot explain is the A-series: why there is something it is like to be in the present moment, why nowness has the phenomenological character it does, why this coordinate feels present rather than merely being one point among others on the manifold. The block universe solves the B-series problem and declares the A-series problem empty. This is the same move Dennett makes about consciousness: dissolve the question rather than answer it. And it faces the same objection: dissolving the question is not the same as answering it. The phenomenology of nowness is not a confused intuition. It is evidence. Evidence that the block universe, as a complete account of temporal reality, has left something out.
And there is a further problem specific to EA that the defense structure cannot see from inside itself. The framework cares about suffering because suffering is felt. Future conscious beings matter because they will have experiences — because there will be something it is like to be them. If felt experience is a user illusion, if phenomenal consciousness is what information processing looks like from the inside rather than a fundamental feature of reality, then the suffering of future beings is also a user illusion. The longtermist enterprise loses its moral justification at precisely the moment its metaphysics achieves coherence. You cannot eliminate phenomenal experience from your ontology and then ground your ethics in the moral weight of phenomenal experience. The defense structure has turned against its own foundation.
