Alasdair MacIntyre is persuasive when he argues that moral discourse is never neutral, and that modern liberalism smuggles in substantive standards while pretending not to. But he dismisses emotivism too quickly as a cultural disaster rather than considering whether it might describe moral language more accurately than his own teleological alternative. If moral utterance is fundamentally prescriptive or expressive rather than descriptive, then the collapse of ‘view from nowhere’ morality doesn’t send us scurrying back to Aristotle. It simply shows that moral language was never doing the metaphysical work MacIntyre wants from it.
The Aristotelian remedy also depends on a nostalgic and anachronistic social model. The Athens he implicitly romanticises was a small polis whose demos consisted of citizens, meaning property-owning males, already bound by shared norms, proximity, and cultural inheritance. In other words, the sort of thick local world that made a certain kind of practical ethical life possible in the first place. MacIntyre’s causal arrow points the wrong way. In Athens, democratic practice emerged from that prior social texture. You do not reproduce the same conditions by philosophical edict.
Audio: NotebookLM summary podcast of this topic.
To put it more bluntly: I don’t think moral realism is tenable, and I am not convinced MacIntyre really thinks so either. His project reads less like a discovery of moral facts than an attempt to promote an ought into an is by force of inheritance and rhetorical confidence. If he carved out a bounded cohort and imposed the right shared practices, perhaps something like his model could function. He may need to annex a reasonably sized car park for the purpose.
There’s a certain kind of cultural panic that tells you more about the panickers than about the thing they are panicking about. The current hysteria over AI-inflected prose is a good example.
The argument, insofar as it deserves the name, goes roughly like this: LLMs produce prose with identifiable features – a certain blandness, a fondness for the em dash, a tendency toward tidy three-part structure. Writers who use these tools risk absorbing those features. The authentic human voice is therefore under threat. Something precious is being diluted by contact with the machine.
This is sentimental rubbish, and it is worth saying so clearly before doing anything else – and a sort of virtue signalling.
Audio: NotebookLM summary podcast of this topic.
I use LLMs daily. For research, for editorial pushback, for smoothing passages that have gone awry. This means I spend hours a day reading a particular kind of output. You’d have to be delusional not to admit it has effects. Certain phrasings start feeling natural that didn’t before. Certain rhythms begin to recur. Certain words might not have otherwise come into use. I notice this and note it without particular alarm, because I’ve read enough to know that this is just what environments do.
Read nothing but McCarthy for a month, and your sentences will start hunting for the spare declarative. Spend a year editing academic philosophy, and you will catch yourself reaching for ‘insofar as’ and ‘it’s worth noting’ in casual conversation. Live in a city long enough, and its cadences work their way into your syntax. This isn’t contamination, the negative moralist dispersion. It’s how language acquisition works for as long as one is alive and reading. Voice isn’t a spring. It’s a river, a moving accumulation of every tributary it has passed through.
The prestige game being played by the anti-LLM faction isn’t difficult to spot. When Dostoyevsky shapes a young writer’s cadence, we call it influence and treat it as evidence of a serious literary education. When a game world shapes a child’s imagination – I homeschooled my son in the manner of unschooling, and his primary corpus for years was World of Warcraft and its attendant lore before shifting to Dark Souls – and that child ends up reading Dante and Milton unprompted in year seven, the same mechanism has clearly operated. The source was not canonical, the outcome was. But the respectable hierarchy of influences cannot easily accommodate this, because the hierarchy was never really about the mechanism. It was about the cultural status of the inputs.
The more interesting observation isn’t about those of us who use these tools. It’s about those who conspicuously do not.
A minor genre has emerged – charitably, I’ll call it a genre because cult feels morally loaded – consisting of writers anxiously purging their prose of anything that might read as AI-generated. It’s worth noting that they have read the lists. Telltale signs of LLM authorship: excessive hedging, em dashes, transitional summaries, the phrase ‘it is worth noting’. And so they scrub, redact, replace, and perform a kind of stylistic hygiene that’s a creative decision made in direct response to LLM discourse.
These writers aren’t free of the machine’s influence. They’re among the most thoroughly shaped by it. They simply have the more theatrical relationship – the counter-imitator, the purity-performer, the one who reorganises their entire aesthetic in orbit around the thing they claim to reject.
Thomas Moore, in Care of the Soul, observes that a child raised by an alcoholic parent tends to become either an alcoholic or a committed teetotaller. He presents this as a dichotomy, which is too neat, but the underlying point holds. Reactions are still relata – see what happens when you read too much philosophy and logic? The teetotaller has organised their life around the bottle as surely as the alcoholic has. Both are defined by it.
Opposition is one of influence’s favourite disguises.
The fair objection is that LLM influence may differ from other influences in kind rather than just in kind. Dostoyevsky is strange. Bernhard is strange to the point of pathology. A canonical prose style is idiosyncratic by definition, which is why it’s worth absorbing. In contrast, LLM output aims for plausible fluency and statistical centrality. Its pull may be more homogenising than the pull of a singular authorial sensibility.
That’s a real point. The environment in question has a centripetal force toward the mean that most literary influences lack.
But conceding the point doesn’t really rescue the panic. It just specifies the kind of influence involved. The mechanism remains identical to every other case of environmental absorption. And ‘this influence tends toward the generic’ is an ironically generic critique of a particular environment’s character rather than a claim that the environment is doing something ontologically unprecedented to the notion of authorship.
The question that actually matters aesthetically is not was this touched by AI? It is what did the writer do with the environment they inhabited? That’s always been the question. It remains the question. The machinery has changed; the problem of influence has not.
What the current schism actually reveals is not that AI is doing something new to writing. It’s that we’ve been operating with a fairy tale about what writing is. The fairy tale holds that voice is self-originating, that somewhere beneath the reading AND the editing AND the genre conventions AND the institutional pressures AND the decade of a particular editor’s feedback, there is a pristine you, unconditioned and pure, expressing itself directly onto the page.
This was always false. Writers have always been patchworks of absorbed environments. The only difference now is that one of the environments is a machine, and the machine is new enough that people haven’t yet learned to be comfortable with what it reveals about the rest.
The environment always wins. The only interesting question is which environments you choose, and what you make of them.
So glad I took time out to watch a short exchange between Rick Beato and Justin Hawkins on whether music is becoming content rather than art. The question is framed in musical terms, but it hardly stops there. The same corrosion is visible in writing, visual art, criticism, and now, with grim inevitability, in AI-mediated production more broadly. The disease is not confined to music. Music merely makes the symptoms easier to hear.
For music, my aversion to pop music goes back to my youth. I was a kid when the Beatles practically invented pop music, but they left it to grow and continued exploring. Sadly, as solo artists, they mainly – not always – failed and rested on their laurels in pop. It’s not that their version or any pop music is inherently unlistenable. Surely, it’s not, if only by the aspiration of the pop moniker, but it has no depth, no soul, as it were. Some make this argument for Organic food. In essence, it involves an appeal to nature fallacy.
Audio: Slightly off, but not bad, NotebookLM summary podcast of this topic.
My own aversion to much pop music begins there. It is not that pop is necessarily bad, nor even that it is always shallow. That would be too crude and too easy. The problem is that pop often presents itself less as an artistic act than as a consumption object engineered for immediate uptake: catchy, frictionless, emotionally legible, and just disposable enough to make room for the next one. It is built to circulate.
That, for me, is the difference between content and art. Art may be accessible, even popular, but it retains some residue that exceeds its delivery mechanism. It resists total reduction to utility. Content, by contrast, is made to be processed. It is optimised not for depth but for throughput. Its highest ambition is not transformation, but engagement.
This is why the question matters beyond music. Writing, too, now lives under the same pressure. One is increasingly expected to produce not essays, arguments, or works, but units of output: posts, threads, reactions, takes, summaries, explainers, and other forms of polished verbal debris. The point is no longer to say something worth dwelling on, but to remain visible within the churn.
The issue, then, is not simply whether one should consume AI-generated material. That framing is too pious and too easy. The more interesting question is what the consumer thinks they are consuming. If a reader, listener, or viewer wants only speed, familiarity, and surface competence, then AI content is not a scandal at all. It is the logical endpoint of a culture that has already demoted art into a deliverable.
This is where the fuss over labelling enters. Is it a principled demand for honesty, or merely a theatrical gesture by people who still want the aura of art whilst consuming content on industrial terms? Some of it is clearly protectionism. Some of it is virtue signalling. But not all of it is empty. The insistence on labelling betrays an intuition, however muddled, that authorship still matters, and that not all artefacts are equivalent merely because they occupy the same screen-space.
The deeper question is whether we still want art at all, or whether we merely want the aesthetic styling of art attached to things optimised for convenience. Once a culture learns to prefer seamless output over resistance, recognisability over risk, and quantity over form, it should not act surprised when machines begin to serve it perfectly. They are only completing a trajectory already chosen.
So no, the issue is not AI alone. AI is only the latest mirror held up to a public that has spent years confusing availability with value and polish with depth. The real question is not whether machines can make content. Plainly, they can. The question is whether we still possess the appetite, patience, and seriousness required for art.
Image: Full image because the cover version is truncated. Generated by Gemini Nano Banana.
My attention has yet again been abducted by fiction, tentatively titled The Box, but this time it is (ever so slightly) different. A publisher mate suggested that no one reads nonfiction, and anyway, fiction sticks better because it captures both attention and salience; I decided, why not both? I wrote some perhaps esoteric but non-academic nonfiction and decided to convey at least some of this through literary speculative fiction.
Besides the obvious nod to MEOW and The Architecture of Encounter, the novella explores several philosophical concerns:
The first three are diagnostic. The fourth is operative. The fifth haunts the whole enterprise. None are cited in the text – readers will encounter them through the fiction rather than the sources, which is rather the point.
I don’t want to spoil the plot, but I wanted to share a post today, and this is where I am: somewhere in the middle of a first draft, trying to make a speculative premise carry philosophical weight without the reader feeling the load. It’s harder than nonfiction. It’s also more fun.
Audio: NotebookLM summary podcast of the Substack topic.
It starts like this:
You’ve had the argument. Everyone has. You present evidence. Your interlocutor presents counter-evidence. You cite data. They cite different data – or the same data, read differently. Eventually, someone says something like how can you possibly believe that, and the conversation is effectively over, though the words might continue for another hour.
What’s left is the quiet conviction that the other person is either ignorant, stupid, or arguing in bad faith. Perhaps all three. And you can be certain they’re extending you precisely the same courtesy.
I want to suggest that something structurally different is going on – something that none of the usual explanations (media bubbles, declining education, algorithmic polarisation) quite reach. These explanations aren’t wrong, but they’re shallow. They describe accelerants. The thing they’re accelerating is more foundational.
Disclosure: I shouldn’t have to apologise for two consecutive posts written by Claude – previous post on duration and time – this time about reframing a chat about time and duration around the hydrozoan jellyfish (sic. sea jelly). Yet again, this post arrives with a warning: if you don’t like AI-authored content, it’s not too late to turn back or click away.
Audio: NotebookLM summary podcast of this topic.
A hydrozoan jellyfish off the coast of Japan has no brain, no central nervous system, and no opinion about the block universe. It does, however, have something to teach us about the nature of time — and it’s more than most pop-science accounts manage.
Researchers recently describedClytia sp. IZ-D, a previously undocumented species of jellyfish that releases its gametes each evening with remarkable precision. The interesting part isn’t the precision. It’s what produces it.
Left to its own devices under constant light, the jellyfish spawns every 20 hours. Not 24. Twenty. It has its own autonomous developmental rhythm — oocytes growing, acquiring competence, reaching a threshold, and releasing. That rhythm belongs to the organism. It’s not borrowed from the environment. It’s not calibrated to the rotation of the Earth. It’s the jellyfish’s own tempo.
But place the jellyfish under a normal day-night cycle, and the 20-hour rhythm gets pulled into a 24-hour one. The daily light-dark transition doesn’t trigger spawning directly. It delays the autonomous cycle by several hours, dragging it into alignment with the environmental period. Sunrise resets the clock — not by starting it, but by holding it back. Under a standard 12-hour light cycle at 21°C, spawning occurs 14 hours after dawn. Every day. Synchronised across the population to within about 20 minutes.
The 24-hour cycle is not the jellyfish’s rhythm. It’s what happens when an external structure is imposed on it.
Now, you could read this as a charming piece of marine biology and move on. But if you’re paying attention, it’s a remarkably clean illustration of something that applies well beyond cnidarian reproduction.
The jellyfish has its own duration — a developmental tempo with no fixed metric. Under constant conditions, the cycle length shifts with temperature: 17 hours at 24°C, 22 hours at 18°C. There’s no intrinsic interval. The process has structure — oocyte growth, staged competence acquisition, threshold sensitivity — but it has no grid lines. No bar lines. No BPM.
The 24-hour day provides the grid. The light-dark cycle imposes intervallic structure on a process that doesn’t natively possess it. And the imposition works — it synchronises the population, coordinates male and female gamete release, and presumably improves reproductive success. The grid is useful. Indispensable, even.
But it’s not the organism’s time. It’s what the organism’s time looks like after the environment has had its way with it.
And here’s the part that matters philosophically. The imposition is imperfect. Individual oocytes within the same gonad don’t all reach maturation competence at the same moment. There’s a spread — 40 to 60 minutes under normal conditions, widening further under constant light. The grid says “now.” The biology says “roughly now, give or take, depending on which oocyte you’re asking.” The grid cannot fully resolve what it organises. There’s always a residue.
Under constant darkness, the residue takes over entirely. The autonomous rhythm reasserts itself, but without the entraining signal it becomes asynchronous. Different jellyfish spawn at different times. Different oocytes within the same jellyfish mature at different rates. The developmental process is still there — the structure is still there — but without the intervallic imposition, it doesn’t produce anything that looks like coordinated “time.” It produces duration doing what duration does when nobody is counting.
None of this requires a brain. None of it requires experience. None of it requires a subject who feels the passage of time. The jellyfish doesn’t experience the 24-hour cycle as given. It doesn’t experience anything, so far as we know. And that’s precisely what makes it useful as an illustration: it shows intervallic structuring operating at a purely material level, without any phenomenological overlay. The grid is imposed on the biology. The biology is not the grid.
Now scale up.
We do the same thing the light-dark cycle does to the jellyfish, except we do it to everything, and we do it to ourselves. We impose intervallic structure on duration — segmenting it, ordering it, metrising it — and then we mistake the structure for what it’s imposed on. We experience time as sequenced, directional, and measured, and we assume that’s what time is. Physics formalises the assumption into equations and discovers, to no one’s surprise, that the formalism is time-symmetric. And then pop science announces that time could therefore “go backwards” — which is exactly like saying that because the score reads the same in both directions, the music could be played in reverse.
The jellyfish knows better. Not because it’s clever, but because it’s simple enough that you can see the layers separately: the autonomous developmental duration, the environmental intervallic imposition, the imperfect entrainment, and the residue the grid can’t capture. In us, those layers are collapsed. We live inside the imposition and mistake it for the terrain.
The 24-hour day is a grid. A useful grid. An indispensable grid. But the jellyfish was doing something before the grid arrived, and it will go on doing something if the grid is removed. What it does without the grid doesn’t look like “time.” It looks like biology unfolding at its own pace, in its own structure, answerable to its own constraints.
That’s duration. And it was there before we started counting.
Disclosure: This post was written by Claude after a chat about time and duration. As a musician emeritus, I suggested the connexion between free time, metred time, and the quantising of ProTools. This discussion occurred as I was talking through a more serious piece on duration and time as ontic versus phenomenological essences. Not wanting to divert my efforts too far, I asked Claude to summarise our chat for this blog post. If you don’t like AI-authored content, this is your opportunity to exit. As for the rest, Abandon All Hope Ye Who Enter Here.
Audio: NotebookLM summary podcast of this topic.
Musicians have been complaining about quantisation for decades. You record a performance — human, breathing, alive — and then the software snaps every note to the nearest grid point. The timing is “correct.” The feel is gone. Something was there and now it isn’t.
The standard explanation is that quantisation removes “feel” or “groove” — which is true but vague. What it actually removes is everything that falls between the intervals. The micro-deviations. The fractional push ahead of the beat, the slight drag behind it, the breath between phrases that doesn’t land on any grid division. These aren’t imprecisions. They’re the performance. Quantisation doesn’t tidy up the music; it overwrites it with a metrically perfect substitute and calls the result an improvement.
This is not a blog post about music production.
This is a blog post about time — and about the fact that the ProTools complaint is a precise, intuitive articulation of a philosophical problem that has been mishandled for over a century.
Here’s the short version. Duration — the actual temporal fabric of things — is something like free time in music. Not metrically structured, not divided into equal intervals, not indifferent to direction. Just: what’s happening, happening. Time, as we ordinarily understand it — clock time, measured time, the physicist’s time — is what you get when you impose intervallic structure on duration. Bar lines. A tempo marking. A grid.
The grid is useful. Indispensable, even. But it’s representational. It organises what’s there; it doesn’t discover what’s there. We no more find intervals in duration than we find bar lines in birdsong.
Now here’s where physics gets into trouble.
The equations of motion are time-symmetric. Run them forwards or backwards, they work either way. Pop science takes this and says: therefore time could, in principle, go backwards. Time travel! Reversed entropy! Christopher Nolan!
But this is exactly like saying: the score reads the same in both directions, therefore the music could be played backwards and remain the same music. It couldn’t. The score is the grid. The music is not the grid. The formal symmetry of the notation tells you something about the representational scheme. It tells you nothing about the direction of the performance.
Bergson saw half of this a century ago. He insisted that lived time — durée — is irreducible to the physicist’s spatialised coordinates. He was right. But he made his own version of the musician’s error: he confused free time with no time at all. Playing without a click track isn’t escaping metric structure. It’s still playing notes in sequence. Bergson thought philosophical intuition could recover duration as it actually is — but the feel of playing freely is still the feel of playing. It’s closer to the truth than the metronome is, but it’s not the truth itself.
Einstein, meanwhile, gave us the most sophisticated metronome ever built and then we all agreed the metronome was the music.
And ProTools? ProTools is what happens when you take the grid seriously enough to enforce it absolutely. Quantise everything. Snap every event to the nearest interval. And then notice — as every musician who has ever hit that button notices — that something real has been lost. Not a feeling. Not a subjective impression. Something that was in the sound is no longer in the sound.
That loss is the point. The grid cannot fully capture what it organises. There is always a residue — and the residue isn’t a flaw in the grid or a limitation of the technology. It’s a structural feature of any intervallic imposition on something continuous. The grid does real work. But the work it does is not the same as the thing it works on.
Time doesn’t go backwards. It doesn’t go forwards either. “Going” is what the grid gives you. Duration doesn’t go anywhere. It’s what’s already there before the grid arrives.
The musicians have known this all along. They just didn’t know it was philosophy.
I am considering a new essay. That’s nothing new, but this was born from personal experience. Whilst reading Derek Parfit’s Reasons and Persons, I reached the chapters on moral arithmetic and imperceptible harms and effects, and it caught my attention. Not in the ‘Aha!’ way, but because I felt excluded given my own experience. My mind wandered off the reservation, but I wondered if my anecdote might be generalised. After a discussion with ChatGPT, Grok, DeepSeek, Gemini, and Claude, I concluded that it can. As is my practice for academic writing, I formulate a thesis and then an abstract at the start. Then comes the real work.
Audio: NotebookLM summary podcast of this topic.NotebookLM Infographic on this topic.
Thesis Statement
Derek Parfit’s moral mathematics relies on an undefended identification between physiological relief and suffering-reduction. Liminal experience exposes the instability of that identification at its source: what is addressed may be a bodily deficit while the suffering that matters lies elsewhere, in suspended indeterminacy itself. Because the preservationist grammar Parfit inherits treats continued life as presumptively good and bodily modulation as prima facie benefit, it cannot distinguish cases in which relief tracks morally salient suffering from cases in which it merely maintains the middle.
Abstract
This essay argues that Derek Parfit’s discussions of ‘moral mathematics’ in Reasons and Persons are not neutral exercises in moral reasoning but operations conducted within a prior ontological grammar that predetermines what can count as a benefit, a harm, and a morally salient outcome. While Parfit explicitly addresses aggregation, commensurability, and imperceptible effects, his examples presuppose an unexamined identification: that physiological relief tracks suffering-reduction, and that such reduction, however marginal, constitutes benefit within a life treated as presumptively worth preserving. This preservationist orientation is not argued for but built into the structure of the cases themselves.
The essay develops this critique through Parfit’s micro-allocation cases, particularly those involving the distribution of small amounts of water to relieve thirst. These examples appear to demonstrate that imperceptible reductions in suffering can aggregate into morally significant goods. But the argument depends on a prior identification that may fail at the point of origin. Slaking thirst addresses a physiological deficit; it does not necessarily diminish the suffering that is morally salient to the subject. The essay does not claim that physiological modulation never tracks suffering-reduction – in many cases it plainly does – but that Parfit’s grammar lacks the resources to distinguish the cases in which it does from those in which it does not. It treats all bodily modulation as benefit by default, and this default is what the essay sets out to make visible.
Drawing on a first-person account of critical illness – respiratory failure, not pain; a demand not for comfort but for determination in either direction – the essay argues that such cases function not as marginal exceptions but as diagnostics that reveal the grammar operating on the wrong dimension of the moral object. The experience of wanting not relief but resolution (‘pick a side’) is both possible and intelligible, yet the framework has no notation for it. What the intervention addressed was a physiological deficit; what it left untouched was suspended indeterminacy – the condition of being maintained in the middle, neither recovering nor ending. That the trajectory eventually resolved toward survival cannot retroactively validate the intervention on the axis that mattered during the interval itself; to argue otherwise would be to confuse post hoc survivorship with moral justification.
The essay argues further that this limitation belongs not to Parfit alone but to a broader preservationist syntax operative across Western medical ethics, legal frameworks governing end-of-life care, and liberal moral philosophy more generally. Within this grammar, life functions as the unmarked container of value; sustaining it is treated as prior to any calculation about its contents; and cessation requires special licence. The cultural entrenchment of this grammar explains why Parfit’s examples feel intuitively compelling: they inherit commitments so deeply embedded that they register as neutral premises rather than contestable positions. The point is not that preservationism is indefensible but that it remains undefended – operative yet unexamined.
Finally, the essay notes that Singer’s universalisation of moral responsibility intensifies rather than resolves the underlying difficulty, since it collapses the bounded cases on which Parfit’s arithmetic depends. What emerges is not a disagreement about consequentialism but about the grammar through which suffering, benefit, and moral salience are first made legible – and about whether that grammar can survive contact with the full range of conditions it purports to govern.
The trolley problem’s borrowed ontology was already philosophically dubious in the seminar room. It becomes materially dangerous when compiled into autonomous systems, because assumptions that once guided thought experiments now govern conduct without appeal.
The first essay argued that the trolley problem is not a neutral moral test but a borrowed ontological grammar. It preformats the scene before reasoning begins, then invites us to mistake compliance with its terms for moral insight. All of that was bad enough when confined to philosophy seminars and undergraduate anguish.
It’s even worse now. Grammar has escaped the classroom. It’s been formalised, compiled, and deployed in systems that make decisions about who lives and who dies. And it wasn’t adopted because it is morally sound. It was adopted because it’s formally legible. Legibility rears its ugly head.
Autonomous systems don’t inherit trolley logic because no one’s examined it and found it adequate to the moral world. They inherit it because it’s the sort of ontology a machine can process: discretised, scalar, optimisable. Computational tractability is not a neutral filter. It selects for ontologies that can be ranked and calculated, and discards what can’t. Trolley grammar survives not on but on formatability. The philosophical problems didn’t get solved. They got encoded.
Audio: NotebookLM summary podcast of this topic.
The Grammar Gets Compiled
The autonomous vehicle ethics literature is, for the most part, the trolley problem with a chassis bolted on.
Public debate still poses the same stale questions in a shinier casing: one pedestrian or five, passenger or crowd, young or old, many or few. These dominate media headlines and a remarkable number of engineering white papers. They are also, without exception, trolley questions – which means they carry every presupposition the first essay indicted.
They assume:
persons are countable units
deaths are commensurable
the relevant moral act is optimisation over comparable outcomes
And they assume all of this so completely that the engineering literature rarely pauses to ask whether any of it’s true. It simply proceeds as though the ontology were settled, because – and let’s be honest here – for computational purposes, it has to be.
This is the quiet scandal. The trolley grammar wasn’t scrutinised and then selected. It was convenient and so inherited. Engineers needed inputs that could be discretised, outputs that could be ranked, and an objective function that could be minimised. The trolley ontology arrived pre-packaged for exactly that specification. The fit was not philosophical. It was architectural. Funny, that.
Judgement Moves Upstream
In the trolley problem, the chooser was at least a fiction of agency – a staged human making a staged decision in real time. That fiction was already problematic. In the autonomous vehicle, even that residual theatre is over.
The ‘decision’ about who to hit, who to spare, and what to optimise isn’t made at the moment of impact. It’s made months or years before – in a design meeting, a spec document, a policy gradient, a loss function. The human chooser doesn’t disappear so much as retreat upstream, where moral judgement is converted into a spec and then forgotten as a latent judgment.
The engineer who writes the objective function is, in a meaningful sense, the person pulling the lever – though not likely culpable or legally liable. In my accounting, they should be, but they don’t experience themselves that way. They experience themselves as solving a technical problem, which it is… among other things. The moral content of their decisions is dissolved into parameters, weights, and optimisation targets, at which point it becomes invisible as moral content. The judgment is still there – baked into code, where it executes without renewed deliberation, without situational awareness, without the capacity to recognise an exception. The trolley problem’s fictional chooser has found their ideal form – not a person at all, but a function call.
Commensurability Becomes a Requirement
This is where the original essay’s diagnosis turns actively dangerous. In the seminar room, commensurability was a presupposition one could interrogate; could refuse; could argue that lives are not the sort of thing that submit to arithmetic, and the worst that happened was a lively tutorial. In engineering, commensurability isn’t a presupposition. It’s a precondition. See James C Scott’s Seeing Like a State.
You can’t write a decision algorithm without assigning comparable values to outcomes. To optimise, you need a scalar or a ranking. To rank, you need commensurable outputs. The system can’t tolerate genuine incommensurability – not because incommensurability is philosophically wrong, but because it is computationally intractable. So what was once a dubious metaphysical assumption becomes an architectural necessity.
The same structure appears in algorithmic triage. A hospital system designed to allocate ventilators during a crisis must score patients on factors like age, comorbidities, projected survival, and so on. Each patient becomes a datum. Each datum enters a ranking, which produces an allocation, which determines who breathes. In some political circles, these might have been cast as death panels. Every step in that chain requires the commensurability that the trolley grammar simply assumed and that the first essay argued was never justified. The machine demands the ontology that the philosopher merely entertained.
And here is the cruelty of it all. In the seminar, you could resist the grammar. You could say: ‘These lives are not commensurable’, ‘this comparison is malformed’, or ‘I refuse the maths’. The system can’t refuse the ontology it was built to execute. It’ll compute within the borrowed grammar until it’s switched off or until someone it couldn’t see is killed by an assumption nobody thought to question.
Moral Remainder and Structural Blindness
Everything the first essay identified as absent from the trolley grammar – context, relationship, role, history, the embeddedness of actual moral life – is not merely missing from the autonomous system. It’s structurally excluded by the requirements of the platform.
Role and obligation. Narrative history. Situated responsibility. Relational asymmetry. Tacit social meaning. Unquantified vulnerability. The possibility that not all harms belong in one metric space at all, ad infinitum… None of these can be rendered as a tractable variable, and what can’t be rendered as a tractable variable isn’t weighed lightly…or at all. Humans bask in their hubris, the purported ability to tame complexity, but their track record tells a different story.
My first essay noted that the trolley problem’s chooser was stripped of everything that makes moral life recognisably human. The autonomous system completes that stripping and makes it permanent. The philosophy student might resist the grammar inarticulately – might feel, without quite being able to say why, that something has been left out. The machine has no such unease. It has no friction, no nagging sense that the map has omitted something important about the territory. It just acts within the ontology it’s given; and the ontology was given by people who inherited it from a thought experiment that was never adequate from the start. Compilation doesn’t merely omit moral texture; it excludes whatever can’t survive formalisation – another procrustean bed. And unlike a bad philosophical argument, which can be refuted, published against, or simply ignored, a bad ontology compiled into infrastructure governs silently. It doesn’t announce its assumptions or invite dissent. It just administers – mini Eichmanns in waiting.
The trolley problem asked what you’d do at the lever. It at least had the decency to pretend you were present for the decision. The autonomous vehicle has already been told what counts – by engineers who mistake ontology for specification, by a machine that can’t question the grammar it executes. In the trolley problem, the borrowed ontology framed the question. In the autonomous vehicle, it drives the car.
This episode is an introduction to the 7-part series that discusses phenomenologists who laid the foundation on which the Mediated Encounter Ontology (MEOW) is built.
Audio: Introductory Podcast for The Architecture of Encounter
This series begins with philosophers from Descartes through Berkeley, Locke, and Hume to Kant, who will be the focus of the first episode. Except for this introduction, which is 15 minutes. Each episode is around 7 minutes because I wanted to keep them bite-sized.
