// Cydonis Heavy Industries, (C.H.I), Ltd.

🌌

Category: Essays

  • Siloes

    Siloes

    If I Ran Silo 1: How I’d Redesign the Algorithm

    1,350 words


    Spoiler warning: this post covers Silo up to and including the Season 3 finale, “Troy”.

    For three seasons, the Algorithm has been the most powerful presence in Silo, an unseen system judging the stability of Silo 18 and deciding whether 10,000 people live or die. Season 3 finally pulled back the curtain. There is no machine. There is Victor Crnkovich in Silo 1, speaking through a voice modulator, with Rosalind Thurman as director and a memory-wiped Daniel Keene, now “Troy”, woken from cryosleep whenever a silo needs putting down.

    The moral case against Silo 1 hardly needs making. What interests me is that, even on its own chilling terms, the system is badly designed. If the goal is to keep 50 isolated populations obedient for centuries, the founders made a string of mistakes any strategist should have spotted. So here is a thought experiment: if I were the Algorithm, what would I have done differently?

    The core flaw: power built on a lie
    Silo 1’s control rests on a shared false belief. The residents are not only told that the outside is lethal and that the Algorithm is infallible; they are kept from their own history altogether. The project’s stated aim is to free humanity from its past.

    The trouble with ruling through a false belief is that it is a coordination problem waiting to be solved. Obedience depends less on what each resident privately believes than on what they think everyone else believes. Once the truth becomes common knowledge, when people know that others know, compliance can collapse almost overnight. Nearly every crisis in the show follows from Silo 1 creating ways for that knowledge to spread.

    1. Don’t send dissenters to the one place they can test the lie
    Cleaning is meant to be the ultimate public deterrent: challenge the system and you walk out to certain death while everyone watches. It works perfectly, right up until someone survives.

    It is an astonishingly risky design. The punishment puts the most curious, rule-breaking people exactly where the fake display on the screens can be exposed, and one survivor, Juliette, was enough to shatter it. A deterrent with even a small chance of disproving your entire narrative in front of the whole population is not worth having. I would handle punishment inside the silo, out of sight of the outside world, or at the very least make sure no engineer in Mechanical could quietly defeat the suit.

    2. Never give one person all the knowledge
    In Silo 18, the head of IT is effectively the only person who knows the truth. That puts the whole regime at the mercy of one person’s temperament, ambitions and succession politics. Bernard’s paranoia, his clumsy choice of successor and his eventual turn against the system show how badly that can go.

    A sturdier design spreads knowledge across several people who don’t fully trust each other and can check each other’s behaviour. Just as important, Silo 1 needs independent ways of watching what is going on. In the finale, Camille’s feint worked because Victor was relying on a single human channel he could not verify. An Algorithm that can be bluffed is not much of an Algorithm.

    3. Make the threat automatic, or keep it secret
    In game theory, a deterrent is most credible when the side issuing it cannot back down. Silo 1 got this the wrong way round. The Safeguard ran through a physical pipe that residents could find and cap, so they knew the threat was real and also that it could be beaten. Worse, it was run interactively by a person, which told everyone there was someone at the other end who could be negotiated with, delayed or tricked.

    I would choose one of two coherent options. Either the threat is fully automatic, triggered by fixed and published conditions with no human discretion and no physical point of failure, or its existence is never revealed at all. The half-hidden, human-operated, breakable version Silo 1 used combines the weaknesses of both.

    4. Punish in proportion
    The Directive treats contamination between silos as the one unforgivable offence, and entire silos are expendable to prevent it. That is the equivalent of a strategy that punishes everyone for a single defection, and over repeated rounds it has two fatal problems.

    First, it destroys the very thing the project exists to protect. With 50 silos and centuries to go, wiping out whole populations steadily drains the stock of humanity the founders claim to be saving. Second, it removes any reason for restraint. Once a silo believes it is already condemned, it has nothing left to lose, which is exactly the point at which people rise up. Graduated, targeted sanctions keep deterrence working while leaving people a reason to stop.

    5. Take the mystique out of the past
    The total ban on relics was meant to cut people off from history. In practice it made every surviving object enormously significant. A forbidden children’s book or a single photograph becomes proof that something was hidden, and hidden things invite investigation.

    I would do the opposite: allow a carefully chosen supply of harmless old-world objects so they become ordinary. Curiosity also needs a safe outlet. The most inquisitive residents are the most dangerous if you fight them and the most useful if you recruit them. Bernard half understood this with Juliette, but by then trust had already broken down. A system that routinely brought its brightest dissenters into IT, rather than sending them out to clean, would face far fewer rebellions.

    6. Don’t build what you’ve banned
    If isolation between silos is the top priority, a network of tunnels connecting them is a baffling weakness. It is like announcing that doors are the gravest threat to the regime and then fitting every wall with one. Either don’t build them, or make their existence the most closely guarded fact in the entire project. By the end of Season 3, Juliette knows about them, and that is catastrophic for Silo 1.

    7. Keep the enforcer reliable
    Troy is Silo 1’s troubleshooter, the person trusted with the hardest decisions. He is also plainly deteriorating: each time he is woken, flashes of Helen come back and he asks more questions. Victor’s response is simply to put him back to sleep.

    That is putting off the problem, not solving it. The person with the most lethal authority in the system should be its most predictable part, and Silo 1 is relying on memory suppression that visibly fails. Sooner or later, the enforcer remembers.

    8. Make promises that can be checked
    In the finale, Troy offers Silo 18 a deal: he will leave the Safeguard off as long as they avoid contact with the other silos. Juliette seems to accept, while planning to reach Silo 1 through the tunnels.

    Of course she does. Silo 1 has lied to her people for generations and has shown it will kill entire populations. A promise from an actor with that record is worthless unless it can be verified or enforced. If I genuinely wanted a settlement, I would give her something concrete and lasting: real control over the capped pipe, independent monitoring, or even a seat in the project itself. Commitments only hold when breaking them costs the one who made them.

    The uncomfortable conclusion
    What strikes me about this list is that almost every change makes the system less cruel as well as more stable. That isn’t a coincidence. Regimes held together by deception and collective punishment are fragile precisely because they give people every reason to resist and nothing to lose by doing so. Legitimacy is far cheaper to maintain than terror.

    The single most powerful move open to Silo 1 would not involve pipes, voice modulators or cryo-sleep. It would be offering residents a believable future: a real, verifiable date when the doors will open, and a reason to wait for it. People will put up with a great deal for something they believe in. Silo 1 chose fear instead, and with Juliette heading for its front door in the final season, it may be about to find out how poor a choice that was.

    Wherever I go, there I am, I am, I am…

  • The Box

    The Box

    How to Actually Destroy the Box: An Engineering Post-Mortem on Dark Matter’s Central MacGuffin.

    1,119 words



    Every season of Dark Matter eventually arrives at the same problem. Someone, usually Jason, decides the box has to go. It’s the thing that ruined his life, ended Daniela’s marriage to the “right” version of him, and now threatens to keep spawning doppelgangers for the rest of their days. The solution always sounds simple when a character says it out loud: destroy the box. What the show is much less interested in is how, which leaves the door wide open for anyone with a passing interest in materials science to have a go at the problem properly.

    What ‘the box’ actually is:

    Strip away the drama and the box is a fairly modest piece of engineering. It’s primarily a sealed metal chamber. A traveller climbs in, gets injected with a compound (the show calls various versions of it Lavender Wings, Lilac Wings, and so on) that inhibits activity in the prefrontal cortex, and this induces a state of quantum superposition, aided by technology embedded into the box’s inner hull. From inside that state, the traveller perceives a corridor of doors, each one leading to a different branch of the multiverse, and picks their way toward whichever reality they’re after.

    Crucially, none of that magic lives in the metal. The box is an isolation chamber, not a spellbook. Its entire function is to hold a human body still and undisturbed long enough for an altered brain state to do the actual work. That distinction matters enormously once you start thinking about how to disable it.

    The real weakness, according to the show itself

    Season 2 quietly hands the audience the box’s actual vulnerability. When a duplicate box gets relocated within Chicago, moved only a few miles from where it originally stood, Amanda works out that the move has severed its quantum connection to what she calls the box’s “prime power source”: some kind of anchor point shared across every multiversal copy of the box. Break the connection to that anchor, and you’re left with, in her words, just a box. No corridor, no doors, no superposition, just a large metal container taking up warehouse space.

    The show has also shown us the second, blunter method works too: a tornado is depicted destroying a box outright elsewhere in the timeline. So canon actually gives us two independently validated failure modes:

    Structural annihilation – enough force or heat to physically wreck the chamber. No chamber, no isolation, no portal.
    Severance – moving the box far enough from its anchor point that the link degrades to nothing, without needing to damage it at all.

    Anything you propose as a destruction method should be judged against those two mechanisms, because they’re the only ones the text has actually confirmed.

    Why concrete doesn’t cut it

    The instinctive answer, entomb it in concrete, fails both tests. Pouring concrete around the box doesn’t move it relative to its anchor point, the coordinates are unchanged, you’ve simply added mass nearby. And it doesn’t structurally destroy the chamber either. Burying something isn’t the same as crushing it. What you’re left with is access denial: nobody can open the door and climb in, which has some value, but it’s containment rather than destruction. And containment is precisely the failure mode the show keeps punishing its characters for relying on, because a box that still physically exists is a box that a sufficiently motivated alternate version of yourself, with unlimited time and resources, eventually digs up again.

    A better stratagem: helicopter, Iceland, volcano

    This is where things get properly satisfying, because a lava drop stacks both validated mechanisms at once rather than betting on just one.

    Severance is trivially achieved. If a few miles across a single city was enough to sever the link in canon, then several thousand kilometres to a volcanic peninsula in the North Atlantic isn’t a marginal improvement, it’s overkill by several orders of magnitude. Strictly speaking you wouldn’t even need the volcano just to disable the box functionally; the flight there already does that part of the job.

    Structural destruction is where the physics gets interesting. Icelandic basaltic lava, the kind produced by the Reykjanes Peninsula’s fissure eruptions, typically sits around 1,100–1,200°C. Mild and structural steel melts at roughly 1,370–1,510°C, meaning a steel box dropped into an active lava lake might not fully liquefy on contact. What it will do is warp under thermal stress, oxidise rapidly, and get chemically attacked by dissolved sulphur compounds, all while being progressively swallowed by the flow. Any internal wiring or chemical-injection hardware has essentially no chance of surviving that environment regardless of what the outer shell is made from. The practical result is a box that is destroyed in every sense that matters, inoperable, and buried under solidifying rock, which arguably beats pure concrete containment, since it removes the long-term “someone eventually excavates it” risk that burial alone leaves open.

    The two wrinkles any serious multiverse-ending team should plan around *wink*:

    Timing. Lava requires an actual eruption, and Iceland doesn’t oblige on demand. As of late 2026 the Reykjanes Peninsula’s Sundhnúkur crater system has been quiet since its last eruption in mid-2025, though magma has continued accumulating beneath Svartsengi and the region is expected to remain volcanically active for decades. If the schedule doesn’t cooperate, Kilauea in Hawaii is the more dependable choice; it has erupted dozens of times in the past two years alone and is far less likely to leave you waiting around with a box strapped to a cargo hook.

    Payload. Nothing in the show gives exact dimensions for the box, but it’s consistently moved by forklift and truck rather than crane, which puts it plausibly in the low single-digit tonnes. That’s comfortably within the external sling-load capacity of a heavy-lift helicopter such as a Sikorsky Skycrane. The genuinely dangerous part of the operation isn’t the lift at all, it’s holding a stable hover near an active vent long enough to release accurately, given the sulphur dioxide plumes and violent thermal updrafts that come with any close approach to a lava lake.

    The uncomfortable postscript

    Even a technically perfect execution of this plan only solves a local problem. Destroying one instance of the box, however thoroughly, does nothing about the countless other Jasons across the branching multiverse who have built, or will eventually build, boxes of their own. That’s really the show’s quiet argument underneath all the corridor-running and doppelganger drama: there is no engineering solution to infinite branching possibility. You can win the battle against one box. You cannot win the war against the concept of it.

    Still, if you only need to win the battle: helicopter, Iceland, volcano. It’s sound.

    Cydonis. Whimsy versus despair.

  • Shopping Trolleys In The Polycrisis

    Shopping Trolleys In The Polycrisis

    The ‘Trolley Problem’: What a Shopping Cart Can Teach You About End Stage/Terminal Stage Surveillance Capitalism…

    1,448 words


    There is a particular kind of fury reserved for a supermarket trolley with a mind of its own. You line it up, push forward, and it veers sideways as though possessed, or worse, one wheel has gone square and every rotation produces a sound like a dying accordion. It seems a trivial complaint. It is not. The humble trolley is, in fact, a masterclass in design philosophy, and once you start looking at it properly, it opens a door onto something much larger: how supermarkets think about their customers, and how far that thinking now extends into your data, your face, and your habits.

    The Trolley as a Design Artefact

    Not all trolleys are created equal, and the differences are not accidental.

    Take the large chains that dominate the UK high street. Their trolleys are free to take, no deposit required, and the wheels are, frankly, an afterthought. Anyone who has wrestled a cart with one wobbling caster and one that has fused into a rigid, non-rotating brick knows the particular indignity of shopping while fighting your own equipment. There is no real incentive for the retailer to fix this. The trolley gets you round the shop and back to the till; wheel quality does not appear on any balance sheet that matters to head office.

    Contrast that with the discount chains that require a pound coin (or a special token) to release a trolley from its bay. These trolleys tend to glide. The wheels are true, the frame doesn’t rattle, and the whole experience is noticeably smoother. This isn’t a coincidence, and it isn’t really about the trolleys themselves. It’s a master-stroke of behavioural design. Nobody wants to lose their pound, so trolleys get returned to their bays rather than abandoned in car parks or nearby streets. The chain saves on staff time spent rounding up stray trolleys, and the deposit mechanism largely pays for itself in reduced losses and lower collection costs. The customer experiences it as “these trolleys are just better”, when really what they’re experiencing is a company that has quietly outsourced its trolley-logistics problem to a one-pound psychological nudge.

    It is a small, almost invisible example of a much bigger idea: that the shopping experience you are given is never neutral. Every element, from trolley wheels to aisle layout to the placement of the bakery smell, has been considered, tested, and optimised for something. Usually that something is not your convenience. It’s the retailer’s bottom line, dressed up as convenience.

    And then there’s the reward scheme layered on top, offering money off fuel at a station the chain itself doesn’t operate anywhere near your area. It is a strange kind of loyalty gesture: a benefit engineered for an average customer who may not exist in your postcode, handed out anyway because the loyalty app’s back end doesn’t know or particularly care whether it’s useful to you specifically. Which brings us to the app itself, and the far larger design decision behind it.

    Loyalty Cards: The Discount That Costs More Than It Saves

    Loyalty schemes are sold to us as a straightforward trade: scan your card, get a bit of money off, everybody wins. In practice, the exchange is wildly asymmetric. You get a few pence off tea bags. The retailer gets a granular, timestamped, endlessly cross-referenced record of your entire shopping life.

    Consider what a till receipt actually contains once it is tied to a loyalty number. Not just what you bought, but when, how often, in what combinations, and how that changes over time. Do you buy own-brand or premium? Do you stock up before bank holidays? Do you only ever buy the reduced yellow-sticker items after 8pm? Did your alcohol purchases spike, or your nappy purchases start, or stop? None of this requires anything sinister like a hidden microphone. The receipt alone tells a remarkably intimate story, and loyalty schemes exist specifically to attach a name and a household to that story rather than let it stay anonymous.

    That data has value well beyond “let’s send this customer a coupon”. It can inform stock decisions, sold on to third parties, or, more controversially, used to personalise pricing. The idea that two shoppers might be shown different prices for the identical tin of beans, based on what an algorithm has inferred about their price sensitivity, sounds like speculative fiction. It isn’t. It’s a live discussion in retail and competition circles precisely because the data infrastructure to do it already exists, sitting quietly behind every loyalty card scan.

    The pitch, “save money by joining”, inverts the actual transaction. You are not saving money. You are selling behavioural data at a steep discount, and the “savings” are your own money being partially returned to you as a rebate for having handed over something a lot more valuable.

    The Watching Aisle

    If loyalty cards are the polite, consensual end of retail surveillance, the far end is considerably less polite and a good deal less consensual: cameras that don’t just record you, but attempt to recognise you.

    Facial recognition in UK retail is no longer hypothetical.

    Facewatch, a system used by a number of retailers, checks faces against a shared watchlist, ostensibly to flag people previously involved in theft or antisocial behaviour. Framed narrowly, this sounds like sensible loss prevention. The trouble is that “narrowly” is doing a lot of work in that sentence, and nobody outside the company holds a clear, verifiable picture of where the narrow use ends and a broader profiling capability begins. You are, in effect, asked to trust the system is only doing the specific thing it claims, with very little visibility into whether that trust is warranted.

    It is worth understanding, purely as a matter of how the technology works, why some of these camera systems behave the way they do. Most digital cameras use CCD or CMOS sensors that are sensitive not only to visible light but also to near-infrared wavelengths that the human eye cannot perceive. This is why older or budget CCTV units, and infrared night-vision cameras generally, can be affected by strong infrared light sources in ways that visible light does not affect them: the sensor “sees” wavelengths a person in the room does not.

    This has been a known quirk for a long time, well-known enough that it has shaped how modern security cameras are built. Many newer units, especially ones installed specifically with this vulnerability in mind, now include IR-cut filters, physical filters that block near-infrared wavelengths from reaching the sensor at all. It is a neat, small example of an ongoing technical arms race: a known sensor limitation becomes public knowledge, manufacturers respond by filtering out the exploit, and the old trick stops working on new hardware, while remaining effective against the huge installed base of older or cheaper cameras that predate the fix. The broader lesson isn’t really about cameras. It’s a reminder that any system built to observe people is built on physical hardware with physical limits, and physical limits eventually became public knowledge and adjusted for.

    The Trade Nobody Explicitly Agreed To:

    Pull the thread from squeaky trolley wheels to loyalty card data to facial recognition, and a pattern appears. At every step, the retail environment has been engineered to extract something from you, whether that’s your compliance in returning a trolley, your purchase history, or your face, and at every step the retailer frames it as a benefit to you: convenience, savings, security. Sometimes it genuinely is a mutual benefit. A well-designed trolley bay really does mean fewer abandoned trolleys cluttering the car park. A discount really is a discount, however small.

    But the honest version of the pitch would sound rather different: “Give us a pound and behavioural economics will make sure you bring the trolley back.” “Give us your shopping history and we’ll give you a fraction of its value back in vouchers, keep the rest, and possibly use it to charge you and your neighbour different prices for the same shop.” “Let us build a facial record of your visits, and trust usssss to only use it the way we say we will.” :’-(

    None of this requires outrage, exactly. It mostly requires noticing. The trolley wheels are not really about trolley wheels. They’re a small, physical demonstration of a much larger principle that runs through modern retail: the environment you shop in has been designed, in exhaustive and often invisible detail, and very little of that design was done with your interests as the primary consideration. Once you see it in the trolleys, you start seeing it everywhere else too…

  • The Observer Effect

    The Observer Effect

    Dr. Elena Vasquez had spent eleven years trying to reconcile general relativity with quantum mechanics, and she had come to the conclusion, quietly, over many sleepless nights, that the problem wasn’t in the mathematics. The problem was in her. Some perceptual limitation, baked into the wetware, that made it impossible to hold curved spacetime and probabilistic wavefunctions in the same thought without one collapsing into the other.

    She had not come to the retreat centre in the hills outside São Paulo to fix physics. She told herself that, repeatedly, on the drive up. She had come because her husband had left in March, because her postdoc had been denied, because she had not published anything worth reading in three years, and because a colleague – a string theorist who had gone quiet and strange after his own retreat, and then oddly brilliant – had told her, *just go, you’ll understand something you can’t get from equations.*

    She had not planned to combine anything. The facilitator, a soft-spoken woman named Iara who had run these ceremonies for two decades, had been firm about that. But on the second night, in the dark after the first purge had passed and the visions had not yet properly begun, Elena had made a decision that was not quite a decision – the kind of thing that happens in an altered state, where intention and impulse blur past the point of separation. She had brought her own mushrooms. She took them anyway, without telling Iara, four grams ground into the last of her water, on top of a ceremonial dose of ayahuasca that was already working through her blood.

    —

    The nausea arrived first, as it always did with the vine, but this time it did not stay in her stomach. It became a property of *space itself.* The walls of the maloca breathed in slow peristaltic waves, and Elena understood, with total clinical certainty, that she was inside something that was digesting her.

    She reached for the bucket. She did not make it. Iara’s assistant came with a cloth and murmured something in Portuguese that Elena’s brain, disassembling in real time, translated not as words but as a shape – a soft blue parabola of care, arcing over her and then gone.

    *I should tell someone,* she thought. *I took something else. This isn’t standard protocol.* But the thought arrived already fragmented, already being folded into something larger, the way a sentence gets pulled apart by strong wind before it’s finished leaving your mouth.

    Her heart was doing something she didn’t like. Not racing exactly – more like it had stopped agreeing to run on a single rhythm and was now negotiating between several tempos at once, and she felt this not as fear but as *data.* Somewhere very far away, a small remaining sliver of Dr. Vasquez noted: *tachycardia, possible serotonergic excess, this is exactly the interaction you warned your students about.* That sliver was calm. It observed. It did not have hands anymore, or a mouth, so it could not do anything but watch.

    —

    Then the visions became geometric, and this was the part she would spend the rest of her life trying, and failing, to write down properly.

    She saw – no, *saw* was wrong, there was no seeing, there was only *being the fact of* – a lattice. Not a metaphor for a lattice. The actual underlying structure, she felt certain, of everything. It was not spacetime as she’d modelled it, smooth and differentiable. It was granular, foaming, appearing and disappearing at a rate she could not measure because the concept of a measuring rate required a stable observer, and she was not stable. She was inside the foam.

    Voices came through the geometry. Not English, not Portuguese, not language really – more like the vine itself was old and had opinions, and the opinions arrived as pressure, as intention, as a very ancient patience regarding her smallness. She had read about “the mother” in every ayahuasca account she’d skimmed, half-embarrassed, before the trip, filing it under *anthropomorphisation, interesting but unscientific.* Now the distinction between *anthropomorphised* and *actually a mind* dissolved along with everything else, and she wept, not from fear, but from something closer to being *known.*

    The mushrooms arrived on top of this like a second frequency modulating the first – faster, brighter, more fractal, less narrative. Where the vine gave her structure and voice, the mushrooms gave her recursive visual detail: the lattice sprouted eyes, then sprouted lattices of eyes, each one a further zoom, and she understood – again, with that same false clean certainty that altered states hand out like candy – that she was looking at *herself*, at the structure of a mind trying to observe its own observation, an infinite regress of a physicist trying to physics her way out of the trip and only sinking further into it.

    *This is just the Copenhagen interpretation having a nervous breakdown,* she thought, and somewhere a version of her that still had a sense of humor almost laughed, except laughing required a body with reliable diaphragm control and hers was busy being folded through eleven dimensions of something that might have been grief.

    —

    Because underneath the geometry, the grief was the real content. It always is, people said, and she hadn’t believed them before this. She saw her husband not as a memory but as a live variable, still coupled to her, refusing to decouple no matter how far the equations of her life had drifted from his. She saw her father, who had wanted her to be a doctor, a real one, hovering disappointed at the edge of every equation she’d ever solved. She saw the postdoc rejection not as a professional setback but as a small, precise wound that had been standing in for a much larger question – *does the universe, in fact, want you here* – and the vine, or the mushroom, or the seething foam of everything, answered that question not with words but with an unbearable, oceanic *yes* that broke something loose in her chest and she cried in a way she had not cried since she was a child, ugly and total, snot and tears and bile.

    Her pulse would not settle. At some point Iara was kneeling in front of her, taking her wrist, her face carved with real concern, saying words that landed as concrete instructions even through the dissolving membrane of Elena’s ordinary comprehension – *breathe with me, you are safe, you are in a body, breathe* – and something in the specific gravity of Iara’s voice, the way it refused to be swept into metaphor, gave Elena a handhold. She breathed. The world did not stop being a foaming lattice of recursive eyes, but her heart, mercifully, began to remember a single rhythm.

    —

    Hours passed that did not behave like hours. At the peak – if it was a peak, if peaks were even the right shape for a place with no vertical axis – she experienced something she could only describe afterward, badly, as *becoming the wavefunction instead of measuring it.* All her career she had stood outside the equation, a physicist, a knower, a subject regarding an object. Now there was no outside. She *was* the superposition, unresolved, every possible Elena – the one who’d stayed married, the one who’d left academia, the one who’d never come to Brazil at all – coexisting without collapse, and it was not distressing, not by this point; it was the first time in a decade she had felt something like rest, because if she was every possibility at once, none of them could disappoint her, and none of them could leave.

    She thought, with sudden and startling clarity: *this is why they call it medicine.* Not because it fixed anything. Because it showed her the shape of the thing she’d been carrying, all lit up at once, so she couldn’t pretend anymore it was only a professional problem, or only a marriage problem, or only a mathematics problem. It was one problem. It had always been one problem.

    —

    The comedown, when it finally came, came slowly and with its own kind of violence – a headache like a fist, a body that felt wrung out and slightly betrayed, a thirst she couldn’t satisfy for an hour. Iara sat with her as the light came up grey through the maloca’s gaps, and asked, gently, without judgement, what exactly Elena had taken.

    “Both,” Elena admitted. “I took both.”

    Iara didn’t scold her. She just nodded slowly, the way people do when a thing they suspected has been confirmed, and said, “Your heart worked very hard last night. You are lucky it is a good heart.” Then, after a while: “What did you see?”

    Elena tried to explain the lattice, the recursive eyes, the wavefunction, and it came out sounding like nonsense, like every account she’d ever rolled her eyes at in a journal she wasn’t proud of reading. She laughed, hoarse and real. “I don’t have language for it. I have language for quarks and I don’t have language for this.”

    —

    Back in her apartment two weeks later, on a Tuesday, staring at a whiteboard she hadn’t touched since before Brazil, Elena found herself writing not an equation but a question, in small letters at the corner of the board, the kind of note you leave for the version of yourself who comes back tomorrow: *what if the observer was never outside the system to begin with?*

    She did not solve unification that year, or the next. The mathematics did not care what she had seen in the dark, foaming and recursive and briefly infinite. But something had shifted in the way she held the question – less like a puzzle to be cracked from a safe distance, more like a thing she was standing inside of, had always been standing inside of, the way she had, for one unbearable and clarifying night, been standing inside a vine’s old and patient mind.

    She never combined them again. Once had been enough to nearly stop her heart and completely rearrange the rest of her. She kept the memory the way you keep a scar – proof of something real that happened, evidence she was careful not to need again.
  • The Photon Falsifiability Gap: The Mystery of the Tiny Click

    The Photon Falsifiability Gap: The Mystery of the Tiny Click

    admin
    1,192 words



    Imagine you are a detective. Your job is to listen for a very, very quiet knock on a door. The knock is so quiet that you can barely hear it. Now imagine that the door also makes random creaking noises all by itself, and those creaks sound almost exactly like the knock.

    How do you know if what you just heard was someone knocking, or just the door being noisy?

    That question – that exact puzzle – is at the heart of a real science mystery. Scientists gave it a name: the Photon Falsifiability Gap. It sounds like a big, fancy phrase, but once you break it into pieces, it’s actually a story about light, invisible energy, and a very clever machine that sometimes fools itself.

    What is a photon?

    Light is made of tiny little packets, kind of like light is made of Lego bricks. Each one of these tiny packets is called a photon. You can’t see a single photon with your eyes – they’re much too small and light for that. But scientists have built machines that can detect a single photon, one at a time. That’s an amazing feat, like being able to hear a single grain of sand hit the floor.

    Meet the super-listener: the photomultiplier

    The machine that can “hear” a single photon is called a photomultiplier tube (you can just call it a PMT for short). Think of it like a super-sensitive microphone, except instead of listening for sound, it’s listening for light.

    Here’s how it works, in simple terms:

    A single photon flies into the tube and hits a special metal surface.
    That hit knocks loose one tiny electron.
    The tube then multiplies that one electron into millions of electrons, like a snowball rolling downhill and picking up more snow.
    All those electrons create an electrical “click” that a computer can count.

    So a photomultiplier tube turns one whisper-quiet flash of light into a loud, countable click. Pretty amazing, right?

    The invisible energy problem

    Now here’s where our mystery begins. Some materials in nature give off a very weak, very quiet form of energy. One of these is called tritium. When tritium releases its energy, it’s incredibly faint – much fainter than most other radioactive materials. It’s like a whisper compared to a shout.

    Scientists want to detect this whisper using their super-listener, the photomultiplier tube. And it can pick up that whisper… but here’s the catch.

    The catch: the door creaks too

    A photomultiplier tube isn’t perfectly silent when nothing is happening. Every once in a while, all on its own, it produces a tiny click – even when no real photon ever arrived. Scientists call this dark noise, because it happens even in total darkness, with nothing there at all.

    And here’s the twist that makes this a true mystery: a dark noise click and a real tritium click look exactly the same. Both are just one tiny electrical pulse. There is no way, just by looking at a single click, to tell which one you’re looking at.

    Go back to our detective story. You heard a knock. Was it a real knock, or just the door creaking? If both sounds are identical, you can never be 100% sure – not for that one single sound.

    Why this is called a “falsifiability” problem

    In science, there’s an important idea called falsifiability. It means that for an idea to be truly scientific, you need to be able to test it in a way that could prove it wrong if it actually is wrong.

    Here’s the problem: if you hear one click and say “that was tritium!” – there is no way to prove that statement wrong. It might have been dark noise. And if you say “that click was just dark noise!” – there’s no way to prove that wrong either. It might have been real. Neither guess can be tested for a single click. That’s the “gap” – a gap where our normal rules for testing scientific ideas don’t quite work, at least not one click at a time.

    So how do scientists solve the mystery?

    If you can’t trust a single click, what do you do? You stop listening for one knock and start counting lots of knocks over a long time.

    Here’s the trick:

    First, scientists measure how many random clicks (dark noise) happen when they know for certain there’s no tritium around. This tells them the normal “creaking” rate of the door.
    Then, they measure the click rate when tritium might be present.
    If the second number is clearly, reliably higher than the first – not just by one or two clicks, but by a lot, over and over – then they can be confident that real tritium signals are hiding inside all those clicks.

    It’s a bit like this: you can’t know if one specific creak was a knock. But if you count 1,000 creaks on a quiet night and then count 1,000 creaks plus 300 extra clicks on another night, you can be pretty confident something extra was happening on that second night – even though you still can’t point to any single click and say “that one was definitely the knock.”

    Why does any of this matter?

    This isn’t just a fun puzzle – it matters for real science:

    Finding ancient objects: Scientists use radioactive materials to figure out how old rocks, fossils, and artefacts are.
    Medicine: Doctors use tiny, safe amounts of radioactive material to see inside the human body without surgery.
    Hunting for dark matter: Physicists build giant, super-sensitive detectors – using the same kind of photomultiplier tubes – to search for mysterious particles that make up most of the universe. They face this exact same “which click was real?” problem, just on a much bigger scale.

    Every single one of these fields depends on scientists being clever enough to work around the falsifiability gap, even though they can never fully close it.

    Quick recap:


    A photon is a tiny packet of light.


    A photomultiplier tube is a machine that can detect a single photon and turn it into a loud click.


    Tritium gives off a very weak, whisper-quiet signal.


    The tube also makes random clicks by itself, called dark noise, and these look identical to real signals.


    Because of this, no single click can ever be proven to be real or fake – that’s the falsifiability gap.


    Scientists solve this not by trusting one click, but by comparing patterns of many clicks over time.


    Try it yourself:

    Here’s a fun way to feel this mystery for yourself. Get a friend and a set of headphones, or just sit in separate rooms. Have your friend randomly tap a table softly, mixed in with the normal sounds of the house (footsteps, a fan, a fridge humming). Try to guess, sound by sound, which taps were real. You’ll probably find that any single sound is a guess – but if you count for five whole minutes, you’ll likely get a much better sense of how many taps really happened, even without ever being sure about any one of them.

    That’s the same trick scientists use to solve the Photon Falsifiability Gap – not by being certain about one click, but by being smart about all the clicks together.

  • What Is the Word for That Which You Cannot Think Of?

    What Is the Word for That Which You Cannot Think Of?

    admin avatar
    admin
    3,452 words

    Nescience, cognitive closure, and the difference between a wall and a puzzle…


    There is a gap in ordinary English that you generally only notice when you need it.

    We have a rich vocabulary for not knowing things. *Ignorant*, *unaware*, *uninformed*, *clueless*, *in the dark*. But every one of these words carries a hidden assumption: that the thing not known is the sort of thing that *could* be known by the person not knowing it. Ignorance is a deficit against a standard. It implies a book unread, a lesson unlearned, a fact that was sitting there available. You can fix ignorance by handing someone a pamphlet.

    What we lack is a common word for the other case – where the failure is not in the person’s diligence but in their equipment. Where the thing isn’t merely unknown but *unthinkable by that kind of mind*, in the way that a smell is unthinkable by a thermometer. Not a gap in the map, but a limit on what can be mapped at all.

    This is not a merely lexical complaint. The word you reach for commits you to a claim about the world, and the claim is usually much stronger than you intended. So it is worth going slowly.

    I. The available words, and what each one smuggles in

    **Nescient** is the best single word. It comes from Latin *nescire*, to not-know, and it has been doing serious work in theology and epistemology for centuries. Its virtue is that it names a *condition* rather than a *failure*. To be nescient of something is not to have neglected it. Nescience is closer to a state of the organism than to a gap in its education. It is also usefully rare, which means it hasn’t been worn smooth by casual use the way *ignorant* has.

    **Incognizant** is the plainer cousin – serviceable, slightly bureaucratic, and it drifts back towards ordinary unawareness. **Insensible to** is nicely physical, with a suggestion of sensory rather than intellectual failure, which may be exactly what you want. **Benighted** is archaic and faintly insulting; it implies a darkness that someone could have walked out of. **Oblivious** is the weakest of all – it means inattentive, and inattention is a moral failing in miniature, which is precisely the wrong note.

    Then there is a second family of words, for a subtler thing: not the unawareness itself, but the unawareness *of* the unawareness. The recursive case. Because the truly closed-off mind does not experience itself as missing anything. It experiences the world as complete.

    **Anosognosia** is the clinical term, coined by Babinski in 1914, for a patient’s unawareness of their own deficit – the stroke patient with a paralysed arm who sincerely reports that the arm is fine, and who will produce elaborate reasons why they simply don’t feel like moving it right now. It has escaped the clinic and is now used, sometimes loosely, for any blind spot about a blind spot. **Scotomised**, from *scotoma*, the blind spot in the visual field, does similar work with a psychoanalytic accent: a perceptual blankness where something intolerable ought to be. Freud was suspicious of the term and preferred to talk about disavowal, but *scotoma* has survived because the image is so good. Your own retinal blind spot does not appear to you as a hole. It appears as nothing at all, seamlessly filled in. You have to hunt for it with a card and a dot to prove it is there.

    That is the phenomenology we are after. Not a hole in the world, but a world with no hole in it.

    II. Cognitive closure

    The philosophical term of art is **cognitive closure**, and it belongs to Colin McGinn, who set it out in a 1989 paper asking whether we can solve the mind-body problem. His answer was no, and his reason was structural rather than pessimistic.

    Every mind has a range. A dog’s mind is well adapted to a rich world of scent, social hierarchy, and object permanence, and it is closed to arithmetic – not because dogs are stupid in a way that could be remedied with better schooling, but because the concept of a prime number is not a thing a canine cognitive system can form. The dog is not frustrated by this. There is no dog-shaped ache where mathematics should be.

    McGinn’s move was to ask why we should assume our own range is complete. Our brains are organs shaped by selection pressures that had nothing to do with metaphysics. It would be a startling coincidence if the set of true things happened to line up exactly with the set of things a primate social-cognition engine can represent. Some truths may simply fall outside our range, and – this is the important part – we would have no way of detecting their absence. Our world would look complete, because it always does.

    The position acquired the name **mysterianism**, coined half-jokingly by Owen Flanagan after the 1960s garage band ? and the Mysterians, and it has been fought over ever since. Daniel Dennett’s objection is the sharp one, and worth keeping in your pocket: an argument from cognitive closure is very hard to distinguish, from the inside, from a failure of imagination. “I cannot conceive how X could be explained” is a fact about you. Treating it as a fact about the universe requires an extra step that mysterianism has never quite managed to justify. Every hard problem looks closed until the week before it opens.

    Hold onto that objection. It is going to matter later.

    —

    III. Umwelt: the sensory version

    If you want the version with less metaphysics and more biology, the word is **Umwelt**, from the Baltic-German biologist Jakob von Uexküll, writing in the 1930s.

    Uexküll’s insight was that an organism does not live in *the environment*. It lives in the environment *as constructible by its own sensory apparatus* – its Umwelt, its self-world. And these worlds are astonishingly narrow. His famous example is the tick, which lives in a world assembled from three cues: the smell of butyric acid, which tells it a mammal is passing beneath; a temperature of around 37 degrees, which tells it where to burrow; and the feel of hair, which tells it where not to. That is the whole world. There is no colour in it, no sound, no shape. A tick can wait years on a branch in a world made of three facts.

    The great virtue of *Umwelt* is that it makes the point without any philosophical commitment at all. It is an empirical claim about sensory ecology. The tick is not being denied access to a richer reality by some Kantian veil; it simply has three receptors. And nothing in its world indicates that anything is missing.

    If your interest is in intelligences with different bodies, *Umwelt* may be the more honest word than *cognitive closure*, because it locates the limit in the sensorium rather than in the intellect – and sensory limits are the kind of thing you can actually go and measure.

    —

    IV. The bat, and why it is not about bats…

    The essay everyone eventually arrives at is Thomas Nagel’s “What Is It Like to Be a Bat?”, published in *The Philosophical Review* in 1974, and it is worth being precise about what it argues, because it is very frequently misremembered.

    Nagel is not saying that bat echolocation is hard to imagine. He is not saying that we lack data about bats. Neurophysiologists know a great deal about the bat’s auditory cortex and can say in detail how the returning echo is processed into a representation of a moth’s position and velocity. The information is not the problem.

    His claim is about the *structure of imagination*. When we try to imagine being a bat, what we actually do is imagine ourselves with modifications: ourselves, but hanging upside down; ourselves, but with a sonar sense grafted on. That is imagination working the only way it can, by extrapolation from our own case. But it delivers only what it is like *for us* to be bat-like. It cannot deliver what it is like for a bat to be a bat, because the extrapolation always starts from the wrong place. And there is no second method available.

    So the barrier is not informational. It is not even, strictly, a barrier of intelligence. It is that the subjective character of experience is only available from a particular point of view, and the tool we use for occupying other points of view is a tool that works by dressing ourselves up.

    This is why *nescience* and *cognitive closure* are the right register and *ignorance* is not. There is no pamphlet.

    —

    V. Four more words, for four different jobs

    Once you are in this territory, several traditions offer terms that overlap without being synonyms. They are worth separating, because each carries a different diagnosis of *why* the thing is inaccessible.

    **Noumenal** (Kant, 1781). The thing as it is in itself, as opposed to the thing as it appears. For Kant the barrier is not our species’ particular equipment but the very structure of experience: space, time, and the categories are the forms our intuition imposes in order to have experience at all, so they cannot also be read off as features of reality behind that experience. Use *noumenal* if you want the inaccessibility to be permanent and principled rather than contingent on our being primates. Note that this is a *stronger* claim than McGinn’s – a smarter alien would not do any better, because it would have its own forms of intuition.

    **Radical alterity** (Levinas, *Totality and Infinity*, 1961, and after him a great deal of anthropology). Otherness that cannot be absorbed into the categories one already holds. Levinas’s point was ethical before it was epistemological: the failure to comprehend the Other is not a defect to be repaired but the very thing that makes the Other a claim on you rather than an object for you. If your interest is in what it would *mean* to encounter an intelligence you cannot model – rather than in whether you could model it – this is the vocabulary.

    **Incommensurable** (Kuhn, 1962). No shared measure. Two frameworks fail to translate not because the dictionary is incomplete but because the terms are individuated differently at the root, so that agreement and disagreement both become ill-defined. Kuhn spent the rest of his career walking back the strongest readings of this, which is itself instructive: near-total incommensurability turns out to be hard to defend, and what survives is a claim about difficulty and loss in translation rather than impossibility.

    **Apophatic** (Pseudo-Dionysius, and the *neti neti* – “not this, not this” – of the Upanishads). Not a description of the ignorance but of the *only available mode of speech* about it: saying exclusively what the thing is not. It is worth knowing that negative theology is the most developed technique humanity has ever built for talking at length about something held to be constitutively beyond conception. If you find yourself writing about an unimaginable mind, you will end up doing apophasis whether or not you have a name for it. Better to know you are doing it.

    —

    VI. Lem, or: the mirror is the only instrument we have

    The novelist of this problem is ‘Stanisław Lem’, and he is more rigorous about it than most philosophers.

    *Solaris* (1961) is usually described as a novel about an alien ocean. It is more accurately a novel about a library – the enormous, futile, century-old discipline of “Solaristics“, with its schools and schisms and competing taxonomies of the ocean’s formations, all of it elaborate and none of it contact. The ocean produces the visitors, apparently from the scientists’ own suppressed memories, and the humans cannot determine whether this is communication, experiment, cruelty, indifference, or a reflex with no more intent than a knee-jerk. The book’s cruelty is that it withholds even the certainty that there is even a question being asked.

    *His Master’s Voice* (1968) does the same thing with a signal from space, which a team of the best minds available decodes into two substances of ambiguous properties and cannot get further, and cannot tell whether their partial success is genuine or a Rorschach blot. Lem’s recurring point, stated bluntly in both books, is that we do not go looking for other minds. We go looking for mirrors, and we call the mirror a discovery. Projection is not a mistake we could avoid with more discipline; it is the only instrument in the kit.

    VII. Now the interesting part: chirality is not the case you think it is…

    Here is where the argument gets useful, because there is a very natural example of “an intelligence I could never comprehend” that turns out, on inspection, to be a completely different kind of problem – and the difference is the whole lesson.

    Suppose you propose two candidates for radically alien minds: a non-carbon intelligence, and a *mirror-image* intelligence, built from the enantiomers of our own biochemistry.

    The first is a genuine Nagelian case, at least potentially. A different substrate might well support a different phenomenology, and we would have no method for checking, and no method for checking whether we had failed to check.

    The second is not. Chirality is a fact about molecular geometry, not about experience. Terrestrial life is homochiral – we use left-handed amino acids and right-handed sugars, an asymmetry that is very likely a frozen accident from a chemistry that could have gone either way. A perfectly mirrored human being would be biochemically incompatible with us in interesting ways: they could not digest our food, our drugs would not fit their receptors, some of our smells would smell like something else to them. But they would not be *phenomenologically* alien. They would have your memories, your emotional range, your sense of humour. They would find your novels moving. There is no hard problem here, only a divergence & supply-chain problem.

    So the mirror case is not a wall. What it is instead – and this is far more interesting – is a **communication** problem, and it has a name and a history.

    VIII. The Ozma problem

    Martin Gardner named it in *The Ambidextrous Universe* (1964), after Project Ozma, Frank Drake’s 1960 attempt to listen for interstellar signals at Green Bank, which Drake had named after the princess in Baum’s Oz books.

    The problem is this. You are in radio contact with a civilisation somewhere distant. You share no objects. You can send only signals – that is, information, not things. You can establish a common vocabulary for numbers, for the elements, for the wavelength of a hydrogen transition, for anything you can define by its structure. Now: **communicate to them which of their hands is the left one.**


    It seems, on first inspection, to be a perfect sealed impossibility. Every method you try fails. You cannot say “the hand on the same side as the heart”, because you cannot establish that their hearts are on the side ours are, and in any case you’d need to define “same side”, which is the thing you are trying to define. You cannot appeal to any geometrical description, because every geometrical description of a left hand is satisfied equally well by a right hand – that is what enantiomorphism *means*. You cannot appeal to astronomy, because they cannot tell whether they are looking at their sky or its mirror image. Every asymmetry you reach for turns out to be a convention or a local accident.

    The Ozma problem looks exactly like cognitive closure. It has the right shape: the failure recurs at every attempt, and for what appears to be a principled reason rather than a contingent one.

    And then it was solved. Not by philosophy – by an experiment.

    In 1956, Tsung-Dao Lee and Chen Ning Yang pointed out that parity conservation, universally assumed, had never actually been tested for the weak nuclear interaction. In early 1957 Chien-Shiung Wu tested it, cooling cobalt-60 nuclei to near absolute zero and aligning their spins in a magnetic field. If parity held, the beta-decay electrons should have come out symmetrically with respect to the spin axis. They did not. They came out preferentially in one direction. The universe, at the level of the weak force, distinguishes left from right.

    Which means the Ozma problem has an answer, and the answer is a recipe: *run this experiment, note which way the electrons go, and now we can both define “left” without ever pointing at anything.*

    Gardner, being honest, flagged the remaining loophole in his own book: if your correspondents are made of antimatter, the recipe reverses, and you have merely traded the handedness ambiguity for a matter/antimatter ambiguity. That gap closed too. In 1964 Cronin and Fitch found CP violation in neutral kaon decay, which gives an absolute, convention-free distinction between matter and antimatter – and therefore, in principle, a fully unambiguous definition of “left” transmissible over a radio link to strangers.

    The wall fell twice, in eight years, both times because someone went and looked at a nucleus.



    IX. So: how do you tell a wall from a puzzle?

    This is the thing worth taking away, (if you have read this far, well done, dear reader) and it is somewhat uncomfortable.

    The Ozma problem *felt* closed. It had every hallmark: repeated failure, a structural-seeming reason for the failure, an air of “you cannot get there from here”. Sophisticated people spent decades treating handedness as the paradigm case of something communicable only ostensively – something you could only teach by pointing, never by describing. They were wrong, and the correction did not come from thinking harder about the problem. It came from an unrelated corner of physics noticing that a symmetry everyone had assumed was simply an untested assumption.

    The bat may be different. Nagel’s argument has survived fifty years of attack in a way that most such arguments do not, and the reason is that it doesn’t rest on our failing to imagine a mechanism; it rests on a structural feature of how imagination works at all. But you should hold even that lightly, because Dennett’s objection stands: from the inside, a genuine wall and a puzzle you haven’t cracked look precisely identical. That is what it is to be nescient rather than ignorant. Nescience does not announce itself. There is no ache in the shape of the missing thing.

    Which means the vocabulary carries a bet, and you should know which bet you are placing when you choose a word.

    Say *ignorant* and you have claimed the thing is knowable and someone has slacked. Say *nescient* and you have claimed the not-knowing is a condition rather than a failure, without yet saying whether it is curable. Say *cognitively closed* and you have made a strong empirical claim about the architecture of a mind – a claim which, historically, has a poor track record. Say *noumenal* and you have made a metaphysical claim that no future experiment can touch, which is either admirable rigour or a way of insulating yourself from being proved wrong, depending on the day. Say *radical alterity* and you have changed the subject from knowledge to ethics, which is sometimes the right thing to do.

    The most defensible position, and the least satisfying, is probably this: we can identify with confidence the cases where our imaginative method breaks down, and we cannot reliably tell which of those breakdowns are permanent. The honest word for the state is *nescient*. The honest posture is apophatic – say what it is not, and keep the account open. And the honest example to keep beside the bat is the Ozma problem, precisely because it is the one that got away.

    —

    Appendix: a working glossary

    | Word | Use it when the emphasis is on… |
    |—|—|
    | **nescient** | not-knowing as a condition rather than a failure; no implication of blame |
    | **incognizant** | plainer register; unawareness without the theological weight |
    | **insensible to** | the failure is sensory rather than intellectual |
    | **anosognosic** | unawareness of one’s own deficit; the recursive blind spot |
    | **scotomised** | a blankness where something intolerable should be; psychoanalytic flavour |
    | **cognitively closed** | a specific mind’s architecture cannot represent the concept (McGinn) |
    | **outside its Umwelt** | the sensorium cannot construct the relevant world (Uexküll) |
    | **noumenal** | inaccessible in principle to any experiencing subject (Kant) |
    | **radical alterity** | otherness that resists absorption into existing categories (Levinas) |
    | **incommensurable** | two frameworks with no shared measure (Kuhn) |
    | **apophatic** | the mode of speech: only negations available |
    | **ostensive** | learnable only by pointing – what Ozma was thought to be, and wasn’t |

    —

    Sources worth reading directly, & citations:

    – Thomas Nagel, “What Is It Like to Be a Bat?”, *The Philosophical Review*, 1974
    – Colin McGinn, “Can We Solve the Mind-Body Problem?”, *Mind*, 1989
    – Jakob von Uexküll, *A Stroll Through the Worlds of Animals and Men*, 1934
    – Martin Gardner, *The Ambidextrous Universe*, 1964
    – Stanisław Lem, *Solaris* (1961) and *His Master’s Voice* (1968)
    – C. S. Wu et al., “Experimental Test of Parity Conservation in Beta Decay”, *Physical Review*, 1957

    Buy me a coffee!
  • The Missing Specification (For Humanity) And The Great Polycrisis Filter.

    The Missing Specification (For Humanity) And The Great Polycrisis Filter.

    Why moral bio-cybernetic/bioenhancement fails at the design document, not the ethics committee.

    admin avatar
    admin
    2,290 words


    Proposals to biologically engineer human moral dispositions – to make people more compassionate, more cooperative, less prone to defection – are usually met with ethical objections. Consent. Autonomy. Authenticity. Value lock-in. The long shadow of eugenics.

    These objections are serious and several of them are quite decisive. But they are also, in a particular sense, premature. They engage the proposal as though the technical programme were ready and the only remaining question were whether we ought to run it. That framing flatters the proposal. It grants it a maturity it does not have.

    There is a more revealing test, and it is the one engineers use on any large proposal before arguing about whether to fund it: **try to write the specification.**

    Not a manifesto. Not a research agenda. A design document – the artefact that lets someone else build the thing, and lets a third party check whether it worked. Every field filled in, every acceptance criterion stated, every assumption made explicit enough to be falsified.

    When you actually attempt this for moral bioenhancement, something instructive happens. The document does not turn out to be *controversial*. It turns out to be **blank**. And the pattern of which fields are blank is more informative than any of the ethical arguments.

    —

    The construct problem: you are not turning up a dial

    The first field in any specification is: *what, precisely, is being modified?*

    “Compassion” is a folk-psychological term. It is not a variable. Before you can build anything you must commit to a decomposition – typically something like an affective component (felt concern at another’s distress), a motivational component (disposition to act at personal cost), a cognitive component (accuracy of your model of the other’s state), a behavioural output (rate and magnitude of costly prosocial acts), and – critically – a **scope function**: to whom does it extend, and how does it decay with social distance?

    That last component is where naive versions of the project die.

    Human prosociality is not a scalar quantity with a gain knob attached. It is a gradient over social distance: steep and high near kin and in-group, falling off quickly with distance, and effectively flat at the level of statistical strangers. This is why we are simultaneously the species that will run into a burning building for a neighbour and the species that can read a famine death toll over breakfast.

    Which means the thing that would actually change outcomes at civilisational scale is not the *amplitude* of caring. It is the **shape of the discount function**. You do not want a higher curve. You want a flatter one.

    Nobody has a genetic or neural handle on the shape. And the best-known intervention that raises amplitude appears to make the shape *worse*. Oxytocin spent about a decade as the “moral molecule” before the picture complicated: alongside well-known affiliative effects, a body of work found it strengthening in-group bonding while in some paradigms increasing out-group hostility or defensive aggression. The replication record across this literature is mixed enough that no single result should be leaned on hard. But the direction of the concern is the point. Raising the gain on a parochial system plausibly yields more effective parochialism – more devoted tribalists, better at their tribalism.

    There is a second constraint that most versions of the proposal omit entirely: **stability under exploitation**. Any modified disposition must be viable in a mixed population that still contains unmodified defectors. A disposition toward unconditional cooperation is not a stable strategy; it is removed from the population – economically, socially, reproductively – by the people who lack it. Making people kinder inside an unchanged incentive landscape does not produce a kinder world. It produces exploitable people.

    So the real target is not “more compassion.” It is something closer to *conditional cooperation, with an unbiased scope function, and with defection-detection and sanctioning capacity fully intact.* That is a much stranger object than the one people imagine. It is also, notably, much closer to what humans already have than to what the proposal would install.

    **Status of this field: unresolved – and not primarily as an empirical matter.** It is a conceptual problem that must be settled before measurement is even meaningful.

    —

    The measurement gate

    This is the field that stops the programme, and it stops it completely.

    Any intervention specification requires a primary endpoint: a measure with construct validity (it measures the target, not social desirability), test–retest reliability sufficient to detect your expected effect size, sensitivity to within-individual change over the intervention window, resistance to demand characteristics (subjects must not be able to score well by inferring what you want), and ecological validity (it must predict field behaviour, not merely laboratory behaviour).

    What actually exists falls into three families, and all three have known, documented problems:

    – **Self-report instruments.** Transparently gameable. Correlate substantially with how respondents wish to be seen.
    – **Economic games** (dictator, ultimatum, public goods, trust). Behaviour in these correlates weakly-to-modestly with real-world prosocial behaviour. The lab-to-field transfer problem here is one of the more uncomfortable open sores in the literature.
    – **Confederate-based laboratory paradigms.** Better ecological validity, but poor scalability and severe single-use problems – you cannot re-run them on the same subject.

    The psychometric reliability of these instruments is nowhere near what would be required to detect the modest effect sizes any realistic intervention would produce.

    The comparison that makes this vivid: a cardiovascular intervention has LDL cholesterol as a validated surrogate endpoint, blood pressure as a second, and hard endpoints – infarction, mortality – ascertained at registry scale with near-perfect reliability.

    **There is no LDL of compassion. There is no mortality-equivalent hard endpoint.** There is nothing you could enter in the “primary outcome measure” field of a trial registration that a competent reviewer would not reject.

    This is not a difficulty. It is a category failure. Without a validated endpoint there is no dose-finding, no efficacy claim, no safety signal, and no way to distinguish a working intervention from a broken one. You would be optimising against a function you cannot evaluate.

    —

    The causal chain has no established arrows

    A specification requires a causal model with every link established and quantified:

    `intervention → molecular change → circuit change → systems-level change → psychological change → behavioural change`

    What exists is a correlational sketch of one link. The empathy/compassion dissociation work – Singer, Klimecki and colleagues – implicates anterior insula and anterior cingulate cortex in empathic distress, and medial orbitofrontal cortex, ventral striatum and affiliation-associated regions in compassion. This is genuinely interesting, and the finding that compassion training increases positive affect while empathy training increases distress and burnout is one of the more useful results in the area.

    But it should be read as suggestive, not as a wiring diagram. It is correlational. It is spatially coarse – a functional imaging voxel contains on the order of a million neurons. Sample sizes are typically small, and this subfield has documented reproducibility problems for precisely this class of finding.

    What is missing is any **causal** manipulation that reliably, durably, and selectively increases the construct. Oxytocin was the strongest candidate and its literature partially collapsed under replication pressure; even the question of whether intranasal administration achieves meaningful central nervous system delivery remains contested. Contemplative training produces real effects, but modest ones, requiring ongoing practice – a behavioural intervention, not a lever a biological one could be built on.

    —

    Genetic architecture: no editable targets

    For any germline proposal, the specification requires target loci with established causal effect, characterised effect sizes, a complete pleiotropy map, and characterised epistasis and gene–environment interaction.

    For prosociality-adjacent traits – agreeableness, self-reported empathy – SNP-based heritability is modest and polygenic scores explain a low single-digit percentage of variance in independent samples. (Treat specific figures as approximate and check current sources; the direction is not in doubt.) The architecture is massively polygenic – thousands of variants of individually negligible effect – heavily pleiotropic, and poorly transferable across ancestries and environments.

    Then there is a recursion problem that is rarely acknowledged: **a genome-wide association study is only as good as its phenotype.** Run against the invalid instruments described above, what you recover is the genetic architecture of *scoring highly on a questionnaire*. That is not the target. It may not even be adjacent to the target.

    Multiplex editing at the scale of thousands of loci, with uncharacterised epistasis and an unmapped pleiotropy burden, is not a hard engineering problem awaiting effort. It sits outside the space of things currently attemptable.

    —

    The safety instrument is inside the system it monitors

    This is the field I find genuinely novel, and it has no analogue in ordinary medicine.

    Post-market drug safety rests on adverse event reporting. Patients notice something has gone wrong and report it. The system assumes the patient’s evaluative faculty is intact and independent of the intervention.

    For a values-modifying intervention, that assumption fails by construction. The adverse event class *includes changes to the faculty that generates the report*. If the intervention shifts what a person values, then self-report is compromised as a safety instrument in exactly the failure mode you most need to detect. A population successfully modified toward a particular specification of compassion may no longer contain anyone disposed to recognise the modification as a harm.

    You would therefore need an external, non-self-report harm criterion, specified in advance, held by someone outside the modified population. Nobody has (yet…) proposed a workable one.

    Note that this is not a philosophical objection dressed up as an engineering one. It is a missing section in the safety file. And it generalises: irreversibility is not merely one cost to be weighed against others, because it removes the mechanism by which anything gets weighed later. Ordinary bad policy is reversed because those harmed by it object. This is the one class of intervention that can eliminate the constituency capable of identifying the error.

    —

    What the blanks tell us:

    Lay the fields out and the completion state is stark. Delivery technology has partial content and active research behind it. Nearly everything else is empty – and the two most upstream fields, construct definition and outcome measurement, are empty in ways that no amount of funding or intelligence resolves from a single location. They are filled by cohorts, instruments, longitudinal data, and decades.

    Two things follow.

    **First: the ethical objections and the technical emptiness point the same way.** This is worth noticing rather than treating as coincidence. The consent problem, the value lock-in problem, and the pharmacovigilance problem are the same structural fact appearing in three registers – an intervention that alters the evaluator cannot be evaluated by the altered. That the technical specification is blank at precisely the points where the ethics is most troubling is not an accident. It reflects that we do not understand the object well enough to specify it *or* to consent to it.

    **Second: the causal premise is probably wrong anyway.** The proposal assumes that destructive collective behaviour is primarily a psychological trait being expressed. But humans are already extraordinarily cooperative by primate standards – we punish unfairness at cost to ourselves, we cooperate with strangers we will never meet again. What competitive systems do is *select* for defection at the level of firms and institutions, largely independent of the dispositions of the people inside them.

    The evidence for this is not subtle. When emergency conditions suspend normal procurement controls – competitive tender, due diligence, published contracts, audit trails – fraud losses jump by orders of magnitude. Same population, same dispositions, different controls. Removing the checking is what changes the behaviour.

    None of which means dispositions are irrelevant. Some people are cruel, some enjoy it, and the variance is real. But what institutions and norms determine is how much *scope* those dispositions get – whether cruelty is costly or licensed, marginal or ambient. Both halves are true, and the tractable half is the second one. Ostrom’s work on commons governance showed groups solving defection problems through monitoring, graduated sanctions, and local rule-making, with nobody’s psyche altered at all.

    And where genuinely catastrophic risk is the concern, it concentrates in a very small number of people with access to weapons systems, engineered pathogens, or critical infrastructure. Screening and constraining that population is orders of magnitude more tractable than modifying a species. It has real problems – who screens the screeners, capture risk – but they are the ordinary problems of institutional design rather than the irreversible rewriting of a lineage.

    —

    Where the real problem is

    If you take the specification exercise seriously, the interesting frontier turns out not to be where the proposal points.

    The measurement field is a live, unsolved, genuinely deep problem: how to construct a valid and reliable instrument for a latent construct that resists direct observation, where the act of measurement perturbs the thing measured and the subject has incentive to game the readout.

    That is a problem in **measurement theory** more than in biology. Psychology has been notably bad at it – partly because the field’s training does not emphasise what a physicist’s or metrologist’s does: error propagation, calibration, sensitivity limits, distinguishing signal from instrument artefact, and knowing when your resolution cannot support your claim.

    Solving it would be valuable regardless of what anyone concluded about enhancement. It is upstream of clinical trials in psychiatry, of policy evaluation, of most of behavioural science. It is where someone with quantitative training would have a genuine edge.

    The specification exercise is not, in the end, an argument for despair about the underlying goal. It is a redirection. The document is blank at the top, and the top is where the work is.

    So let that work begin.



    *Further reading and citations: Persson & Savulescu, ***Unfit for the Future*** (the strongest case for the affirmative); John Harris’s reply on the freedom to fall; Paul Bloom, ***Against Empathy***; Elinor Ostrom, ***Governing the Commons***; Singer & Klimecki on the empathy/compassion dissociation; Habermas, ***The Future of Human Nature***.*

    Buy me a coffee!
  • The Quantum Crucible: How Magnetic Rocks, Spin-Polarised Electrons, and Deep-Sea Vents Solved the Origin of Life’s Greatest Paradox

    The Quantum Crucible: How Magnetic Rocks, Spin-Polarised Electrons, and Deep-Sea Vents Solved the Origin of Life’s Greatest Paradox

    The chiral CISS engine of life…

    admin avatar
    admin
    7–10 minutes

    Picture the Earth four billion years ago. The Hadean eon is in full swing. The atmosphere is a choking haze of carbon dioxide and methane, and a vast, acidic, iron-rich ocean covers the globe. Deep at the bottom of this abyssal sea, far removed from the violent ultraviolet radiation of the young Sun, alkaline hydrothermal vents spew scorching, mineral-laden fluids into the frigid waters.

    For decades, astrobiologists and biochemists have suspected that these deep-sea chimneys—specifically, the towering “black smokers” and alkaline vents—acted as the chemical wombs of the first living cells. They provided the heat, the chemical gradients, and the building blocks necessary to kickstart primitive metabolism.

    But there was a glaring, seemingly insurmountable problem with the deep-sea hypothesis: The Chirality Paradox. If life started in the pitch black of the ocean floor, how did it achieve molecular handedness without light? Recently, a revolutionary quantum mechanical concept known as the Chiral Induced Spin Selectivity (CISS) effect has provided a staggering answer, fundamentally rewriting the state of academic opinion on abiogenesis.

    The Mirror-Image Menace: The Problem of Homochirality

    Before we can build a cell, we must confront a fundamental geometrical quirk of the universe: chirality.

    Many of life’s most essential molecules—like amino acids (the building blocks of proteins) and sugars (the backbone of DNA and RNA)—are chiral. This means they exist in two distinct, non-superimposable mirror-image forms, much like your left and right hands. We call these forms enantiomers (L- for left-handed, D- for right-handed).

    If you synthesize these molecules in a standard, sterile laboratory without biological enzymes guiding the process, the laws of thermodynamics dictate that you will produce a racemic mixture—an exact 50/50 split of left- and right-handed molecules.

    Yet, all known life on Earth is strictly homochiral. Our biological machinery exclusively uses L-amino acids to build proteins and D-sugars to build nucleic acids. If you introduce a right-handed amino acid into a growing protein chain, the entire structure folds incorrectly, and the biological machine breaks down.

    Therefore, before the first true cell could ever arise, nature had to find a way to take a chaotic 50/50 chemical soup and ruthlessly filter out one of the enantiomers.

    The Death of the Sunbeam Hypothesis

    Historically, the leading theory to explain this initial filtering relied on Circularly Polarised Light (CPL). It was theorized that CPL—perhaps generated by neutron stars irradiating molecular clouds in space, or by intense UV light hitting shallow tidal pools on early Earth—acted as a cosmic asymmetric destroyer. The light would preferentially break down one handedness of a molecule while leaving the mirror image intact.

    This theory works beautifully for shallow pools (“warm little ponds”). But it is fatal to the hydrothermal vent hypothesis. At the bottom of the ocean, under kilometers of water, there is absolutely zero ultraviolet light. If the CPL hypothesis was the only way to achieve homochirality, life could not have started at the vents.

    The origin of life research community was caught in a stalemate: the vents had the perfect thermodynamics for early metabolism, but no apparent mechanism to sort molecules by their handedness.

    The Quantum Saviour: Chiral Induced Spin Selectivity (CISS)

    The paradigm shifted with the discovery of the Chiral Induced Spin Selectivity (CISS) effect, a quantum phenomenon that proves you don’t need spin-polarized light to sort molecules, if you have spin-polarized electrons.

    The CISS effect dictates a profound relationship between a molecule’s physical 3D shape (its chirality) and the quantum spin of electrons moving through it. When an electron attempts to travel through a chiral molecule, the molecule acts as an aggressive quantum filter. Depending on the molecule’s handedness, it will only permit electrons of a specific spin state (spin-up or spin-down) to pass efficiently.

    This works in reverse, generating a phenomenon called electrochirogenesis. If you provide a surface saturated with electrons that are all spinning in the exact same direction, and you wash a racemic 50/50 mixture of prebiotic molecules over it, the surface will strongly bind and crystallize one handedness while violently repelling the other.

    At a hydrothermal vent, where do we find a massive, continuous supply of spin-polarized electrons? Magnetic rocks.

    As the alkaline fluids of the vent mix with the acidic, iron-rich Hadean ocean, they precipitate trillions of nanoparticles of magnetite ($Fe_3O_4$) and greigite ($Fe_3S_4$). These are natural ferromagnetic minerals. Thanks to the early Earth’s geodynamo (its magnetic field), as these minerals crystallized out of the scorching vent fluids, they underwent Thermoremanent and Chemical Remanent Magnetization. They locked the Earth’s magnetic field into their very atomic lattice, becoming permanent magnets.

    The towering walls of these hydrothermal chimneys became vast, uniform sheets of spin-polarized electrons. As the chaotic, racemic soup of early amino acids and RNA precursors (like ribose-aminooxazoline) washed over the rock face, the quantum spin-exchange interactions allowed only molecules of a single handedness to safely dock, crystallize, and concentrate. The rocks themselves were the sorting machines.

    Experimental Validation: Proving it in the Lab

    This is not just theoretical physics; it is heavily tested laboratory reality. Teams led by researchers like Ron Naaman, Yossi Paltiel, and Dimitar Sasselov have successfully recreated this deep-sea quantum sorting engine.

    The experimental setup is breathtakingly elegant:

    1. The Substrate: A layer of magnetite is placed over a strong permanent magnet (e.g., 0.42 Tesla), forcing all the electron spins on the surface into a uniform alignment (mimicking the permanently magnetized Hadean vent rocks).
    2. The Soup: A supersaturated, completely racemic 50/50 solution of an amino acid or RNA precursor is introduced.
    3. The Crystallisation: As the solution cools, the molecules hit the surface. Because of the CISS effect, molecules with a handedness that opposes the surface spin are repelled (due to high-energy, parallel triplet-like states). Molecules with the correct handedness form low-energy, antiparallel singlet-like bonds and crystallize.
    4. The Result: Researchers harvest the resulting crystals and analyze them using Circular Dichroism (CD) spectroscopy and magnetic conductive Atomic Force Microscopy (mc-AFM). The results consistently show near-total enantiomeric purity. Flip the magnetic field, and the surface selectively crystallizes the exact opposite handedness.

    Enclosing the Miracle: Lipid Vesicles

    Achieving homochirality of building blocks is only part of the puzzle. Life requires a boundary—a cell membrane. Hydrothermal vents continuously synthesize simple amphiphilic molecules (like fatty acids), which naturally self-assemble in water into lipid vesicles or protocells.

    However, the thermodynamics of self-assembly present a stark warning: a membrane built from a racemic, mixed-handedness jumble of lipids is structurally chaotic, highly permeable, and dangerously leaky. It cannot hold an electrochemical gradient.

    But when lipid vesicles self-assemble on or near these spin-polarized magnetic surfaces, the CISS effect biases the chirality of the lipids themselves. The resulting homochiral lipid bilayers pack tightly and securely. They become stable enough to capture the vent’s natural proton gradients, providing the necessary battery power to fuel the very first primitive, pre-enzymatic metabolisms.

    The Cosmic Perspective: How Rare is the Quantum Crucible?

    If this highly specific, highly intricate geochemical and quantum mechanical machine is the true origin of life, how likely is it that this process is happening elsewhere in the Milky Way?

    To calculate this, we use a Fermi estimation approach to establish $N_{CISS}$: the number of planets currently capable of operating a magnetic hydrothermal vent abiogenesis engine. The governing equation is:$$N_{CISS} = N_{hab} \times f_{water} \times f_{dynamo} \times f_{vent} \times f_{chemistry}$$

    Where:

    • $N_{hab}$ = Number of rocky planets in the habitable zone.
    • $f_{water}$ = Fraction that retain optimal surface water oceans.
    • $f_{dynamo}$ = Fraction possessing an active internal magnetic field.
    • $f_{vent}$ = Fraction featuring active tectonic/hydrothermal recycling.
    • $f_{chemistry}$ = Fraction possessing the specific iron-rich, reducing atmospheric conditions for magnetite precipitation.

    The Optimistic View

    If we assume 4 billion Sun-like stars in our galaxy, optimistic Kepler telescope data suggests there could be 2 billion habitable-zone planets ($N_{hab}$). If we assign generous probabilities to standard planetary geology ($f_{water} = 0.20$, $f_{dynamo} = 0.25$, $f_{vent} = 0.20$, $f_{chemistry} = 0.50$):$$N_{CISS} = 2,000,000,000 \times 0.20 \times 0.25 \times 0.20 \times 0.50 = 10,000,000$$

    Under this view, there are 10 million planets in the Milky Way acting as vast, active prebiotic laboratories.

    The Pessimistic View (The Rare Earth Hypothesis)

    However, planetary geophysics is often brutally unforgiving. What if Earth is a statistical freak?

    • What if a planet’s oceans are usually too deep, forming an impenetrable barrier of high-pressure Ice VII at the ocean floor, suffocating the rock?
    • What if a geodynamo requires a highly improbable, cataclysmic Mars-sized planetary impact (like the one that formed our Moon) to keep the core spinning?
    • What if the crust is almost always a stagnant, unmoving lid like Venus, preventing tectonic vents entirely?

    If we take the strictest lower bounds for habitable planets ($N_{hab} = 280,000,000$) and assign a brutal 1% (0.01) probability to the four geological filters:$$N_{CISS} = 280,000,000 \times (0.01 \times 0.01 \times 0.01 \times 0.01)$$$$N_{CISS} = 280,000,000 \times 10^{-8}$$$$N_{CISS} = 2.8$$

    Under the most mathematically pessimistic, geologically rigorous interpretation possible, there are fewer than three planets in the entire Milky Way galaxy equipped to run this ‘origin-of-life engine’.

    Conclusion

    Yet, even if the pessimistic math holds true, the observable universe contains upwards of two trillion galaxies. Even at a rate of 3 planets per galaxy, there are still trillions of worlds scattered in the dark, where deep-sea vents and magnetic rocks are quietly, inevitably, sorting the chaos of chemistry into the architecture of life.

    The paradox has been largely solved. The darkness was not an obstacle to life; armed with quantum spin, the darkness was the very mechanism that built it.

  • A Secular Humanist’s Take on Trump’s Mt. Rushmore Sermon

    A Secular Humanist’s Take on Trump’s Mt. Rushmore Sermon

    The Menace That Wasn’t: A Secular Humanist, Green Party Take on Trump’s Rushmore Sermon

    7–11 minutes
    1,669 words
    admin avatar
    admin


    There’s a particular genre of American political theatre that never quite goes out of style: the mountain, the flag, the borrowed marble faces of dead presidents, and a man at a podium warning you that the enemy is already inside the gates. On July 3, 2026, on the eve of the country’s 250th birthday, Donald Trump stood in front of Mount Rushmore and reached for the oldest prop in that theatre’s closet – communism – and dusted it off for an audience that, statistically, has never lived under it, never studied it seriously, and would be hard-pressed to name a single line from the *Communist Manifesto*. That’s not an accident. The speech wasn’t really about communism. It was about who gets to define “American,” who gets cast as a threat to it, and who gets to stand at the centre of the frame while doing the casting.

    I’m writing this as a secular humanist and a GPEW & UK Green Party Member – which means I come to this with two specific moral and political objections. One is to political rhetoric that fuses patriotism to a particular ‘god'(s), and treats disbelief, doubt, or a different faith as a moral defect. The other is to language that smears environmental and economic reform as a totalitarian plot, because that language has spent forty years being used against exactly the kind of politics I care about.

    The Return of an Old Formula

    Red-baiting has a rhythm to it, and Trump’s speech followed the rhythm precisely: name an external ideology, claim it has infiltrated the homeland through newcomers, strip it of any economic or historical specificity, and then attach to it every imaginable evil – theft, godlessness, lawlessness, murder – until “communism” stops meaning an actual set of economic ideas and starts meaning simply *the enemy*. It is the McCarthy playbook with better production values. The tell is in the vagueness. Nowhere in the speech is there an actual communist policy platform, an actual party programme, an actual piece of legislation. There’s only atmosphere: menace, resurgence, mass control. That vagueness isn’t a flaw in the rhetoric – it’s the whole function of it. A vague enemy can absorb almost anyone the speaker wants it to: a socialist mayor, an immigrant, a public school teacher, a climate scientist, a union organiser, a Green.

    And that absorption is precisely why this matters to those of us who spend our political lives arguing for climate policy, wealth redistribution, and structural reform of capitalism. We have watched this same conflation trick used against basic proposals – public healthcare, a carbon tax, tenant protections – for decades. “Socialism” and “communism” get used almost interchangeably in speeches like this one, which is either ignorance of or indifference to the fact that Scandinavian social democracy, American-style progressive taxation, eco-socialism, and Soviet-style single-party state ownership are not the same thing, do not share a history, and do not share an outcome. Collapsing them into one bogeyman isn’t an argument. It’s a way of making argument unnecessary.

    “They Don’t Love ‘God(s)’…” – The Theological Loyalty Test

    The line that should trouble anyone who takes secularism seriously is the claim that the supposed communist newcomers “don’t love God,” “don’t want God,” and have no interest in religion – as though these were self-evidently disqualifying traits in an American, and as though love of ‘God(s)’, (fantasies and delusions) were a reliable predictor of respect for the rule of law. This is not a description of communism. Historical communist states were genuinely, often violently, hostile to organised religion, and that history is real and worth reckoning with. But the rhetorical move here isn’t a critique of state atheism as state policy – it’s an implication that irreligion itself, present-tense, in an ordinary citizen or newcomer, is evidence of moral rot.

    That should unsettle a lot more people than it seems to. Non-religious Americans are one of the fastest-growing demographic categories in the country. Only a portion of them are politically left of centre, and essentially none of them are Soviet apparatchiks. To imply that lack of religious devotion correlates with lawlessness and “mass murder” isn’t a factual claim about political ideology – it’s a loyalty test with a cross on it, dressed up as a warning about Marx. A secular humanist ethic doesn’t need ‘God'(s )to generate a respect for law, human dignity, or “your God-given rights,” to use the speech’s own phrase; it grounds those things in the observable, arguable, revisable project of human reason and empathy instead. A speech that treats a categorical love of God as the load-bearing wall of civic virtue is not just historically sloppy about communism – it is quietly, casually exclusionary towards tens of millions of its own citizens.

    Who Gets to Be a “Newcomer”

    The line linking the “communist menace” explicitly to “newcomers to our country” deserves its own paragraph, because it is doing quiet, ugly work. Immigration anxiety and communism panic have been braided together in American rhetoric since at least the first Red Scare of 1919-20, when Attorney General Palmer’s raids targeted immigrant labour organisers as much as they targeted any coherent Bolshevik threat. The pattern repeats here: an ideological menace is described, and then it is given a face, and the face belongs to the person who arrived more recently than the speaker’s own ancestors did. It’s worth noting, as an aside that the speech itself doesn’t, that nearly every American is descended from a newcomer at some point, and that the timeline of “how long ago you got here” has never been a meaningful predictor of political ideology.

    The Irony of “Mass Control”

    Here’s where the Green in me can’t help but notice the mirror. The speech describes communism as “an ideology of mass theft, mass control, mass lies, and mass murder” – and follows it, in the same set of remarks, with a call to eliminate the Senate filibuster and pass a specific piece of legislation so that a single party need “not lose an election for 100 years.” Whatever one thinks of the filibuster as a procedural tool, a call for a hundred years of uncontested one-party rule, delivered in the same breath as a warning about “mass control,” is not a minor rhetorical stumble. It’s the thing the speech claims to be warning against, worn as a lapel pin.

    This is the pattern that Greens and civil libertarians alike should recognise regardless of who is doing it: concentrated, unaccountable power is the actual danger, whether it wears a hammer-and-sickle or a flag pin. A one-party state is a one-party state whether its founding myth is proletarian revolution or American exceptionalism. The environmental movement in particular has learned this lesson the hard way – ecological collapse has been accelerated as readily by unaccountable state bureaucracies (the Soviet Union’s environmental record is genuinely catastrophic, from the Aral Sea to Chernobyl) as by unaccountable corporate power in market democracies. The threat was never the label. It was the absence of checks.

    Stolen Land, Real History

    The speech also takes a swing at people “who tell our children that we live on stolen land or that our heroes were oppressors,” framing this as a communist lie about heritage. This is worth pausing on simply because it’s checkable. The displacement of Native nations from their land through treaty violation, forced removal, and military conquest is not a Marxist interpretation of American history; it’s the documented factual record, taught in university history departments with no particular ideological bent and available in the treaties themselves. Refusing to look at that record doesn’t make the country’s founding more secure – it just makes the founding myth more brittle, because myths that can’t survive contact with evidence eventually break all at once instead of bending gradually. A humanist approach to history says the honest version, oppressors and all, produces a more resilient civic identity than an insistence on unblemished virtue. You cannot build a durable patriotism on a historical record you’re not allowed to examine.

    None of this is to say communism as historically practised is above criticism – quite the opposite. The gulag system, the Holodomor, the Cultural Revolution, and the killing fields are real, and their death tolls are not political footballs; they are among the worst human-caused catastrophes on record and deserve unflinching moral reckoning, not what-about-ism from the left and not cartoonish flattening from the right. But precisely because that history is so serious, it deserves to be invoked accurately – as an argument about the specific dangers of one-party rule, command economies, and the suppression of dissent – rather than as a floating epithet applied to socialists, environmentalists, secularists, and immigrants alike because they are politically inconvenient this news cycle.

    A Green politics that argues for public transit, a carbon price, universal healthcare, and stronger labour protections is not a request for a politburo. It is, if anything, a bet that democratic institutions – the same institutions this speech claims to be defending – are strong enough to legislate ambitious change without collapsing into the very authoritarianism the speech warns against. Painting that politics with the same brush as Stalinism isn’t vigilance. It’s an [ironically] bad faith way of making sure ambitious reform never gets argued on its merits at all.

    The Actual Choice

    The speech ends on a binary: you can be a communist, or you can be a patriot, and you cannot be both. It’s a tidy line, and tidy lines are usually where the thinking stopped. The real choice on offer in American politics has never been between Marxist revolution and the status quo. It’s a choice among a wide field of positions on how much the public sector should do, how power and wealth should be distributed, how religious pluralism should be protected, and how honestly a country tells its own history. Collapsing that whole field into a loyalty oath – love ‘God(s)’, distrust newcomers, don’t question the founding myth, or be filed under “menace” – isn’t patriotism. It’s a disingenuous shortcut around the argument, aimed at anyone who might otherwise have made one.

    Buy me a coffee!
  • Mirror Life. Mirror Dangers.

    Mirror Life. Mirror Dangers.

    The Mirror World That Could Kill Us: Inside the Race to Stop Synthetic Mirror Life.

    12–18 minutes
    admin avatar
    admin


    In December 2024, something unusual happened in the normally collegial world of synthetic biology. Thirty-eight scientists – including a Nobel laureate, a co-creator of the first synthetic cell, and several of the most influential figures in the field – published a paper in *Science* asking the world to **not** build something. Alongside it they released a technical report running to nearly 300 pages. Their message was blunt: a category of artificial organism that does not yet exist, and cannot yet be built, may be so dangerous that humanity should decide *now*, before the capability arrives, never to create it.

    The organism in question is “mirror life.” And the reason it frightens the people best equipped to understand it is not that it would be a cleverer pathogen than anything in nature. It is that it would be playing a completely different game – one our immune systems, our ecosystems, and four billion years of evolution have never encountered.

    This post explains what mirror life is, why the alarm is so unusual, what specific risks have scientists worried, what the sceptics say in response, and how the world is now scrambling to govern a technology that may still be decades away.

    —

    First, the strange physics of “handedness”

    To understand mirror life, you have to start with one of the deepest and oddest facts about biology: life is one-handed.

    The property is called **chirality**, from the Greek word for hand. A chiral object cannot be superimposed on its own mirror image. Your left and right hands are the classic example – they are mirror images, but no matter how you rotate one, you can’t lay it perfectly over the other. Many of the molecules that make up living things are chiral in exactly this way. They come in two mirror-image versions, conventionally labelled “left-handed” (L) and “right-handed” (D).

    Here is the remarkable part. Although both versions are chemically possible and equally stable, **all known life uses only one orientation for each class of molecule.** Proteins are built from left-handed amino acids. DNA and RNA use right-handed sugars and twist in a consistent direction. This uniformity is called **homochirality**, and it is universal – bacteria, fungi, redwoods, blue whales, and humans all share it. Louis Pasteur discovered molecular chirality in 1847, and the consistency of life’s handedness has been one of biology’s quiet constants ever since.

    Why life settled on one set of orientations rather than the other is still debated. But *that* it did is not in question, and it has a profound consequence: biology is built to recognise and process molecules of a specific handedness. An enzyme shaped to grip a left-handed amino acid will not grip its mirror image, just as a left glove won’t fit a right hand. Handedness is the lock-and-key logic running underneath nearly everything living things do.

    —

    What “mirror life” actually means

    A mirror organism would be a living cell in which **every chiral molecule is flipped to its opposite orientation.** Mirror DNA, mirror RNA, mirror proteins, mirror sugars, mirror lipids – a complete inversion of the molecular handedness of an ordinary cell.

    Crucially, a mirror bacterium would not be a genetically engineered version of an existing microbe. It could not arise through mutation or evolution from anything alive today, because you cannot get there one step at a time – a half-mirrored cell wouldn’t function. It would have to be constructed from the ground up, molecule by molecule, as a built artefact. The scientists behind the *Science* paper describe this as a feat of biological engineering far beyond anything yet accomplished.

    And that is the point of the warning. Because mirror life cannot evolve naturally, it does not exist anywhere on Earth, which means nothing in our biosphere has ever had to defend against it.

    —

    Why this isn’t science fiction any more.

    For decades, “mirror cells” lived in the realm of speculation. Geneticist George Church mused about mirror humans in his 2012 book *Regenesis* – beings that might be immune to all ordinary viruses precisely because no virus would recognise them.

    What changed is that the building blocks stopped being hypothetical. Over the past two decades, chemists have synthesised mirror-image proteins, mirror-image DNA and RNA, and a working mirror-image version of an enzyme that copies genetic material. Researchers have produced a mirror-image polymerase and demonstrated mirror-image transcription. In 2019, the U.S. National Science Foundation awarded a roughly $4 million grant to a team explicitly aiming to design and build synthetic mirror cells with all key molecules in their non-natural orientation.

    No one is close to a complete, self-replicating mirror bacterium. The hardest single component – a functioning mirror-image ribosome, the molecular machine that manufactures proteins – remains, by the assessment of researchers in the field, the most formidable obstacle, and is itself years away. The consensus estimate is that a full mirror organism is likely **decades** off, if it is achievable at all.

    But the trajectory is what matters. The relevant fears of past decades – particle colliders spawning black holes, nanotech “grey goo” – concerned exotic or highly theoretical physics. Mirror molecules are neither exotic nor theoretical. They are real, they have been made in laboratories, and the enabling technologies are improving steadily. The line between speculation and feasibility is being crossed in increments, which is exactly why scientists wanted to start the governance conversation before the capability fully matures rather than after.

    —

    The central danger: a pathogen our immune system can’t see

    The single most serious concern is **immune evasion**, and it follows directly from the logic of handedness.

    The human immune system, like that of virtually all complex organisms, detects invaders by recognising the shapes of their molecules. A large part of innate immunity works by spotting characteristic microbial molecules – sometimes called microbe-associated molecular patterns – using dedicated receptors such as Toll-like receptors and others. The catch, as the technical report emphasises, is that **almost all of these recognition systems are themselves chiral.** They are tuned to molecules of a specific handedness.

    A mirror bacterium would present mirror-image versions of those molecules. The immune system’s pattern-recognition machinery might simply fail to register them – the lock would not accept the flipped key. The same problem extends to adaptive immunity and even to the enzymes our bodies use to digest and break down bacteria, many of which are also handedness-specific. The worry is not that a mirror pathogen would be especially aggressive, but that it could spread while remaining substantially **invisible** to defences that have protected animals for hundreds of millions of years.

    This is the qualitative difference that sets mirror life apart from ordinary engineered pathogens. A conventional dangerous microbe is still a microbe our biology recognises as foreign. A mirror microbe might not trigger the alarm at all.



    It wouldn’t stop at humans…

    The same reasoning extends across the living world, which is what elevates mirror life from a public-health concern to a potential ecological one.

    Plants and animals rely on chirally-specific immune mechanisms too. A mirror bacterium that could draw nutrients from the environment might be able to infect or colonise a wide range of hosts – crops, livestock, wildlife – without provoking effective defences in any of them. Because no existing organism has co-evolved with mirror biology, the usual checks that keep bacterial populations in balance might not apply.

    Consider how ordinary bacteria are kept in check in nature. They are eaten by predators such as protozoa, and they are killed by viruses called bacteriophages, which are astronomically abundant and which constantly prune microbial populations. Both predation and phage attack typically depend on recognising molecular features of the target – features that, in a mirror organism, would be flipped. A mirror bacterium might be poorly recognised by natural predators and effectively immune to the phages that would otherwise control it. Released into the environment, such an organism could potentially persist and spread in soil, water, and living hosts with few of the natural brakes that constrain ordinary microbes.

    The technical report is careful here: it does not claim certainty that a mirror organism would be an unstoppable superbug. Mirror life would also face real disadvantages – it could only consume nutrients that happen to be present in a usable mirror form, which might limit where it could grow. But the authors argue the plausible worst cases are severe enough, and irreversible enough, that they cannot be waved away. An environmental release could not be recalled.

    —

    Why “just keep it contained” may not be enough

    A natural response is to say: fine, build it if we must, but lock it down. The trouble is that every proposed safeguard has a known failure mode.

    One idea is **synthetic auxotrophy** – engineering the organism to depend on an artificial nutrient that exists only in the lab, so it dies the moment it escapes. The *Science* authors acknowledge this could reduce risk, but note that organisms evolve, and engineered dependencies can be lost through mutation or defeated by human error. Building in *multiple* such dependencies lowers the odds of escape further, but does not eliminate them.

    The other line of defence is **physical containment** – high-security laboratories of the kind used for the most dangerous known pathogens. But the historical record is sobering: laboratory accidents and accidental releases happen with some regularity, even in the most secure facilities, because human error is irreducible. For a self-replicating organism that could spread through the environment and resist natural controls, a single containment failure could be catastrophic and permanent. The asymmetry between the difficulty of perfect containment and the severity of a single failure is precisely what makes many researchers conclude the organism is better off never built.

    —

    The case for caution isn’t unanimous – and that matters

    Responsible coverage of this topic has to take the sceptics seriously, because the scientific debate is genuinely live and the alarm, however well-credentialed, rests partly on projections rather than observations.

    Some researchers argue the *Science* commentary painted too dire a picture. David Perrin, a synthetic chemist at the University of British Columbia, has contended that the headline framing overstated the danger relative to the more measured technical report, and that the discussion gave too little weight to the immune system’s genuine capacity to respond, to the complex biology of what actually makes a pathogen virulent, and to the large pharmacological toolkit that could be brought to bear against a mirror infection. A pathogen, on this view, needs far more than immune invisibility to become a successful disease-causing agent; virulence is hard-won, and a from-scratch organism would likely be fragile.

    There is also pushback against the idea of restricting *basic research* prematurely. Ting Zhu, whose laboratory has pioneered mirror-image molecular biology, has said publicly that he has never sought to build a living mirror cell and remains far from the components that would make one possible. In a 2025 opinion piece he acknowledged that fully realised mirror organisms could be harmful while welcoming open debate – and cautioned against halting foundational science based on a distant and uncertain threat. Even among those who agree mirror *organisms* would be dangerous, opinions differ sharply on where exactly to draw the line, and whether work on individual mirror components (like a mirror ribosome) should itself be off-limits.

    This tension – catastrophic potential versus speculative timeline, precaution versus open inquiry – is the real heart of the policy problem. The risks are projected from first principles, which is unusually strong as scientific reasoning goes (mirror molecules have identical chemistry to their natural counterparts, just reversed geometry, so a great deal can be inferred without building anything). But “we can reason it out in advance” is not the same as “we have seen it happen,” and reasonable scientists weigh that gap differently.

    —

    A distinction that the whole debate turns on: molecules vs. organisms

    If there is one point that gets lost in alarming headlines, it is this: **mirror molecules and mirror organisms are not the same thing, and the concern is overwhelmingly about the latter.**

    Mirror-image *molecules* are not just harmless – they are genuinely promising. Because the body’s degradation machinery is handedness-specific, a mirror-image drug can resist being broken down and may remain stable and active far longer than its natural counterpart. Researchers are pursuing mirror-image proteins, nucleic acids, and peptides as candidate therapies for metabolic disease, inflammation, cancer, and infection, and as durable tools for diagnostics. Mirror antimicrobial peptides are being explored as a weapon against antibiotic resistance, and mirror enzymes have potential industrial uses such as breaking down plastics. At least one mirror-chemistry-based drug is already approved and in clinical use.

    The near-consensus that has emerged is therefore narrower and more workable than “ban mirror biology.” It is roughly: encourage research on mirror molecules for their real benefits, while drawing a firm line against research aimed at assembling a complete, self-replicating mirror organism. Getting that boundary right – distinguishing genuinely safe component research from “dual-use research of concern” that lowers the barrier to building a full organism – is the technical crux that governance bodies are now wrestling with.

    —

    The world’s response: dialogues, not decrees (yet)

    What makes this episode historically interesting is that scientists raised the alarm about their own field, pre-emptively, before any dangerous capability existed. The closest precedent is the 1975 **Asilomar Conference**, where biologists paused to set safety norms for the then-new technology of recombinant DNA – norms that shaped decades of biosafety regulation. The mirror-life community has explicitly invoked that model, and 2025 marked Asilomar’s 50th anniversary, lending the comparison extra resonance.

    Rather than push immediately for binding law, the original *Science* working group launched the **Mirror Biology Dialogues** effort to convene scientists, policymakers, industry, and the public through a series of international meetings. The first was held at the Institut Pasteur in Paris in June 2025 – a fitting venue, given Pasteur’s own discovery of chirality there. Further meetings followed at the University of Manchester in September 2025 and the National University of Singapore, with the explicit aim of clarifying red lines, articulating principles for responsible research, and producing governance recommendations.

    Governments and international bodies are now engaging in parallel. The U.S. Congressional Research Service has examined whether existing biosafety oversight is adequate, and whether a moratorium on creating mirror life might buy time for deliberation. The UK government convened a roundtable in early 2025; a notable conclusion was that while officials regard the risks as real, some felt the evidence base was not yet sufficient for decisive regulation – prompting careful work to identify which knowledge gaps can be safely filled *without* accelerating the very capability everyone wants to prevent. The European Union has taken up mirror biology in consultations informing its biotech policy, and a United Nations scientific advisory brief has weighed how to translate emerging agreement into actual governance. Proposals on the table include a global moratorium on building self-replicating mirror organisms and an advisory committee under the WHO or UN to classify and oversee research by risk level.

    No binding international ban exists today. What exists is a fast-coalescing norm – that research directly aimed at creating mirror life should not be funded or pursued – and an unusually proactive attempt to harden that norm into governance before, rather than after, the technology arrives.

    —

    Why this story is worth watching

    Mirror life is, for now, a danger that lives in projections and laboratories rather than in the world. A complete mirror organism may be decades away, and might prove harder to build than anyone expects. It is entirely possible the worst scenarios never materialise.

    But the reason serious people are treating it seriously comes down to a particular combination of features that few other risks share. The threat is **inferable in advance**, because the chemistry is well understood. It is potentially **irreversible**, because a self-replicating organism released into the environment cannot be recalled. It would exploit a vulnerability that is **universal and ancient**, because every living thing shares the same molecular handedness and none has ever faced its mirror. And the window to decide how to handle it is **open now**, while the capability is still incomplete.

    That last point is the whole argument. With most catastrophic technologies, society reacts after the fact – after the accident, the release, the proof of harm. Mirror life offers a rare chance to make the decision the other way around: to look clearly at a thing that does not yet exist, judge it too dangerous to create, and choose, deliberately and in advance, not to build it. Whether the world takes that chance is a question still very much being written.

    —

    *This post is a general-audience explainer drawing on the December 2024 *Science* Policy Forum article “Confronting risks of mirror life” and its accompanying technical report, along with subsequent scientific commentary and policy discussion through early 2026. It is intended to inform public understanding of the debate and deliberately does not address methods for creating mirror organisms – an omission shared by the scientists who first raised the alarm.*

    Buy me a coffee!
    2,802 words