Just an Assisted Memo Pad

Category: Opinion

  • Don’t Trust AI Screenshots—Evidence Needs Accountability, Not Incantation

    Let me disclose the embarrassment upfront: this, too, has been run through a proprietary AI pipeline after the “author” lobbed in a bundle of rough notes instead of doing the decent thing and writing the column himself. Yes, the same author who likes to grumble about AI use apparently outsourced his grumbling to a machine because he was, in technical terms, feeling slothful and intellectually underfunded. You can probably tell. No matter how hard he tried, he could not stop the system from producing the faint aroma of AI varnish: the smooth transitions, the suspiciously tidy cadence, the lingering sense that a committee of very articulate office printers has joined the conversation.

    That confession aside, the complaint is real. One of the dreariest habits of the current internet is the rise of the AI screenshot as argument: a glowing chat bubble offered as though it were an affidavit, or the phrase “I asked AI and it says…” delivered with the solemnity once reserved for a footnote. This is a category error. A chatbot response may be useful. It may even be correct. Most of the time, on ordinary matters, it will land somewhere near common sense. That is precisely what makes it dangerous as evidence. Plausibility is not provenance.

    The distinction matters. Knowledge in public argument requires accountability. If a newspaper publishes a claim, there is an editor to blame. If a scholar makes one, there is a bibliography to inspect. If Wikipedia states something, one can trace the references, inspect the edit history and quarrel with the sourcing. The site spent years clawing its way from punchline to a surprisingly robust clearing house of referenced material because it built mechanisms for verification. It learned, through bruising experience, that anonymity alone does not doom a source, provided transparency and citation do the heavy lifting.

    A chatbot gives you almost none of that. The answer arrives as a polished paragraph detached from a visible chain of custody. Even systems that now append links do not always solve the problem; studies have found that generative search and chat tools can misattribute, fabricate or overstate sources, while researchers continue to build specialised citation and retrieval methods precisely because ordinary language models remain unreliable at attribution. In some evaluations, users nevertheless rate such systems as convenient and trustworthy, which helps explain the confidence with which AI-generated claims are now smuggled into everyday disputes. The machine sounds composed, and people confuse composure with authority.

    That is why the AI screenshot feels so objectionable. It resembles evidence while sidestepping the obligations evidence normally carries. It asks the reader to trust a black box whose workings are inaccessible, whose output is probabilistic, and whose mistakes are delivered in the same silky register as its truths. To cite a chatbot in place of a source is to say, in effect, that the forest spirits have whispered a fact into your ear and would rather not be cross-examined.

    None of this requires Luddism. AI is marvellous for many things: drafting, brainstorming, summarising, organising, translating, provoking, nudging. Even this piece began life as a lazy message-to-friend draft handed over to the silicon ghostwriter because the human involved could not be bothered to locate his own verbs. There is no point pretending otherwise. The sensible objection is narrower and sharper. Use AI as a tool for finding the trailhead, not as permission to skip the trail. If a chatbot gives you a statistic, find the report. If it offers a legal claim, locate the statute or judgment. If it states a historical fact, identify the book, archive or article from which the claim can be checked.

    Public discourse is already saturated with frictionless assertion. It does not need a new ritual in which unsupported claims are laundered through a synthetic voice and presented as neutral wisdom. The standard should be simple: if you want to persuade people, show your workings. Bring receipts with names on them. Accountability is what separates information from incantation.

    Sources: Nature; JMIR Cancer; arXiv; New Media & Society; Social Media + Society; OECD.AI.

  • AI Answers Aren’t Evidence: Why Accountability Still Matters in the Age of Chatbots

    This column, in the interests of full disclosure, was itself produced through a proprietary AI pipeline after the author lobbed in a handful of rough notes instead of doing the decent thing and writing it properly. One hesitates to call this efficiency. Intellectual indolence, perhaps. Mechanical loafing. A small outsourcing of conscience by a man apparently too lazy to finish his own sentences, even while grumbling about the machine.

    That irony, unfortunately, is the point.

    A peculiar habit has taken hold online: people post screenshots of chatbot answers as though they were documentary evidence. “I asked AI, and it says…” now appears in arguments with the tone once reserved for an official report, a court filing or a paragraph from a well-edited reference work. The screenshot arrives as a conversation-stopper, a synthetic witness offered to settle the matter. This should alarm anyone who cares about how knowledge is established in public.

    The problem is not simply that AI can be wrong, though it often is. The deeper problem is that a chatbot answer has almost no built-in accountability. A screenshot of a prompt and reply does not tell readers which sources were consulted, whether those sources were authoritative, whether the system blended fact with inference, or whether it quietly invented details with the confidence of a prep-school debater. Generative systems are designed to produce plausible language. Plausibility and provenance are not the same thing.

    Recent audits of AI search and answer engines have found exactly the sort of behaviour one would expect from a technology that mimics authority better than it earns it. Researchers at the Tow Center for Digital Journalism found that leading AI search tools frequently returned incorrect answers and often supplied broken, fabricated or misdirected citations. Other studies have shown that chatbots tend to overgeneralise research findings and flatten nuance, especially in scientific and medical contexts where caveats are the substance. Users experience this not as a technical quirk but as a breach of trust. Once a machine speaks in a smooth, declarative voice, many readers stop asking the old and necessary question: how do you know?

    There is an instructive comparison with Wikipedia. In its early years, invoking Wikipedia in an argument invited eye-rolling. The site was mocked for looseness, amateurism and vandalism. Yet over time it became more defensible, not because crowds became sages, but because the platform developed norms of verification. It built a culture around citations, edit histories, talk pages, reversion and visible disputes over sourcing. Wikipedia’s best principle remains wonderfully austere: the threshold for inclusion is verifiability. Readers can inspect the references, follow the trail and argue with something firmer than a glowing paragraph produced on demand.

    A chatbot screenshot offers none of that architecture. It is an answer severed from an evidentiary chain. Even when the content happens to be accurate, the form encourages intellectual laziness. It asks the audience to trust a performance of knowledge rather than the labour of showing one’s workings. In a healthy argument, claims are attached to institutions, authors, data, documents and methods. Someone can be corrected. Someone can be blamed. Someone can be asked to defend the point. With AI, responsibility disperses into the fog.

    None of this requires a puritan rejection of AI. These tools are useful. They are marvellous for drafting, summarising, brainstorming and helping one organise half-formed thoughts — as this column’s own scandalously idle gestation demonstrates. Yet using AI responsibly means treating it as a starting point for inquiry, not as the inquiry’s final court of appeal. If a chatbot gives you a statistic, find the study. If it summarises a historical event, locate the archive, article or book. If you want to persuade other people, bring them something they can inspect.

    That is what accountability looks like in public discourse. Not a screenshot of synthetic confidence, but a claim with a trail behind it. AI can help us get to the library faster. It should not be mistaken for the library itself.

    Sources: Tow Center for Digital Journalism/Columbia Journalism Review; Wikimedia Foundation materials on verifiability and no original research; Nature’s 2005 comparison of Wikipedia and Encyclopaedia Britannica science entries; Royal Society Open Science research on chatbot overgeneralisation of scientific studies; Scientific Reports on user-reported LLM hallucinations and trust.

  • Citing AI as Authority: When Screenshots Replace Accountability in Public Argument

    Let this be disclosed at the outset, if only to preserve a shred of intellectual honesty: this column, attacking the lazy use of AI as ersatz authority, has itself been drafted through a proprietary AI pipeline at the direction of its shadow author, who was evidently too indolent, too pampered by convenience, or too professionally committed to cutting corners to write it unaided. One imagines the poor creature sighing theatrically into a prompt box rather than facing a blank page like an adult. That irony is the point. AI can be useful. It can also become a spectacular machine for laundering vagueness into confidence.

    A peculiar habit has spread across the internet and, depressingly, into ordinary argument. Someone wishes to prove a point. Rather than cite a study, a report, a court filing, a public dataset or even a decent newspaper article, they post a screenshot of a chatbot response. “I asked AI,” they say, as if they had consulted an oracle with footnotes, editorial standards and legal liability. This ought to strike us as absurd. Instead, it is becoming normal.

    The problem is not simply that AI can be wrong, though it can be. The deeper problem is that AI is unaccountable in the precise way argument should never be. A screenshot of a chatbot answer is epistemically weightless. It presents a conclusion without a chain of custody. Who produced the underlying claim? What source supports it? Was the answer drawn from a reputable publication, a fringe blog, a misread summary, or a fabrication assembled in fluent prose? In most cases, the reader cannot tell. Even the person posting it often cannot tell. The result is a style of debate in which confidence survives while verification disappears.

    There is a useful historical analogy in Wikipedia’s early reputation. In the 2000s, many people treated it warily, often with good reason. Articles could be incomplete, sloppily sourced or vandalised. Yet Wikipedia’s great strength, over time, came from becoming more transparent, more heavily maintained and more deeply tied to citation. Its best pages can now function as maps to evidence rather than substitutes for it. A reader can inspect the edit history, check the references and follow the argument outward to more authoritative material. Even sympathetic studies have found Wikipedia increasingly reliable in many domains, while still warning that it is rarely the endpoint for serious citation. That distinction matters.

    A chatbot screenshot offers none of those safeguards. There is no stable page to inspect, no visible editorial process, no agreed version, no meaningful provenance and often no reproducibility. Change the wording of the prompt and the answer may shift. Ask for sources and the machine may provide genuine ones, invented ones or a muddled blend of both. Researchers and standards bodies have spent the past few years warning that generative AI systems can produce persuasive falsehoods and fabricated citations with unnerving ease. The technology has improved, certainly. Improvement does not amount to accountability.

    That is why “the AI said so” feels like a regression. It resembles the old, bad internet habit of treating easily generated text as evidence in itself. It invites people to outsource judgment while retaining the posture of having done research. It flatters the speaker with an air of technical sophistication while lowering the standard of proof. A machine’s answer can be a starting point, a clue, a shortcut to the relevant literature or a useful summary of material you then verify properly. It should not be smuggled into discourse as though it were a source.

    This is especially maddening because many of us who object to this habit are enthusiastic users of AI. It is perfectly sensible to use these tools for brainstorming, drafting, summarising and searching. One may even, with enough shamelessness, feed a half-formed rant into an automated writing contraption and have it polished into magazine prose. Yet that only sharpens the duty to distinguish assistance from authority. If AI helps you find a claim, go and find where the claim comes from. If it names a paper, read the paper. If it gestures towards a statistic, locate the underlying dataset or institution. If you wish to persuade, present information that can be examined, challenged and traced back to a responsible source.

    Public argument depends on accountability. Evidence should be inspectable. Claims should have owners. Screenshots of chatbot replies have neither quality by default nor responsibility by design. They may be useful in private workflow. They are feeble in public reasoning. By all means use AI. Just do not cite the ventriloquist dummy when you could cite the witness.

    Sources:
    UNESCO, guidance on generative AI in education and research; OECD, work on generative AI risks and integrity; Nature, comparison of Wikipedia and Britannica and later commentary on Wikipedia reliability; Nature Machine Intelligence, research on improving Wikipedia verifiability with AI; Scientific Reports, studies on user-reported hallucinations in AI systems; Nature, reporting on fabricated citations and disclosure concerns in AI-assisted research.

  • AI Is a Great Helper—But Don’t Mistake Its Words for Evidence

    There is an irony here worth stating plainly. This column, which argues against treating chatbot output as evidence, has itself been shaped through a proprietary AI pipeline at the shadow author’s direction. That is offered as both disclaimer and device. The contradiction is real, though perhaps instructive. AI can be useful in drafting, organising and sharpening prose. It becomes far less useful when people promote its answers as if they were proof.

    One of the more dispiriting habits of the past two years has been the rise of the screenshot as argument. A person posts a query to a chatbot, receives a smooth paragraph in reply, then waves that paragraph around like a signed affidavit. “I asked AI, and it says…” As a form of public reasoning, this is shoddy. A chatbot is not a witness, not an archive, not a scholar, not an editor and certainly not a source with legal or moral responsibility for what it emits. It is a machine for generating plausible language.

    That distinction matters. The problem with AI output is not merely that it can be wrong, though it often is wrong in confident and inventive ways. The deeper problem is that it dissolves accountability. A newspaper article can be challenged. An academic paper can be checked. A government report can be scrutinised. Even a mediocre blog post has an identifiable author, a publication date and some traceable chain of responsibility. A screenshot from a chatbot has almost none of that. It is detached from stable provenance, vulnerable to prompt manipulation and impossible to verify unless the user shows the exact query, context, model version and any system instructions that shaped the answer. Even then, reproducibility is shaky.

    In that sense, the better comparison is not with a book or encyclopedia, but with early Wikipedia. In its rougher years, Wikipedia was often treated with suspicion because the line between knowledge and hearsay looked alarmingly thin. Yet Wikipedia matured by doing something chatbots still struggle to do in public argument: it built norms of verifiability. Its articles became studded with citations; its edits became inspectable; its disputes became visible. One could check the footnotes, inspect the revision history and follow the argument back to accountable sources. The site earned a degree of trust not by demanding faith, but by making doubt easier to practise.

    Chatbots pull in the opposite direction. They flatten many sources, of varying quality, into a single polished answer with the mess removed. That polish is precisely what makes them dangerous in argument. They speak in a voice cleansed of uncertainty. Researchers have repeatedly shown that large language models can hallucinate facts and citations, and even the companies building them acknowledge that such systems are optimised to produce fluent responses rather than to preserve a transparent chain of evidence. The technology can be marvellous at synthesis. Synthesis without attribution leaves the user holding something rhetorically potent and evidentially weak.

    None of this requires puritanism. AI is already woven into ordinary intellectual life. People use it to brainstorm, summarise, translate, draft emails and sketch ideas they later refine. Journalists, students, academics and office workers all do versions of this, whether publicly admitted or quietly practised. The sensible norm is not abstinence. It is discipline. If AI points you toward a claim, go and find the claim in the world. If it names a study, read the study. If it describes a law, check the law. If it offers a statistic, locate the dataset or the report. Use the machine as a guide dog, not as a magistrate.

    That is why citing a random AI answer feels so outrageous. It asks the audience to accept language in place of evidence, confidence in place of accountability. It turns a tool for assistance into an oracle. A culture that tolerates this will make itself easier to manipulate, not because people are foolish, but because the form itself encourages passivity. The great task of the information age was learning to ask, “Where did this come from?” The age of generative AI makes that question more urgent, not less.

    So by all means use the tools. I do. This very piece is touched by them. Yet if one wishes to persuade, one still owes the reader the old courtesies: sources, traceability, and a path back to something sturdier than a machine’s immaculate guess.

    Sources:
    OpenAI, “Why language models hallucinate” (September 5, 2025); OpenAI, “Understanding the source of what we see and hear online” (May 7, 2024); Nature Machine Intelligence, “Improving Wikipedia verifiability with AI” (2023); Nature, “AI tidies up Wikipedia’s references — and boosts reliability” (2023); arXiv, “Citation-Enhanced Generation for LLM-based Chatbots” (2024); arXiv, “Hallucination-Free? Assessing the Reliability of Leading AI Legal Research Tools” (2024); Wikipedia, “Reliability of Wikipedia”; Wikipedia, “Wikipedia: Citing sources”; Wikipedia, “Wikipedia Seigenthaler biography incident.”

  • Why We Shouldn’t Treat AI Chatbot Screenshots as Evidence

    There is something faintly absurd about the new habit of posting a screenshot of an AI chatbot as though it were a notarised affidavit. “I asked AI, and it says…” has become a rhetorical flourish of the age: a way to end debate without doing the old-fashioned work of proving anything. The trouble is not that these systems are useless. They are often impressively useful. The trouble is that usefulness has been mistaken for authority.

    That confusion matters. A chatbot answer may be plausible, fluent and even correct. None of those qualities amount to accountability. In public argument, the standard cannot be that a machine produced a sentence which sounded informed. The standard has to be that a claim can be traced, checked, challenged and, if necessary, disproved. A screenshot of a prompt and a polished paragraph offers none of that. It is evidence of what a system generated for one user at one moment, under one set of hidden conditions. It is not a transparent record of the world.

    The comparison with early Wikipedia is revealing, though perhaps kinder to the bots than they deserve. Wikipedia once carried the stigma of looseness: anonymous edits, variable quality, thin citations. Yet over time it developed a culture of verifiability. Its best pages are not trusted because “Wikipedia says so”. They are trusted because readers can inspect the edit history, see disputes, follow footnotes and judge the underlying sources for themselves. The strength of Wikipedia lies less in omniscience than in auditability.

    Generative AI works differently. Large language models are built to predict convincing sequences of words, not to maintain a ledger of truth. Researchers have spent years documenting the problem politely known as hallucination: systems producing false statements, invented references or claims untethered from reliable evidence. This is not a marginal defect. It is a structural feature of a technology optimised for linguistic coherence. When such a system happens to be right, it may still be right for reasons invisible to the user. That opacity is precisely why an AI screenshot should not function as a citation.

    The dangers are no longer theoretical. Courts in the United States have had to deal with filings containing fabricated cases and citations generated by AI tools, prompting judges to emphasise that lawyers remain responsible for verifying every authority they submit. If legal professionals, operating under rules, sanctions and professional duties, can be seduced by fluent nonsense, one hesitates to imagine the standards governing the average social-media argument.

    None of this requires a puritan rejection of AI. On the contrary, the sensible case for these tools is strong. They can help organise ideas, summarise large bodies of material, suggest avenues for research and turn a rough sentiment into a sharper paragraph. They can act as accelerants for thinking. They should not be mistaken for the finished product of thinking. To use AI well is to treat it as a starting point, a draft partner, a research assistant who must never be left unsupervised with the bibliography.

    That distinction feels especially important now because our public culture already suffers from a shortage of accountable speech. Too many claims circulate detached from provenance, repeated because they are catchy, ideological or conveniently screenshot-able. AI risks supercharging that habit by giving unsupported assertions the sheen of composure and technical sophistication. A sentence generated by a chatbot arrives dressed like an answer. The costume is persuasive. It remains a costume.

    So by all means use AI. Ask it for a summary, a counterargument, a cleaner turn of phrase. Ask it to help you think. Then do what serious people have always had to do: find the source, check the quote, read the study, verify the statistic. If you want to persuade others, do not present a machine’s output as though it were self-validating proof. Give the underlying information, and give it with the one thing a chatbot cannot supply on its own: responsibility.

    **Sources:** Nature Machine Intelligence; Nature; Scientific Reports; U.S. District Court for the Southern District of New York; U.S. Court of Appeals for the Fifth Circuit.

  • Can Order Emerge from Chaos? The Surprising Elegance of the Figure-8 Solution to the Three-Body Problem

    There is something almost insolent about the figure-8 solution to the three-body problem. For centuries, the three-body problem has stood as a monument to the limits of prediction: give Newton’s law three mutually attracting masses and, in most cases, the future quickly turns unruly. Henri Poincaré drew from this the modern lesson that deterministic laws need not yield tidy forecasts. The equations are exact; the behaviour is not obedient. Out of that tradition, one expects turbulence, near-collisions, ejections and mathematical grief.

    And then comes the figure-8.

    In this remarkable solution, three equal masses chase one another along a single closed curve shaped like an eight, each body following the same path with a phase shift. The motion is planar, periodic and exquisitely symmetric. No body enjoys pride of place. No one sits at the centre like a sovereign sun. The system becomes a choreography in the literal sense: one line traced by three dancers, each arriving on cue. If the broader three-body problem is often treated as a warning about chaos, the figure-8 is a reminder that chaos does not abolish order; it sharpens our appreciation for how rare and delicate order can be.

    Its existence is itself a story about modern mathematics. The orbit was first found numerically in 1993 by Cristopher Moore, then proved rigorously in 2000 by Alain Chenciner and Richard Montgomery through variational methods. The proof mattered because celestial mechanics has long been haunted by seductive numerical mirages. A computer animation can suggest a heavenly waltz that dissolves under exact scrutiny. What Chenciner and Montgomery showed was that this one was real: hidden inside Newtonian gravity was a periodic orbit of startling elegance.

    Why does that matter beyond the specialist’s delight? Because the figure-8 solution reveals that the universe described by classical mechanics is richer than the old textbook caricature. Popular accounts often present the three-body problem as if it had only two modes: the solvable simplicity of two bodies and the anarchic confusion of three. The truth is more interesting. Between integrable perfection and generic chaos lies a landscape full of islands, symmetries and exceptional structures. The figure-8 is one of those islands, and its beauty comes partly from its improbability. It does not rescue us from complexity. It shows complexity producing form.

    There is also a philosophical sting in it. The scientific imagination often equates understanding with reduction: break a problem into simpler parts, find the governing law, and the rest should follow. Yet the figure-8 demonstrates that even under a law as old and compact as Newton’s inverse-square gravitation, astonishing behaviour may remain concealed for centuries. Knowledge of the rule is not the same as knowledge of its consequences. Whole worlds of behaviour can sit latent inside familiar equations, awaiting the right combination of symmetry, computation and proof.

    That is one reason the figure-8 has travelled beyond mathematics into culture. It lends itself to metaphor too easily to remain confined to journals. Three entities locked in mutual dependence, none controlling the whole, each shaping the path of the others: one can see why the image appeals in an age obsessed with systems, networks and feedback loops. Politics, finance, climate and technology all confront us with situations in which no single actor commands events, yet stable patterns sometimes emerge from interaction. The figure-8 flatters our hope that entanglement need not end in collapse.

    Still, the deeper lesson is sterner. The orbit is beautiful precisely because it is exceptional. It survives through symmetry, equal masses and highly special initial conditions. One should admire it as one admires a cathedral arch: not because every pile of stones forms one, but because this one does. The figure-8 does not repeal the unruliness of the three-body problem. It illuminates it by contrast.

    For that reason the figure-8 deserves to be seen as more than a curiosity. It is a rebuke to intellectual laziness. It tells us that even in domains famous for disorder, structure may be hiding in plain sight. It suggests that elegance is sometimes not the opposite of difficulty, but its most surprising offspring.

    Citations: Scholarpedia, “Three body problem”; American Mathematical Society, “A new solution to the three-body problem — and more”; Chenciner and Montgomery, “A remarkable periodic solution of the three-body problem in the case of equal masses”; Simó et al., linear stability analyses of the figure-eight orbit; Scientific American, “The Three-Body Problem”.

  • What Can an Old Unlocked Bicycle Teach Us About Urban Trust and the Commons?

    There are cities held together by law, by money and by surveillance. Then there are cities that continue, rather improbably, on the strength of tiny, unrecorded acts of restraint. An old bicycle left unlocked outside an apartment block ought to belong to the first category: a thing so vulnerable that only a fool would trust it to the street. Yet sometimes it slips into the second. It becomes, in practice, neither wholly private property nor fully abandoned object, but a modest common resource—used by strangers, maintained by them, and returned with a rough fidelity to where it came from.

    That is what makes the story of the old bicycle so arresting. Its owner does not lock it because it is too shabby to be worth stealing. It is cheap, already old, and only intermittently useful. For months it remains untouched, or seems to. Then one morning it disappears. By afternoon it is back, standing neatly where it always stood. Later, closer attention reveals a pattern: the saddle rises and falls according to the height of unseen riders; a loose seat suggests hasty adjustments; a flat tyre is mysteriously reinflated by the next day. The bicycle is not being stolen in the conventional sense. It is being borrowed by a silent public.

    Modern urban life is supposed to make this impossible. The dominant account of city living insists that anonymity erodes obligation. What is unattended will be taken; what is shared will be neglected; what belongs to everyone will soon belong to no one. Yet scholars of the “commons”, most famously Elinor Ostrom, spent decades showing that communities often manage shared resources successfully without either privatisation or heavy-handed state control. Her examples were forests, fisheries and irrigation systems, but the principle scales down remarkably well. A bicycle can become a commons in miniature if a local culture, however tacit, imposes rules: use it when you need it, do not ruin it, put it back.

    There is a deeper urban lesson here. Jane Jacobs wrote that cities depend on countless small public interactions, habits and watchful routines that generate trust without intimacy. The old bicycle belongs to that world. No committee governs it. No app tracks it. No contract defines the rights of use. Even so, a social order has formed around it. Somebody takes it to run an errand. Somebody else pumps the back tyre. Several people, apparently, feel obliged to return it to the exact same place. The bicycle’s continued existence depends less on ownership than on a shared recognition that this small convenience should remain available.

    That recognition is striking partly because it contrasts with more formal systems of sharing. Contemporary bike-share schemes rely on docks, GPS, payment systems and customer agreements because trust on a metropolitan scale is expensive to produce. When those systems fail, cities can end up with mountains of broken bicycles and a public soured on the promise of frictionless sharing. The unlocked old bicycle, by contrast, works precisely because it is local, legible and humble. Nobody imagines getting rich from it. Nobody mistakes it for a status object. Its value lies in use, not possession.

    This does not mean the bicycle story is sentimental proof that private property is obsolete or that urban life is kinder than we think. Informal arrangements are fragile. One truly selfish person could end the experiment in a single afternoon. Yet fragility is part of the point. The bicycle survives because enough people choose, over and over, not to exploit the opportunity before them. The city reveals itself not only through grand infrastructure and public policy, but through these ordinary decisions made in stairwells, parking areas and courtyards.

    An old bicycle, then, can become an index of civic health. It shows that between strict ownership and outright theft lies a broad, neglected territory of practical morality. In that territory, strangers recognise limits, preserve usefulness and sustain a tiny commons without ever naming it as such. The miracle is not that the bicycle vanished and returned. The miracle is that, in a city full of private need, so many people agreed to leave something behind for the next person.

    Citations:
    Elinor Ostrom, Nobel Prize materials; OECD reports on urban commons and commons governance; Springer research on commoning practices in streetscapes; Urban Studies research on cycling, bicycle parking and theft; historical research on the origins of bike-sharing.

  • 낡은 자전거가 보여준 조용한 신뢰의 사회, 소유와 공유의 경계에서

    도시 생활은 대개 결핍의 감각 위에 세워진다. 더 많은 자물쇠, 더 두꺼운 문, 더 정교한 인증 절차. 현대의 아파트 단지는 서로 스쳐 지나가는 사람들로 가득하지만, 정작 서로를 믿을 이유는 점점 줄어든다. 그래서 낡은 자전거 한 대가 보여주는 풍경은 의외로 낯설다. 누군가 훔쳐 갈 만큼 값나가는 물건도 아니고, 주인이 잃어버려도 크게 상심하지 않을 정도의 물건. 그런데 바로 그 하찮음 덕분에, 그 자전거는 어느 순간 사유재산과 공공재 사이의 기묘한 경계로 미끄러져 들어간다.

    주인은 자전거를 잠그지 않았다. 방심이라기보다 무관심에 가까웠다. 없어도 큰일 나지 않는 물건, 누가 가져가도 세상이 무너지지 않는 물건이었다. 그러자 예상 가능한 일이 벌어졌다. 자전거는 사라졌다. 그런데 더 흥미로운 일은 그 다음에 일어났다. 자전거가 다시 돌아온 것이다. 그것도 늘 있던 자리에 가지런히. 한 번의 해프닝이었다면 우연이라 할 수 있었겠지만, 안장의 높이가 달라지고, 바람 빠진 타이어가 다음 날 다시 채워져 있는 장면은 다른 이야기를 암시한다. 이 자전거는 도난당한 것이 아니라, 묵시적으로 공유되고 있었던 셈이다.

    사회학은 이런 현상을 오래전부터 설명하려 애써 왔다. 마르셀 모스는 교환의 핵심이 가격이 아니라 관계라고 보았다. 사람들은 물건만 주고받지 않는다. 호의, 의무, 체면, 암묵적 약속을 함께 주고받는다. 누군가의 자전거를 잠시 타고 다시 제자리에 갖다 놓는 행위는 법적으로는 애매하고 도덕적으로도 완전히 결백하지 않다. 그럼에도 그 행동에는 묘한 자제가 들어 있다. 가져간 사람은 완전히 훔치지 않았고, 돌려놓는 수고를 들였으며, 때로는 타이어에 바람까지 넣어 두었다. 익명의 편의가 익명의 보답을 낳은 것이다.

    이런 장면은 경제학자 엘리너 오스트롬이 탐구한 공유 자원의 세계를 떠올리게 한다. 공동체는 언제나 중앙의 통제나 사적 소유권의 철벽 없이도 스스로 규칙을 만들어 자원을 관리해 왔다. 물론 여기에는 명문화된 규칙도, 회의도, 합의문도 없다. 대신 더 미세한 질서가 작동한다. 쓰되 망가뜨리지 말 것, 빌리되 돌려놓을 것, 심하게 탐내지 말 것. 이 질서는 CCTV보다 약하고, 법률보다 흐릿하다. 그런데도 어떤 경우에는 놀랍도록 오래 지속된다. 신뢰가 높아서라기보다, 모두가 그 물건의 사소한 가치와 공동의 편익을 직감하기 때문일지 모른다.

    낡은 자전거의 이야기가 흥미로운 까닭은 인간에 대한 낙관을 강변해서가 아니다. 오히려 인간의 도덕성이 얼마나 비공식적이고, 상황적이며, 사소한 사물들을 매개로 형성되는지를 보여주기 때문이다. 사람들은 거대한 원칙 앞에서는 쉽게 냉소적이지만, 생활의 작은 편의 앞에서는 의외로 협조적일 수 있다. 이름도 모르는 누군가의 자전거를 타고 다녀온 뒤 원래 자리에 세워 두는 행위에는 최소한의 양심과 공동체 감각이 스며 있다. 그것은 영웅적 시민정신이 아니라, 생활의 매너에 가깝다. 그래서 더 현실적이다.

    이 이야기는 또한 소유의 의미를 되묻게 한다. 법적으로 내 것인 물건이 실제로는 여러 사람의 필요를 충족시키고, 여러 사람의 손을 거치며, 여러 사람에 의해 유지될 수 있다면 그것은 어디까지가 개인의 것이고 어디부터가 공동의 것인가. 도시인은 대개 자기 물건을 지키는 일에 익숙하지만, 때때로 어떤 물건은 닫힌 소유보다 느슨한 사용 속에서 더 생생한 사회적 기능을 얻는다. 낡은 자전거는 그래서 단순한 이동수단이 아니라, 이름 없는 이웃들 사이를 순환하는 작은 신뢰의 장치가 된다.

    거대한 사회를 믿기 어려운 시대일수록, 이런 사소한 사례는 오래 남는다. 사람을 믿으라는 교훈 때문이 아니다. 사람들은 생각보다 자주 남의 것을 잠시 빌리고, 생각보다 드물게 완전히 가져가며, 때로는 예상 밖으로 잘 돌려놓는다. 공동체는 대개 선언으로 시작되지 않는다. 대단치 않은 물건 하나를 여러 사람이 조용히 쓰고, 아무도 합의서를 쓰지 않았는데도 대체로 선을 넘지 않는 순간, 공동체는 이미 작동하고 있다.

    Citations:
    Encyclopaedia Britannica, “Gift exchange”; Encyclopaedia Britannica, “Generalized exchange”; Encyclopaedia Britannica, “Social capital”; Nobel Prize resources on Elinor Ostrom.

  • Are AI-Generated Bedtime Stories Replacing the Magic of Parent-Child Storytelling?

    There was a time when a child’s most sophisticated bedtime technology was a parent with enough energy left to invent a dragon. Now the evening plea has changed with the age: “Dad, will you generate me a bedtime story?”

    The sentence is comic, faintly absurd, and more revealing than it first appears. Children have always asked adults to conjure worlds on demand. The novelty lies in the verb. “Generate” belongs to the grammar of machines. It suggests infinite supply, personalisation at speed, stories assembled like playlists. A child can ask for a tale about a pirate astronaut who loves mango ice cream and fears thunderstorms, and some obliging system will deliver in seconds. The old scarcity of imagination has been replaced by abundance.

    That sounds convenient, even magical. In some ways it is. Stories told at bedtime have long done far more than fill silence before sleep. Research on shared reading and language-based bedtime routines links them with stronger language development, literacy, emotional regulation and better sleep habits. Reading aloud before bed can also aid word learning, while regular story routines appear to support empathy, creativity and parent-child attachment. For exhausted parents, a tool that helps spin fresh narratives may seem less like cultural decline than domestic reinforcement: a helper in service of an ancient ritual.

    Yet the important part of bedtime storytelling has never been the efficiency of production. It is the texture of attention. A bedtime story works because it arrives wrapped in human timing: the pause before the monster appears, the whispered improvisation when a child looks anxious, the digression that reflects the day’s scraped knee or playground triumph. The value lies partly in the story, and largely in the relationship around it. Studies of story time repeatedly find that interaction matters: turn-taking, questions, shared emotion, the small dance between adult and child. A perfect plot delivered without that exchange would miss the point.

    This is where generative AI presents both promise and risk. Used sparingly, it can be a creative prompt for adults. A father stuck after the third retelling of the same rabbit adventure can borrow a premise, adapt it, embroider it, make it his own. Families already use technology in this way: supervised, intentional, and subordinate to the adults in the room. Some parents and researchers see value in introducing children to such tools carefully, so they learn that AI is neither oracle nor toy but instrument.

    The danger begins when assistance becomes substitution. If the parent’s role shifts from storyteller to button-pressing intermediary, something intimate is thinned out. Bedtime is one of the last defensible frontiers against the industrial logic of frictionless content. Childhood already contains enough automated entertainment, enough screens, enough systems designed to anticipate desire before desire has fully formed. A child who asks for a generated story may be expressing wonder. An adult who answers only with generated output may be surrendering a small but meaningful duty.

    There is also a subtler loss. Children do not merely consume stories; they learn how stories are made. They listen to hesitation, invention, revision and voice. They discover that imagination is a human act, imperfect and alive. An AI can produce endless narrative combinations. It cannot model fatherhood. It cannot show what it means to think lovingly under mild pressure, to create something inadequate but personal for someone you adore.

    So the modern answer to that bedtime request should be neither panic nor surrender. Use the machine if one must, as one uses a torch to find the path. Then tell the story yourself. Change the ending. Add the family dog. Forget a detail. Invent a better one. Let the child interrupt. Let the tale wobble. In that wobble resides the point.

    The future of bedtime may indeed involve generated stories. One hopes it will still involve parents who know that the deeper task is not content delivery. It is to make a child feel, in the dark, that another mind is present and theirs.

    **Citations:** American Academy of Pediatrics via HealthyChildren; NIH/NICHD; Mindell et al., *Sleep Medicine Reviews*; Williams and Horst, *Developmental Science*; Dowdall et al., *Scientific Reports*; Sun et al., *Journal of Developmental & Behavioral Pediatrics*; Xu et al., *Journal of Children and Media*; *The Guardian* reporting on parents introducing children to generative AI.

  • Will Generative AI Serve Creators—Or Strip Their Imagination for Parts?

    The mood around generative AI in the creative world is neither simple euphoria nor straightforward revolt. It is closer to a tense, unstable bargain. Artists, writers, musicians and filmmakers are using these tools in growing numbers, often because they are fast, cheap and increasingly hard to avoid. Yet the prevailing sentiment remains suspicious, defensive and morally unsettled. Creative industries are not rejecting the machine. They are trying to stop it from becoming the landlord.

    That distinction matters. Much of the public conversation has been framed as a contest between romantics and technologists, as though artists object to efficiency on principle. The evidence suggests something more grounded. Creative workers do see utility in generative AI: for ideation, mock-ups, editing, translation, background production and administrative drudgery. Commercial design teams report gains in speed and workflow. Large creator surveys show that many professionals now incorporate AI into at least some part of their process. In practice, adoption has arrived faster than consensus.

    Yet willingness to use a tool does not amount to trust in the system surrounding it. Across sectors, the strongest sentiment is conditional acceptance paired with deep resentment about how the technology has been built and who stands to benefit. Visual artists surveyed on generative image models have expressed concern about workforce impacts, ownership and profit extraction from their work. Writers have been even clearer. In an Authors Guild survey, only a tiny minority said it was acceptable for AI companies to use books without consent or compensation; overwhelming majorities demanded permission and payment. Many were open to licensing arrangements, though only under strict conditions involving control, attribution and continuing remuneration. That is not technophobia. It is a market demanding terms.

    Music reveals the same pattern in harsher form. Songwriters and composers are anxious less because software can help produce a demo than because industrial-scale imitation threatens to flood the market, depress income and blur authorship. Recent surveys in music have found widespread fear that AI-generated tracks will compete with human work while being trained on it at the same time. Forecasts cited by cultural bodies and rights groups warn of substantial revenue losses for creators within a few years if current practices continue. For musicians, the grievance is economic, legal and existential all at once: their style risks becoming raw material for systems that owe them neither credit nor a cheque.

    Film and television have served as the clearest early warning. Hollywood’s recent labour battles placed AI near the centre of the dispute, and for good reason. The Writers Guild of America secured some of the first meaningful contractual guardrails, establishing that AI cannot be treated as a writer and that its use must not erode credit, compensation or creative control. The strike offered a preview of the broader politics of generative AI. Creative workers are not merely contesting a gadget. They are contesting the terms under which human labour is made legible, replaceable and bargainable.

    There is also a cultural objection that economics alone cannot capture. UNESCO and other international bodies have begun warning that AI is advancing faster than cultural governance, raising risks of homogenisation, platform dominance and erosion of cultural diversity. If generative systems are trained on the most abundant and marketable material, they may reinforce the familiar and flatten the marginal. That presents a problem larger than copyright. Culture depends on difference, minority expression, linguistic nuance and stubborn regional oddity. A machine optimised for plausible averages may produce abundance while weakening precisely those forms of originality that make culture worth having.

    The creative world’s sentiment, then, can be read as a three-part verdict. First, generative AI is useful. Second, the present settlement is unfair. Third, the danger lies less in machine creativity than in corporate asymmetry. Creators fear a future in which their labour is harvested without consent, their styles are diluted into generic supply, and their negotiating power collapses under the rhetoric of inevitability.

    That leaves room for a more constructive future, though only if law and policy catch up. Consent, compensation, transparency, provenance standards and enforceable labour protections would not smother innovation. They would civilise it. The central question is no longer whether generative AI belongs in the creative industries. It already does. The question is whether it will remain a servant of human imagination or become the mechanism by which imagination is stripped for parts.

    Citations:
    UNESCO, “A new expert report explores how AI is transforming culture”; UNESCO, “Re|Shaping Policies for Creativity 2026”; Authors Guild, “New Authors Guild AI Survey Reveals That Authors Overwhelmingly Want Consent and Compensation for Use of Their Works”; Lovato et al., “Foregrounding Artist Opinions: A Survey Study on Transparency, Ownership, and Fairness in AI Generative Art”; Brookings, “Hollywood writers went on strike to protect their livelihoods from generative AI”; OECD.AI incident summary on APRA AMCOS warning of revenue risk to Australian musicians; arXiv, “How Professional Visual Artists are Negotiating Generative AI in the Workplace”; Adobe, “Creative pros are leveraging Generative AI to do more and better work.”