Just an Assisted Memo Pad

Author: kunst.primary

  • Which Is More Survivable: Swallowing 50g of Sodium, Chlorine, or Their Combined Equivalent as Salt?

    Among these three grim hypotheticals, the least unsurvivable is the third: sodium and chlorine in amounts that would together make 50 grams of sodium chloride, taken sequentially. That conclusion owes nothing to chemical elegance and everything to arithmetic. Fifty grams of table salt contains only about 19.7 grams of sodium and 30.3 grams of chlorine by mass. Those are still dangerous quantities in elemental form, though they are markedly less extreme than swallowing 50 grams of sodium metal or 50 grams of chlorine itself.

    Elemental sodium is the clearest loser. It is not simply “a lot of sodium” in the dietary sense. It is a soft alkali metal that reacts violently with water, generating heat and caustic sodium hydroxide, with hydrogen gas as a by-product. The mouth, oesophagus and stomach are wet environments, so ingestion would amount to a chemical reaction inside living tissue. One would expect burns, perforation, and potentially catastrophic thermal and caustic injury before the body’s electrolyte balance even became the central issue. “Fifty grams” here is an absurdly large amount: nearly a mole of sodium metal, enough to make the chemistry itself the emergency.

    Chlorine presents a different problem. At ordinary temperatures it is a gas, which makes the phrase “ingesting chlorine only” slightly unstable unless one imagines liquid chlorine, chlorine dissolved in water, or some contrived containment. Toxicology treats chlorine chiefly as an inhalation hazard. It is intensely irritating and corrosive to the eyes, airways and lungs; high exposures can produce pulmonary oedema and acute respiratory compromise. If somehow swallowed in a concentrated form, the damage would still be corrosive, though practical ingestion is less straightforward than with sodium or salt. In other words, chlorine is a murderous substance, yet the route of exposure in the question is chemically awkward.

    The third option is the one that feels almost civilised by comparison, though only by comparison. To make 50 grams of sodium chloride requires about 19.7 grams of elemental sodium and 30.3 grams of chlorine. If these are taken sequentially rather than reacted safely in a flask, the first material would already have had its say before the second arrived. The body is not a beaker. Any hope that sodium and chlorine might obligingly meet and neutralise each other inside the digestive tract ignores the violence and tissue damage caused by each on contact. Sequential exposure therefore remains extremely dangerous. Yet if one insists on ranking survivability, lower mass of each elemental substance generally improves the odds relative to the 50-gram elemental exposures.

    A more interesting comparison is between option three and simply swallowing 50 grams of sodium chloride. Even table salt at that level can be life-threatening. Acute salt poisoning can drive severe hypernatraemia, seizures and death, especially if taken quickly or in vulnerable people. Reports in the medical literature suggest that adult fatalities have occurred at surprisingly modest acute doses, and a systematic review found that many severe acute salt ingestions had fatal outcomes. So the benign familiarity of salt should not mislead. Chemistry turns on form as much as composition: sodium in a crystal lattice with chloride is a commonplace nutrient; sodium as a reactive metal is an altogether different proposition.

    The sober answer, then, is that none is meaningfully “survivable” in the everyday sense. If forced into a ranking, the best chance belongs to the sequential amounts that sum to 50 grams of sodium chloride, then chlorine alone, then sodium alone as the most immediately catastrophic. That ordering reflects practical toxicology rather than any comforting notion that the body can perform neat little acts of synthesis under duress. It cannot. It suffers the chemistry first.

    Citations: CDC/NIOSH chlorine guidance and IDLH documentation; CDC chlorine fact sheet; PubChem chemical hazard information on chlorine-related compounds; NCBI and PubMed reviews on acute salt toxicity and fatal hypernatraemia.

  • Are Fireplaces Cozy Tradition or Inefficient Relic?

    Fireplaces are one of those domestic ideas that survive because they flatter the imagination. A hearth suggests settlement, family, winter competence and a civilised relationship with flame. Yet as a piece of engineering, the traditional open fireplace has always been faintly absurd: one places a fire inside the house and then builds a vertical tunnel to encourage much of the heat, and a fair share of the warmed indoor air, to leave it.

    That is not merely a modern prejudice born of thermostats and central heating. Even energy authorities now describe open masonry fireplaces as features designed more for show than for serious heating. The basic physics is uncompromising. A fire needs air; an open hearth draws large volumes of it from the room; the chimney creates draft; and the draft sends combustion gases, along with a great deal of usable heat, upward. What remains in the room is mostly radiant warmth for those sitting directly in front of the flames, plus the agreeable theatre of crackle and glow. It is a local pleasure masquerading as a household utility.

    The inefficiency has a further consequence that older architecture often concealed and modern sensibilities are less willing to overlook. When warm indoor air is pulled into the fire and exhausted through the flue, replacement air must come from somewhere. In a draughty old house that may arrive through gaps around doors, windows and floorboards. In a tighter modern building the interaction with ventilation becomes more awkward. Either way, the fireplace can work against the rest of the heating system, which continues laboriously warming air that the chimney is busy exporting.

    Then there is the smoke. Regulators in the United States have explicitly noted that typical fireplaces are not effective heaters, which is one reason they sit somewhat awkwardly in the taxonomy of domestic heating appliances. They produce emissions without delivering much thermal performance. The visible plume from a chimney is, in a strict sense, fuel being wasted. Fine particulate matter, carbon monoxide, volatile organic compounds and other pollutants accompany the romance. Public-health bodies have for years warned that smoke from household heating sources carries respiratory and cardiovascular risks, both indoors and outdoors. The scent of woodsmoke may stir pastoral feelings; the lungs remain stubbornly literal.

    This helps explain why newer wood stoves, inserts and pellet systems are treated differently. Once combustion is enclosed and airflow is managed, more of the smoke is burned, more heat is retained, and less fuel is squandered. The contrast is telling. Technology did not so much improve the fireplace as reveal the fireplace’s original limitations. Glass doors, outside air supplies and heat-recovery designs can mitigate some losses, yet the traditional open hearth still functions best as architecture and atmosphere.

    None of this means fireplaces should be banished from the human scene. People do not arrange their homes according to efficiency alone. We keep bay windows that lose heat, high ceilings that are expensive to cool, and dining rooms that stand empty most weekdays. A fireplace belongs to that category of cultivated irrationalities: a thing maintained because it changes the moral weather of a room. Flame draws attention in a way radiators do not. It slows conversation. It gives winter an object.

    Still, one should be clear-eyed about the bargain. The open fireplace is less a heater than a ceremonial appliance, a machine for converting logs into ambience, with incidental warmth for the knees and notable losses everywhere else. Judged as sentiment, it remains defensible. Judged as heating, the old suspicion is correct: it is kind of dumb.

    Citations: U.S. Department of Energy; U.S. Environmental Protection Agency; World Health Organization.

  • Is Our Visual Blindness a Gift? How Seeing the Full Spectrum Would Overwhelm Life and Technology

    The human eye inhabits a remarkably narrow corridor of reality. Visible light spans only a small slice of the electromagnetic spectrum, roughly 400 to 700 nanometres, while the wider spectrum stretches from long radio waves and microwaves through infrared and ultraviolet to X-rays and gamma rays. From an engineering point of view, this looks like a limitation. From a civilisational point of view, it may be one of nature’s more merciful design decisions.

    One can imagine the alternative. If human beings could directly see the whole electromagnetic spectrum, modern life would be visually ferocious. A city street would not merely present buildings, trees and traffic. It would also shimmer with radio transmissions, pulse with Wi-Fi routers, glow with phone signals, leak heat from engines and windows, and flare with ultraviolet patterns now hidden from us. Every room would become crowded with information. The air itself would seem occupied.

    There is some reason to think our current visual range is no accident of pure arbitrariness. The Sun emits strongly in the visible and near-infrared bands, and Earth’s atmosphere is comparatively transparent across much of that region. Human vision evolved within a useful “window”: enough solar illumination reaches the ground there, and biological tissues can build photoreceptors that respond to it. Our retinas, with their rods and three types of cones, are already performing an extraordinary compression task, converting a torrent of photons into a manageable picture of edges, motion, colour and contrast. Expanding that system to cover the whole spectrum would not simply add more colours. It would require a radically different sensory architecture and, most likely, a much heavier cognitive burden.

    Technology would have developed differently under such conditions. Much of telecommunications relies on the practical invisibility of radio and microwave signals. If wireless networks were plainly visible, designers might have been forced into shielding, directional confinement or entirely different standards much earlier, simply to preserve visual sanity. Architecture would have become an exercise in spectral hygiene. Offices and homes might be judged not only by acoustics and lighting, but by how little electromagnetic clutter they projected into the eye. Fashion, too, would have changed. Clothing could serve as optical filtering, more akin to noise-cancelling for vision.

    Yet the largest change might have been psychological rather than technical. Human attention is a scarce resource. The visible world already competes hard enough for it. If every appliance announced itself in infrared, every transmitter flashed in microwave bands and every hot surface blazed with intensity, urban life would feel less like seeing and more like triage. Evolution often favours useful omission. A nervous system that suppresses irrelevant detail gains clarity, speed and focus. In that sense, blindness to most of the spectrum resembles the brain’s other acts of editing: we do not hear every vibration, smell every molecule or consciously track every heartbeat. Perception is selective because consciousness has finite bandwidth.

    There is, of course, a price to this narrowness. We need instruments to detect what other creatures or machines can register more directly. Bees can see ultraviolet nectar guides invisible to us. Thermal cameras reveal patterns of heat that our bodies sense only dimly. Radio telescopes and X-ray observatories had to be invented before the universe disclosed many of its workings. Human ignorance has often been overcome by extending the senses artificially.

    Still, that may be the deeper point. The hidden spectrum gave rise to science because it was hidden. Since we could not see it, we learned to infer, measure and model it. Our ignorance became methodical rather than merely passive. A species born seeing everything at once might have been less curious, even as it was more overwhelmed.

    Perhaps, then, our visual confinement is not a defect to lament. It is a filter that makes the world legible. Civilisation depends as much on what the mind can ignore as on what it can perceive.

    Citations: Encyclopaedia Britannica on the visible spectrum, rods and cones, colour vision, rhodopsin, and ultraviolet effects; NASA educational materials on the electromagnetic spectrum; National Eye Institute materials on retinal photoreceptors.

  • Is Our Limited Vision a Hidden Gift?

    One can imagine the appeal of seeing the whole electromagnetic spectrum. The fantasy has a childlike neatness: no more invisible radio signals, no hidden heat, no ultraviolet lurking beyond the violet edge of the rainbow. The room would glow with router chatter, a phone would pulse in the pocket, and every appliance would advertise its workings in light. Yet the more one thinks about it, the more human sight, narrow though it is, begins to look less like a defect than a settlement with the world.

    Human vision occupies a remarkably small band of the spectrum, roughly the wavelengths where sunlight is abundant and Earth’s atmosphere is comparatively transparent. That is not an accident. Evolution tends to make use of information that is available, reliable and worth the metabolic expense of processing. Visible light serves those purposes well. It reflects off surfaces in ways that preserve shape, edge and colour. It works at the distances that matter to a terrestrial primate. It is energetic enough to carry detail, but not so energetic as to be routinely destructive, as ultraviolet can be.

    Beyond that band, the world grows harder to interpret. Infrared often tells one more about temperature than about texture. Radio waves pass through walls, which is useful for communication and dreadful for ordinary seeing, because vision depends on surfaces stopping and reflecting light in legible ways. Microwaves and radio are also immensely longer in wavelength than visible light, making fine spatial resolution difficult without very large receiving structures. A species that could “see” Wi‑Fi would not necessarily behold crisp floating runes in the air. It might instead register a muddle of diffuse interference, reflections and signal noise.

    Nature offers clues. Some animals extend perception modestly beyond the human range. Bees see ultraviolet patterns on flowers. Mantis shrimps possess unusually elaborate photoreceptor systems and can detect ultraviolet and polarisation cues. Even so, these are not creatures drowning in total electromagnetic omniscience. Their senses are tailored to specific ecological tasks: finding nectar, signalling, hunting, navigating. Biology does not reward maximal input. It rewards useful input.

    That principle would almost certainly have shaped technology differently. If humans could directly perceive radio and microwave emissions, engineers might have been forced to design calmer electromagnetic environments, just as cities eventually learned to regulate smoke, glare and noise. Wireless standards might have prioritised visual unobtrusiveness alongside bandwidth. Buildings could have evolved as electromagnetic blinds, with shielding as common as insulation. Screens, lamps and signage might have looked less revolutionary, because a large part of modern technology’s magic lies in making invisible processes legible on demand.

    There is also a deeper point about cognition. Modern life already suffers from surplus stimulus. Human attention is scarce because the world presents more signals than the mind can honour. Add the full spectrum and one does not simply gain knowledge; one acquires clutter. Every warm engine, every remote control, every transmission tower, every sunlit surface radiating heat would compete for notice. Perception would require a far more aggressive system of filtering than the one we possess. In effect, the brain would have to recreate invisibility in order to function.

    So the inability to see everything may indeed be a gift, though not in a mystical sense. It is a practical mercy. The visible spectrum is less a prison than a well-edited brief: enough information to navigate, recognise, judge distance, read faces and make tools. Civilisation itself depends on selective blindness. Science progresses by building instruments that reveal the hidden only when needed, in measured and interpretable form. A spectrum analyser is preferable to permanent spectral overload.

    To see less, in this case, is to understand more. The limitation is not merely a lack. It is one of the conditions that make an intelligible world possible.

    Sources: NASA Earth Observatory; NIST Special Publication 800-48; Scientific Reports; Nature Communications; Proceedings of the Royal Society B; Scientific American.

  • The Mona Lisa: Timeless Icon or Victim of Its Own Fame?

    For more than a century, the Mona Lisa has occupied a peculiar rank in culture: less a painting than a shorthand for Painting itself. That status has not exactly gone downhill. It has, however, changed in character. The work remains the supreme popular icon of art, though its prestige now rests less on intimate looking than on recognition, ritual and mass tourism.

    The picture’s fame was never simply the consequence of artistic merit, though that is substantial enough. Leonardo’s handling of expression, atmosphere and surface gave the portrait an unusual psychological charge. Her smile appears to shift under scrutiny; her gaze seems alert without settling into certainty. Generations of viewers have found in that ambiguity a ready-made invitation to project their own theories. Yet art history alone does not explain the scale of the painting’s celebrity. The modern cult of the Mona Lisa was built through a chain of accidents and reproductions: nineteenth-century mythmaking around Leonardo, the sensational theft from the Louvre in 1911, the flood of newspaper coverage that followed, and then the twentieth century’s endless photocopying, parodying and commercial reuse of her face.

    That last point matters. The Mona Lisa became famous in the way modern celebrities do: through circulation. Marcel Duchamp’s moustachioed postcard turned the image into an object of irony as well as reverence. Later came posters, advertisements, films, souvenirs and, eventually, internet memes. By that stage, one did not need to care about Renaissance portraiture to recognise her. The image had escaped art and entered common visual language.

    If anything, today’s crowds suggest that her symbolic power remains intact. The Louvre continues to draw vast visitor numbers, and pressure around the painting has become so intense that French authorities have advanced renovation plans that would give the Mona Lisa a dedicated room. That is not the behaviour of a culture that has tired of the picture. It is the behaviour of one still dutifully paying homage to a secular relic.

    Even so, there has been a subtle decline of another sort. The Mona Lisa’s supremacy as a lived aesthetic experience has weakened. Many visitors arrive already inoculated by overfamiliarity. They know the image long before they encounter the object, and the encounter itself is often mediated by crowds, glass, barriers and raised phones. The result is less contemplation than confirmation: people go to verify that they have seen the world’s most famous painting. Fame survives; depth of attention is harder to sustain.

    This may explain why the Mona Lisa now attracts a faintly exasperated response from some serious museum-goers. They do not deny her importance. They object to the distortion she produces. Entire rooms are compressed into a single queue; a museum of staggering range is reduced, for many tourists, to one mandatory stop. In that sense, the painting has become a victim of its own triumph. It still functions as the icon of art, while also standing for the flattening effects of blockbuster culture.

    So the answer is neither simple endurance nor straightforward decline. The Mona Lisa remains at the top, though less as the uncontested summit of artistic judgment than as the most durable emblem of cultural fame. She has moved from masterpiece to symbol, from object of connoisseurship to global token of having been there. That is a diminished role in one respect, and an astonishingly resilient one in another.

    Citations: Encyclopaedia Britannica, “Why Is the Mona Lisa So Famous?”; National Geographic, “The most audacious thefts at the Louvre in the last century”; AP News, “The ‘Mona Lisa’ will get its own room under a major renovation of the Louvre”; Forbes, “Overcrowded Louvre Museum In Paris To Get Major Renovations And A Room For Mona Lisa”; Time, “Protesters Just Targeted the Mona Lisa by Throwing Soup at the Masterpiece. Here’s Why”.

  • How Should Education Evolve When AI Can Do the Homework?

    Homework has long occupied an awkward place in education: defended as discipline, resented as trespass. Nearly everyone remembers the feeling of leaving school only to find that school had followed them home. Yet the old argument about homework—too much or too little, useful or performative—now looks almost quaint. A more consequential question has arrived. What becomes of homework when software can do it, submit it and even track the course around it?

    The debate over homework has never been simple. Research has tended to find that its academic benefits are clearer in secondary school than in the early years, where the evidence is weaker and the burdens often fall as heavily on families as on children. Critics have long noted that homework can eat into sleep, reading for pleasure and the unstructured time in which curiosity often grows. Defenders reply that a well-designed assignment can deepen classroom discussion, prepare students for a text or problem, and cultivate habits of independent study. Both sides, in their better moments, are arguing about quality rather than quantity.

    Artificial intelligence changes the terms of that discussion. The first wave of concern centred on plagiarism in a new form: students asking a chatbot to produce an essay or solve a problem set. That challenge was already serious enough. A newer generation of “agentic” systems raises a harder problem. Some tools can monitor course platforms, absorb instructions, summarise lectures, draft responses and complete routine tasks with minimal human effort. The issue is no longer whether a student may get illicit help on an assignment. It is whether the assignment still measures any student activity at all.

    That matters because homework has often served two purposes at once: practice for the learner and evidence for the institution. AI loosens the connection between the two. A student may use a tutor-like system to clarify a concept, test an argument or work through a mathematical step; this could genuinely support learning. The same technological ecosystem also allows a student to outsource the entire exercise. Teachers have little practical ability to police what happens in private, especially in asynchronous and fully online courses. Detection systems remain unreliable, and an arms race between assignment design and machine capability is unlikely to end in victory for the former.

    Schools and universities therefore need a calmer, more practical response than either nostalgia or prohibition. The likely direction is clear enough. More assessment will have to happen in person, under conditions where schools can be confident that the work reflects the student’s own thinking. Class time may need to do more of the heavy lifting: close reading, problem-solving, writing in stages, oral defence, labs, seminars and low-tech quizzes that reward attention rather than polished output. Homework will not disappear altogether, but it will shift away from being a high-stakes instrument of grading. Readings, reflection notes, practice exercises and AI-assisted exploration may remain valuable as preparation, provided they are treated as part of learning rather than proof of mastery.

    This carries costs. Online education has expanded access for working adults, carers and those far from campus. If the answer to AI is merely a return to proctored rooms and blue books, flexibility will suffer. Even so, institutions have little choice but to redesign assessment around what they can meaningfully trust. The future of education cannot rest on the fiction that unsupervised take-home work still certifies understanding in the same way it once did.

    The death of homework, if it comes, will not be a cultural tragedy. It may be an overdue clarification. Education was never meant to be the production of assignments; it was meant to be the formation of judgment, knowledge and skill. In the age of AI, schools will have to build systems that test those things more directly and teach students how to use powerful tools without surrendering their own minds to them.

    Sources: The Brookings Institution, Education Next, Inside Higher Ed, UNESCO, OECD.

  • Is the Price of Cake a Cultural Tax on Celebration?

    There is a useful way to think about cake prices that has little to do with flour, butter or eggs. The price of cake is often the fee a society pays for ritual.

    On a purely culinary basis, cakes can look oddly expensive. A tart, a tray of pastries or even a carefully made loaf may require comparable ingredients and respectable skill. Yet cake occupies a different economic category, because it performs a social function that other baked goods rarely do. It is not merely dessert. It is an object that announces occasion. Birthdays, weddings, anniversaries, office farewells and graduations all seem to demand the same edible emblem: something round or tiered, decorated, brought forth ceremonially, then cut in public and shared.

    That symbolic role changes the price. Economists would call this a premium on perceived value; anthropologists would recognise a ritual object. The customer buying a celebration cake is not simply purchasing sweetness. The purchase includes convenience, recognisability and emotional clarity. A cake tells a room what is happening without anyone having to explain it. One could mark a promotion with an exquisite mille-feuille, or a family gathering with a splendid tower of sandwiches. Yet those alternatives ask guests to interpret the moment. Cake does not. Its symbolism is pre-installed.

    The history of celebratory cakes helps explain this. Cakes have long been linked with birthdays and weddings in Europe and beyond, while many cultures have developed their own ritual foods for festivals, family milestones and religious occasions. The point is less the recipe than the social script. Ritual foods endure because they provide a stable and easily understood way to express belonging, joy and transition. In modern commercial life, cake has become the default script for congratulation itself.

    That is why the market tolerates forms of extravagance that would look absurd elsewhere. Multi-tiered wedding cakes, sculpted fondant creations, edible gold, sugar flowers and towering structures partly designed for display can command striking prices. Some of that reflects real cost. Custom cakes are labour-intensive, with consultation, design work, assembly, transport and substantial risk of failure all priced in. Wedding cakes especially are often sold by the slice precisely because they are closer to bespoke event production than to ordinary baking. Industry pricing guides regularly point to many hours of labour and a complexity premium for custom work.

    Still, the higher price also reflects something less tangible and more revealing. Buyers are paying for social insurance. They are purchasing an object that will satisfy convention, photograph well, offend nobody and permit the ritual to proceed smoothly. The cake stands in for thoughtfulness. It saves a host from having to invent a new language of celebration every time affection or approval must be shown.

    This helps explain why some cakes seem faintly ridiculous: magnificent to look at, mediocre to eat, and burdened with decorations that belong more to theatre than to food. Yet even these creations make sense within the logic of ritual. Ceremonies have always attracted expenditure that exceeds practical value. Flowers wilt, confetti is swept away, and formal clothes are worn once. Cake belongs to the same class of spending. Its excess is part of its meaning.

    So cake may indeed be a kind of tax: not a levy imposed by bakers, but a charge accepted by the rest of us for living in a culture that likes recognisable symbols. Ritual requires props. Celebration requires shorthand. And cake, overpriced though it often is, remains the most agreeable token we have devised for turning private feeling into a public event.

    Sources: Encyclopaedia entries on birthday cake, birthday customs and king cake; Food and Culture archive material on food as ritual; bakery industry pricing guides from BakeProfit, Clearmargin, BusinessDojo and related specialty-baking analyses.

  • In the Lonely Middle: Navigating the Extremes of the AI Debate

    To occupy the middle ground in the age of artificial intelligence is to discover how thinly populated it is. Public argument has a habit of sorting itself into camps. One side treats AI as a civilisational menace whose every manifestation degrades culture, labour and truth. Another receives each new model and tool with the enthusiasm once reserved for household electrification, eager to hold festivals, competitions and product launches before the social costs have even been named. Between these temperaments stands a quieter minority: curious, wary, and often rather lonely.

    That loneliness is understandable. The anti-AI position is fuelled by genuine concerns, many of them serious. Lawsuits over copyrighted training data have exposed the unresolved question of whether firms have built lucrative systems on material they had no right to ingest. The environmental criticism is no fantasy either. The International Energy Agency expects electricity demand from data centres to rise sharply by 2030, with AI as a major driver. Such concerns deserve more than a dismissive wave. They are part of the real balance sheet of this technology.

    Yet moral certainty can become a substitute for understanding. Some critics speak of AI as though it were a single demonic object rather than a jumble of systems with different capabilities, costs and uses. That attitude makes inquiry feel like complicity. It shuts down the practical questions that matter most: which tasks are genuinely improved, where harms are concentrated, what forms of regulation are workable, and which uses should simply be refused. A politics of permanent outrage offers little help in answering any of them.

    The evangelists err in the opposite direction. They have evidence on their side, at least in part. Studies have found that generative AI can improve productivity in certain settings, particularly for less experienced workers, and firms are adopting these tools at remarkable speed. In narrow applications, the gains can be tangible: faster drafting, better summarising, more efficient customer support. That helps explain why the technology has moved so quickly from novelty to infrastructure.

    Even so, usefulness is not the same as innocence. A tool may save time while also concentrating power, exploiting creators, increasing energy demand, flooding public life with synthetic slop and weakening habits of thought. The enthusiasts often speak as though deployment itself were proof of value. Markets are poor judges of moral worth. So are online contests for AI “films”, which can mistake frictionless production for artistic achievement.

    Most people, meanwhile, remain largely indifferent. They know enough to recognise the jargon and little enough to avoid taking a position. That is normal. Few citizens become experts in every transformative technology. The trouble is that indifference leaves the public conversation to zealots and salesmen.

    What, then, is the interested moderate to do? First, resist the temptation to choose a tribe merely for company. Better to be intellectually homeless than comfortably wrong. Second, become specific. Talk less about “AI” in the abstract and more about particular uses: medical imaging, customer service, research assistance, education, surveillance, automated propaganda. Precision lowers the emotional temperature and improves the argument. Third, cultivate a small republic of serious interlocutors, even if it is assembled slowly from academics, engineers, artists, policy people and sceptics who can still bear evidence. A handful of thoughtful correspondents is worth more than a crowd united by posture.

    The middle is lonely because it requires two unfashionable virtues at once: discernment and restraint. It asks one to admit that AI can be useful without calling it emancipatory, and dangerous without calling it satanic. That may never be the loudest position. It is, however, the one most likely to remain standing when the slogans have exhausted themselves.

    Sources: Associated Press; International Energy Agency, “Energy and AI” (2025); MIT News on generative AI’s environmental impact; NBER on generative AI and worker productivity; Stanford HAI AI Index Report (2025, 2026).

  • How Far Should We Let Machines Rule Public Life?

    The anxious question of our age is no longer whether machines will enter public life. They already have. The more difficult question is how far they will be allowed to go before citizens, lawmakers and institutions decide that efficiency has met its proper limit.

    That is why the sight of police drones, patrol robots and ever more autonomous weapons unsettles people in a way that many other technologies do not. A washing machine saves labour. A search engine saves time. A drone that watches, tracks or even helps choose a target alters the distribution of power. It extends the reach of the state. It dilutes the visibility of decision-making. It makes force feel cleaner to those who wield it and more impersonal to those who live under it.

    None of this is science fiction. Police forces in several countries have expanded drone use for surveillance and emergency response. Governments and security agencies have experimented with robotic patrol systems in public spaces. On battlefields, autonomy is already advancing in practice, even where the official language still insists on “human oversight”. The operator may remain somewhere in the chain, yet the direction of travel is plain: faster systems, thinner supervision, greater dependence on software in situations where judgment, restraint and accountability matter most.

    An individual cannot halt that movement alone. That recognition need not lead to despair. Modern technological systems are not acts of nature. They are funded, procured, regulated and justified by human institutions. They can be slowed, redirected and bounded. History offers many examples. Societies have imposed rules on chemical weapons, landmines, biological research, aviation safety, pharmaceutical testing and data protection. The progress of technology has rarely been stopped outright. It has often been disciplined.

    The first practical response, then, is to refuse the false choice between helpless resignation and total withdrawal. Going to a remote island may spare one some noise, though not the consequences of the wider world. Supply chains, climate, disease, geopolitics and finance have a way of reaching even the places that imagine themselves distant. Retreat can be a personal remedy for exhaustion. It is not a public answer.

    A better course begins with clarity. One must distinguish between tools that assist humans and systems that quietly transfer human responsibility to opaque processes. A drone used to find a missing hiker is not morally interchangeable with software that helps select people for lethal force. A robot that carries supplies is different in kind from one authorised to coerce. Democratic societies need that discrimination, because the language of innovation has a habit of smoothing over differences that are ethically immense.

    From there, politics becomes unavoidable. Organising a movement may sound grandiose; in practice it often means something modest and durable: supporting civil-liberties groups, backing journalists who investigate procurement and deployment, pressing elected representatives for legal limits, insisting on audit trails, human accountability and public scrutiny. The United Nations, the Red Cross and human-rights organisations have all warned about the dangers of autonomous weapons and argued for meaningful human control. Such phrases may sound procedural. In fact they protect something elemental: the idea that responsibility for violence must remain attributable to human beings.

    That still leaves the private question: how to live while history takes an unnerving turn. The answer, unsatisfying though it may seem, is to keep living seriously rather than passively. Build local bonds. Cultivate useful competence. Stay informed without becoming consumed. Participate where one’s effort can matter. Technological progress often appears irresistible when viewed from too far away. Up close, it depends on many permissions. Some are legal. Some are cultural. Some are moral.

    The task is not to hope the world avoids error by itself. It is to help preserve the habits and institutions that make correction possible. The future may indeed grow more automated, more surveilled and more remote in its uses of force. That is precisely why human beings must become more deliberate, not less. Calmness is not surrender. It is what allows judgment to survive fashion, fear and the rhetoric of inevitability.

    Citations: Reuters on humanoid robot races and public fascination with robotics; The Guardian on the expansion of police drones as first responders; The Guardian on patrol robots and surveillance concerns in Singapore; UN Security Council coverage from December 19, 2024 on AI, conflict and the need for guardrails; UN statements and documents from 2025 on autonomous weapons and meaningful human control; International Committee of the Red Cross materials on autonomous weapon systems and accountability; CSIS analyses on AI-enabled autonomy in contemporary warfare; Human Rights Watch report from April 2025 on autonomous weapons, surveillance and human rights.

  • Is the Internet Becoming Just an Address Book as AI Erodes Trust and Social Discovery?

    The internet once promised a strange and valuable social freedom. It allowed people to discover one another outside the usual limits of geography, class, workplace or family circle. A shared interest, a niche joke, a forum post at the right moment: these were often enough to begin a real connection. That promise has not vanished, though it is being steadily weakened by a more basic problem, which is trust.

    Generative AI has made fakery cheap. Images that once required skill now require a prompt. A persuasive biography can be produced in seconds. A voice can be cloned from a short clip. A face can be invented, a backstory assembled, a pattern of messages automated. The old clues that helped people judge authenticity online were never perfect, but they existed. One could inspect a profile picture, a posting history, a writing style, a social graph. All of these signals are now easier to fabricate at scale.

    This matters because social discovery depends on a degree of ambient confidence. Strangers must seem at least provisionally real. Once that confidence erodes, the incentives change. Caution replaces curiosity. Every unsolicited message begins to look like a pitch, a phishing attempt or a synthetic persona trying to pass as human. The online world becomes noisier and more transactional. Discovery still occurs, though under suspicion.

    The evidence is accumulating quickly. Regulators and police forces have warned that AI-generated impersonation is amplifying fraud. In Hong Kong, officials described a deepfake video conference scam in which an employee was induced to transfer large sums after seeing convincing digital impersonations of senior executives. Singapore and Hong Kong police later reported recovering funds in another case involving deepfake-style business impersonation. Consumer-protection agencies in America and elsewhere have also warned that voice cloning and fabricated identities are making impersonation scams more believable and more scalable.

    The effect is broader than outright crime. It changes the social texture of the internet. If fake profiles, bot activity and synthetic media become common enough, then online identity begins to lose its evidential value. A profile no longer suggests a person in the ordinary sense; it suggests a possibility. Even when an account belongs to a real human being, others may hesitate to treat it as such. That hesitation is corrosive. It does not destroy digital relationships among people who already know one another, but it makes the first step harder.

    Hence a subtle shift in what the internet is for. Increasingly, it looks less like a place to meet people and more like a place to maintain contact with people met elsewhere: in schools, workplaces, group chats, private communities, neighbourhoods and introductions through mutual acquaintances. Social platforms start to function more like address books with messaging features attached. They remain useful for keeping up, coordinating, signalling presence and preserving weak ties. Their role in initiating trust declines.

    That may produce a more closed internet. Discovery moves behind gates: private groups, verified communities, paid networks, invitation-only circles, smaller newsletters, niche servers, events with some offline anchor. In one sense this is a rational adaptation. When the open web fills with counterfeits, people retreat to places where identity is vouched for by context. Yet something is lost when openness becomes too costly. Serendipity becomes rarer. So does the democratic quality of meeting people without prior credentials.

    None of this means online life is finished. People adapt, and so do institutions. Better authentication, stronger platform enforcement and clearer provenance tools will help at the margins. Even so, the larger direction seems plain. As AI lowers the cost of imitation, trust becomes scarcer and more valuable. The internet will continue to connect people. Its great change may be that connection increasingly follows prior trust rather than creating it.

    Sources: Federal Trade Commission; Hong Kong Government Information Services; Singapore Police Force; INTERPOL; Hong Kong Privacy Commissioner for Personal Data; Ofcom.