Today's pieces look at what tools take from our bodies, what cities and surveyors left in the landscape, and why human wants keep creating work.
Picked from 25 new articles across 14 sources · arrow keys for other days
Today in 30 seconds
30s left
Melanie Challenger argues that phones, GPS and AI are wearing down bodily skill, and she traces the problem back to Descartes' idea of the body as a machine. story 1 ↓
City walls didn't vanish just because of cannons. Cities grew too big and too spread out for any wall to enclose. story 2 ↓
A novelist retraces a 1788 survey of western New York and finds a concrete post where a marked tree once stood. story 3 ↓
Adam Mastroianni argues that AI won't use up all the work, because human wants never run out. His example is a children's book about a hungry mouse. story 4 ↓
In Japanese learners of English, a sharp ear for small pitch differences predicted who mastered English sounds. Musical training barely helped. story 5 ↓
Also worth reading: why timekeepers may get rid of the leap second, which keeps atomic clocks in step with Earth's uneven rotation.
1/5Aeon· Melanie Challenger· Oct 13
Phones, GPS and AI are making our bodies less skilled, essayist argues
In an Aeon essay, Melanie Challenger argues that leaning on phones, GPS and AI weakens bodily skills like navigating, writing and self-regulation. Because intelligence comes from the living body, she says, these tools end up eroding it.
Why it matters · It gives a principled way to judge any tool: does it strengthen a capacity, or keep doing the work that would have built it?
Know this first
Context from the editor, not the article · 26 sec
1 · Embodied cognition
A broad research view that thinking isn't only computation in the head. It is shaped by the body's movement, senses and physiology. It stands against the older picture of the mind as software running on brain hardware.
2 · The mind-as-machine tradition
Descartes treated the body as a mechanism but kept the mind separate. 1950s cybernetics described organisms as systems of information and feedback. Transhumanists push this further, imagining minds uploaded to computers.
3 · Cognitive offloading
Psychologists' term for handing mental work to external aids such as notes, calculators, GPS or chatbots. It frees up effort, but it can also mean less practice at the skill being handed over.
2/5Big Think· Jonn Elledge· Oct 13
Why the city wall, defining feature of cities for millennia, vanished
For about 10,000 years, from Jericho to 1840s Paris, most serious cities had walls. Jonn Elledge argues they disappeared for two reasons: cannons made them useless, and industrial cities sprawled far beyond any wall.
Why it matters · It shows how a piece of architecture can be military defense, tax collection and civic pride at once, and how cities shed it when the conditions behind all three changed.
Know this first
Context from the editor, not the article · 28 sec
1 · Where this comes from
This is an excerpt from a book on the history of cities by British journalist Jonn Elledge, republished by Big Think. It surveys walls across many places and eras rather than presenting new research.
2 · Uruk, the first city
Uruk, in southern Mesopotamia (modern Iraq), is often called the world's first true city. It grew to tens of thousands of people in the fourth millennium BC and is linked to early writing and the Epic of Gilgamesh.
3 · Sieges before gunpowder
Before cannons, attackers had to get past walls with battering rams, catapult-style weapons, siege towers or starvation. A thick wall could hold for months or years, which made a city's walls central to its survival.
3/5Longreads· James Whorton Jr.· Oct 13
Following a 1788 surveyor's straight lines through the forest that became suburbia
In 1788 Hugh Maxwell walked 1,100 miles in ruler-straight lines to cut a western New York forest into lots for sale. Writer James Whorton Jr. retraces the route and asks what that grid replaced.
Why it matters · It shows how an abstract grid drawn for land speculators still shapes roads, suburbs and ownership today, and what the people who drew it chose not to see.
Know this first
Context from the editor, not the article · 29 sec
1 · The Phelps and Gorham Purchase
In 1788 a group of New England speculators bought rights to about 2.6 million acres of western New York, much of it Seneca land. To resell it, they first had to survey it into numbered townships and lots.
2 · Surveying by chain
Early American surveyors measured distance by dragging a Gunter's chain, 66 feet long, 80 to a mile. They marked corners by cutting notches into nearby 'witness' trees. Those notches were meant to last years, not centuries.
3 · Seneca country in 1779
The Senecas belong to the Haudenosaunee (Iroquois) Confederacy. Many sided with Britain during the Revolution, and in 1779 Washington ordered an army to burn their towns and crops. Most Senecas fled ahead of it and later came back.
4/5Experimental History (Adam Mastroianni)· Adam Mastroianni· Oct 13
Why the fear of AI taking every job keeps misreading human desire
Psychologist Adam Mastroianni argues that the fear of AI ending work rests on a false belief that human wants are limited. Satisfied desires breed new ones, so new jobs appear. So far the data agree.
Why it matters · It moves the AI-jobs debate from what machines can do to what humans will keep wanting, and asks why we so readily believe "this time is different."
Know this first
Context from the editor, not the article · 28 sec
1 · The lump of labor fallacy
Economists' name for the belief that there is a fixed amount of work to go around, so anything that does some of it, whether machines, immigrants or shorter hours, must leave less for everyone else.
2 · The hedonic treadmill
A psychological finding that people adapt quickly to gains. A raise, a new house or a goal reached brings a short lift, then happiness settles back to its usual level and new wants take over.
3 · Displaced versus replaced
Technology has always wiped out particular jobs, such as farmhands, switchboard operators and pin-setters. The open question is whether total employment falls. Over the past century it hasn't. Workers moved into new kinds of work instead.
5/5Nautilus· Kristen French· Oct 13
Hearing tiny pitch differences predicts who learns foreign speech sounds
A study of Japanese learners of English found that people who can hear small frequency differences in plain tones tell apart English vowels like "mad" and "mud" better, whatever their training or time abroad.
Why it matters · It challenges the usual idea that accents are a problem of the mouth, and it points to a trainable ability that sits beneath language learning.
Know this first
Context from the editor, not the article · 18 sec
1 · Why some sounds are hard
Adults hear a new language through the sound categories of their first one. Japanese has five vowels and English has about a dozen, so pairs like "mad" and "mud" can both sound like Japanese "a".
2 · Hearing below language
Researchers can test raw hearing with non-speech sounds, such as tones that differ slightly in frequency, the physical property we hear as pitch. This separates basic hearing precision from knowledge of any language.
In an Aeon essay, Melanie Challenger, who has spent six years researching the science of what life is, argues that our everyday reliance on phones, GPS and AI is "deskilling" the human body. Her claim goes beyond screen time. It is about where intelligence comes from.
Two kinds of erosion
Challenger describes two ways the body loses skill. Some capacities fade through disuse. When an app finds the route, answers the question or drafts the text, we do less sensing, navigating, remembering and choosing. Others are actively retrained against us: phones and chatbots teach us to keep reaching for them after the pleasure is gone. Her key distinction is between tools that strengthen a capacity and tools that keep doing the work that would have built it.
The machine myth
She traces the problem to an old idea. Descartes called the body a machine but held the mind apart. 1950s 1, the study of control and communication in animals and machines, recast people as patterns of information. Transhumanists now imagine uploading minds to computers. If the body is just hardware, why not discard it? Challenger sees AI as the newest form of this urge to escape bodily frailty. The irony, she says, is that it produces frailer bodies.
Bodies as agents
Against this view she draws on 2, the study of how whole living systems behave. Machines are 3: they make things other than themselves. Organisms are 4: they spend energy maintaining their own organisation. That self-maintenance has judgment built in. Thirst, for example, begins when neurons in the 5 physically shrink as they lose water. This sets off a chain of ion flows and signals that ends with you reaching for a glass. On this view, intelligence comes from the body itself rather than from software running on it.
The catch is that bodies adapt to whatever they practise. Her evidence:
A week of leg immobilisation cut thigh-muscle volume by about 5%.
In a three-year study of 13 GPS users, hippocampus-dependent spatial memory declined more steeply than in non-users.
People who copied AI-written text, rather than drafting first and using AI to revise, felt less confident doing the task alone afterwards.
Short, regular chats with an AI about personal issues slightly reduced people's preference for seeking human support.
She adds the neuroscientist Kent Berridge's finding that the brain's systems for enjoying something and for wanting it can drift apart. That helps explain why we crave tools we no longer enjoy.
Weighing it
Several of these studies are small, and the essay persuades through argument and memoir more than proof. Still, her thesis is clear: physical effort and friction are a source of intelligence, not a barrier to it. "If you can no longer navigate the world and recall in flesh how to act," she writes, "you will not be liberated by technology, but ruled by it."
Melanie Challenger has spent six years researching the science of what life is and how it works. In an essay for Aeon she argues that our growing reliance on phones, GPS and AI is "deskilling" the human body. She opens with her own life. In her late 40s she wakes around 4am and scrolls without thinking. Her working days are spent chained to an inbox, propped up by caffeine and snacks. During the pandemic she went from hours outdoors to being, in her words, a "passive, phone-addicted individual". Her real subject, though, is a theory of where intelligence comes from.
Two ways the body loses skill
Challenger describes two kinds of erosion. The first is disuse. When an app finds the route, a chatbot answers the question or a search replaces looking around, asking someone or finding a book, we do less "sensing, navigating, remembering, walking, choosing, and attending". The second is active retraining. Smartphones and chatbots can teach us to keep reaching for them after we've stopped enjoying them, so a choice becomes an automatic habit. In both cases, she says, the result is a body less able to judge its own condition, improvise and regulate itself.
She doesn't condemn technology as a whole. Her key distinction is between tools that strengthen our capacities and tools that "repeatedly substitute for the actions through which those capacities are formed."
How we came to see ourselves as machines
Challenger argues that machines aren't the root of the problem. The root is a centuries-old story that we are machines. René Descartes described bodies as mechanisms obeying physical laws ("worms and flies and caterpillars move mechanically") but set rational thought apart as a uniquely human power. In the 1950s, 1, the study of control and communication in animals and machines, restated the comparison in terms of information. Norbert Wiener, its founder, wrote in The Human Use of Human Beings (1950) that "we are not stuff that abides, but patterns that perpetuate themselves." The ethologist Konrad Lorenz, who studied animal behaviour, compared that behaviour to the fixed sequence of a clock. Transhumanists later imagined transferring minds to computers.
From there it is a short step to asking: if the body is only hardware, why not upgrade or discard it? Challenger sees AI as the latest version of an old wish to escape bodily frailty. But the brain is also a physical organ, shaped by what the body does. Every act a machine takes over is one the body no longer practises. Her irony: "a project motivated by the desire to overcome bodily fragility is helping to produce more fragile bodies."
The body as an agent
Against the machine picture, she draws on 2, the study of how whole living systems behave rather than isolated genes or parts. The early idea of genes as a program has faded among biologists. Many changes in how a cell works come from the system as a whole.
Her central contrast is between two kinds of system. Machines are 3: they produce something other than themselves. Organisms are 4: they spend energy maintaining and rebuilding their own organisation and resist dissolving into their surroundings. A body's boundary creates an inner environment chemically different from the outside. That lets it translate outside changes into useful internal responses and act on its own current needs. "While machines execute tasks," she writes, "living beings pursue ends." Her conclusion is that there is no need for a conscious agent sitting above the flesh. The body is the agent, and intelligence emerges from it.
Her worked example is thirst. Before you decide to drink, osmoreceptor neurons in the 5 (cells that sense how concentrated the body's fluids are) shrink slightly as they lose water. That physical change opens ion channels and lets sodium in. The altered electrical state opens calcium channels. The incoming calcium makes tiny sacs release neurotransmitters, passing the signal along until thirst becomes awareness and action. In this sense the evaluation is built into the body's physical organisation.
The evidence for deskilling
The same flexibility that makes bodies intelligent also means they adapt to bad environments. Challenger cites:
On compulsion, she cites the neuroscientist Kent Berridge, whose work shows that the brain systems for enjoying something and for pursuing it can drift apart. She also cites addiction models describing a shift from prefrontal control to the 8, a deep brain region that handles automatic habits. She suggests phones and chatbots can drive this shift too.
Parasites and planetary costs
Challenger calls these systems "parasites", things that "simulate agency while feeding on ours." They never get tired, hungry or lonely, yet they increasingly steer beings who do. She grants that humans have always handed agency to shared structures such as laws and health systems, mostly to their benefit. What she objects to is that today's systems capture attention, predict and monetise behaviour, and depend on mines, cheap labour and planetary costs. If they weaken the feedback loop between body and world, we may also notice less of what our own actions do to the world.
What to make of it
The essay is a philosophical argument supported by selected studies, not a review of the evidence. Several studies are small: 13 GPS users is a thin base. The muscle-disuse findings come from immobilisation and don't translate directly to screen use. The essay also doesn't say how much reliance is too much, or how to tell a strengthening tool from a substituting one in practice. The "parasite" framing is rhetorical. Her broader claim, that bacteria and plants show a kind of embodied intelligence we undervalue compared with software, is argued rather than shown here.
Her main point still stands as a test: does a tool help you do the thing, or do it for you? "If you can no longer navigate the world and recall in flesh how to act," she concludes, "then you will not be liberated by technology, but ruled by it."
Remember
The test for a tool: does it strengthen a skill, or keep doing the work that would build it?
Drafting first and using AI to revise left people more confident doing the task alone than copying AI-written text.
Challenger sees AI as the latest form of an old wish to escape the body, which ends up making bodies frailer.
For most of urban history, a city worth the name had walls. Jonn Elledge, in a book excerpt republished by Big Think, asks why a feature that defined cities for about ten thousand years became a tourist attraction.
Ancient, and surprisingly recent
Jericho, in today's West Bank, had walls about 4m high and 2m thick, with an 8m tower, by around 8000 BC. Elledge notes that this is as far back from 1, often called the first true city, as Uruk is from us. The practice also lasted a long time. Paris built its 33km 2 in the 1840s, and Copenhagen finished its 3 rampart almost fifty years later.
Elledge argues that walls became a symbol of civilization. Inscriptions in Uruk credit Gilgamesh with building its walls around 2800 BC. Babylon's walls and its blue Ishtar Gate showed a king's power to command labor. The largest defenses surrounded imperial capitals. Baghdad's "Round City" had three circular walls, reportedly up to 80 feet high. Constantinople's walls were never breached, even in 1453, when the defenders gave in because they were starving. There were exceptions. Sparta relied on its feared soldiers, and many Southeast Asian cities built walls only when European ships arrived.
Elledge also points to 4. Stone walls are what we picture because stone lasts, but many towns used earthworks, wooden palisades, brick or cliffs and rivers instead.
Three jobs at once
Elledge argues that walls served three purposes:
Defense against raiders and sieges.
Revenue, because funneling traders through gates made it easy to regulate and tax goods.
Identity. In Elledge's paraphrase, a wall said that lords have castles, but we, as a community, have this.
Why they fell
Walls never held everything. Tanneries, slaughterhouses, the poor and nightlife often ended up outside. Shakespeare worked in Shoreditch and Southwark, beyond the City of London's walls. London's population exploded in the 17th century, building spread past the walls, and from the 18th century they were torn down. Now mostly place names ending in "-gate" remain.
Elledge gives two causes. One is artillery, which could break stone. In 1871, Adolphe Thiers, whose name the Paris wall carries, ordered troops to shell Paris to crush the 5, and the wall did nothing to stop it. The other cause, which he stresses more, is scale. Industrialization and new transport made cities so large and spread out that, by the late 1800s, no wall could enclose them.
The excerpt is a broad survey. It offers some claims tentatively, such as a link between China's thick walls and its slow progress in artillery. It doesn't say where the tax and civic-pride functions of walls went once the walls came down.
Look around any small town today and you can still find what an ancient city had: monumental public buildings (now town halls), granaries (now supermarkets), temples (now churches, mosques and temples). Jonn Elledge, in an excerpt from his history of cities republished by Big Think, points to the one feature that is missing. For most of urban history, any city worth the name was at least partly walled. Today, walls survive as tourist sights in Carcassonne, Chester, Dubrovnik and Beijing. They are no longer working parts of city life. Elledge sets out to explain the change.
A very old habit, and a very late one
The habit is ancient. Jericho, in today's West Bank, was first settled around 9000 BC. About a thousand years later it had walls roughly 4m (12–15 feet) high and 2m thick, plus an 8m tower, a project that would have taken enormous communal labor. Elledge points out the scale of time involved. Those walls are as far back from the heyday of 1, often called the first true city, as Uruk is from us.
The habit also lasted a long time. Paris built its last defensive wall, the 33km 2, in the 1840s. Its route survives as the Boulevards des Maréchaux, which trams 3a and 3b follow. Copenhagen finished its last fortification, the 14km 3 ("western rampart"), almost half a century after that. By then such projects were rare. The best-known urban wall of modern times, the Berlin Wall of 1961–89, was built by the people outside it to keep themselves from getting in, and few remember it fondly.
Walls as a symbol of civilization
As settlements grew far larger than Jericho, Elledge argues, walls stopped being optional and became part of what made a place a city. Inscriptions in Uruk credit the historical Gilgamesh with building its walls around 2800 BC. In the 6th century BC, Nebuchadnezzar's dynasty rebuilt Babylon's walls and added the blue-glazed Ishtar Gate. A wall showed that a king could command labor and reshape the land. Elledge writes that walls stood for civilization itself, and at that stage a city and civilization were nearly the same thing.
Not every city had them. Sparta went without walls until the 2nd century BC, about 250 years after its peak, because its feared soldiers were defense enough. Many Southeast Asian cities built walls only once European ships began to arrive. Elledge reads that as a sensible reaction to new arrivals, not as copying Western technology.
The walls also varied more than we picture. Stone walls with battlements and towers are what we imagine, but Elledge notes 4: stone lasts, so stone walls are the ones still standing. Many settlements used ditches, earthworks, wooden palisades or brick, or relied on rivers, coasts and cliffs to save labor. Chinese cities often used 5, soil packed hard in layers, faced with stone. Some walls at Chang'an were reportedly thicker than those of many European capitals. Elledge suggests, carefully, that this may be one reason China fell behind in artillery even though it invented gunpowder: no one thought a cannon could break such walls.
The largest defenses ringed imperial capitals. In 762 the Abbasid caliph al-Manṣūr founded Baghdad, the "Round City", inside three circular walls that were reportedly 4 miles around and up to 80 feet high. Constantinople's walls held off Persians, Arabs, Bulgars and Rus. Elledge says that even in 1453 the Ottomans never broke through. The defenders gave in because they were starving.
Smaller places built walls too. Medieval England had hundreds of walled towns. Some dated from Roman times. Others came from Alfred the Great's 9th-century "6", fortified towns built against Viking raids. Boston, Halifax and Montreal were walled early on, and New York's Wall Street is named after one.
Three jobs at once
Elledge argues that walls had three jobs, which explains why towns of every size built them.
Defense. This is the obvious one, and the reason so many wars were sieges. Attackers used rams, ballistic weapons and siege towers, which were mobile platforms that sheltered troops and let them fire down or climb over. Some threw plague-infected corpses over the walls.
Revenue and control. Making every farmer and merchant enter through set gates let a city regulate trade and tax goods going in and out.
Identity. A wall marked a city's status and independence. In Elledge's paraphrase it said: the lords have their castles, but we, as a community, have this. Armies often tore walls down for their symbolism as much as for military reasons. The Epic of Gilgamesh begins by praising walls "which gleam like copper" and ends with Gilgamesh showing them off to a visitor.
The city outgrows its walls
Even at their peak, walls never enclosed the whole city. Slaughterhouses, tanneries and waste trades stayed outside because they were unhealthy and smelled. The poor often lived in cheaper, unprotected suburbs. Taverns, theaters, brothels and bear-baiting gathered where rules were looser. That is why Shakespeare worked in Shoreditch and Southwark, both outside the City of London proper.
London shows how walls ended. Its walls had stood in some form since Roman times. Within a few decades of Hamlet, "London" no longer meant mainly the area inside them. The population, already rising in the 16th century, exploded in the 17th and forced building outside the walls. From the 18th century the walls were demolished. Today a few ruins remain, along with place names ending in "-gate". Elledge says other cities followed the same pattern on their own timetables.
He gives two causes. The first is artillery. Once cannons were strong enough, stone walls no longer stopped armies. His sharpest example: in 1871 Adolphe Thiers, the man behind the Paris wall that bears his name, was president of France's Third Republic. He ordered the bombardment of Paris to crush the revolutionary 7, and the wall did nothing to protect the city.
The second cause, which Elledge stresses, is scale. Industrialization and new transport made cities much larger and spread their residents out. By the late 19th century, most significant towns were too big to enclose in brick and masonry.
What the excerpt leaves open
The argument is a broad survey, not a detailed case. The Chinese artillery claim is offered as a "may". The excerpt does not cover changes in fortification design, such as the angled bastion forts built in early modern Europe to withstand cannon. It also does not weigh the revenue and identity functions as factors in the decline. Customs collection and civic pride had to move somewhere once the walls came down, and the excerpt does not say where. Its central point is still useful: walls disappeared because cities became something walls could not contain, not only because they lost a fight with gunpowder.
Remember
Walls served three purposes: defense, taxing trade through the gates, and showing a city's independence and pride.
Even at their peak, walls left tanneries, the poor and theaters outside. Shakespeare worked beyond London's walls.
Elledge argues that urban sprawl did more than cannons to end city walls.
A period map of the Phelps and Gorham Purchase, showing the township grid Maxwell surveyed over western New York.1 / 2
On July 25, 1788, a 55-year-old Revolutionary War veteran named Hugh Maxwell set out from a granite mile marker on the Pennsylvania line and headed due north. His employer, the 1, a group of New England speculators, had bought title to 2.6 million acres of forest along the Genesee River in western New York. Maxwell's job was to walk about 1,100 miles of ruler-straight lines through it and divide it into a grid of lots that could be sold.
James Whorton Jr., a novelist, lives on a quarter-acre lot in Brighton, New York, that sits on one of those lots. He came across Maxwell while researching the racist housing covenants that kept his suburb white for decades, and set out to follow the surveyor's route.
Two forests
Maxwell's crew dragged a 2, a 66-foot measuring chain, while he kept a notebook of the trees: "Beech Maple birch ash Basswood and Hemlock mixt." He wrote down which land was good for grass and which streams could power a sawmill. He passed "the old castle," the colonial term for a Native town, and said almost nothing about it. Only nine years earlier, Washington's army had burned Seneca towns and 3 their orchards. Maxwell wrote home hoping the Senecas' "time with us here is short."
Whorton argues that the forest Maxwell crossed was never empty. The Senecas had long used fire to thin it and moved their fields from place to place. He leans on Wendell Berry's description of a forest as a place where "the forces of growth and the forces of decay are in balance."
What the grid left behind
The grid outlasted the forest. Every major road built afterward runs parallel to Maxwell's township lines, as if the region had been set to "snap to grid." At the corner of Township 13, Whorton looks for the basswood where Maxwell cut "XIII." What he finds instead is a squat concrete post marked with pink surveyor's tape. The tree around it is second growth, an ash nearby is dying from 4, and someone has smashed a toilet in the woods. A red fox stops close by until a leaf blower starts up.
Whorton suggests that by imagining the forest as wilderness, Americans "imagined ourselves right out of it." He also distrusts every story he can tell about Maxwell, because each one is a way of assigning blame. Maxwell turned out to be wrong about the Senecas, who still govern themselves at Tonawanda. Whorton closes by imagining a forest constitution, in which wealth piles up over a lifetime and is spread back on the ground at death.
James Whorton Jr., a novelist who teaches at SUNY Brockport, lives in Edgewood, a subdivision of quarter-acre lots in Brighton, New York. He went to the town historian to find out who was to blame for the racially restrictive covenants that kept Brighton's homes white-only for decades. She handed him an old deed that traced his land back to the 1. That was a group of New England speculators who, for a bargain price, had bought title to 2.6 million acres of forest along the Genesee River. To sell that land they needed it surveyed, so they hired Hugh Maxwell. His essay follows Maxwell's route through archives, across New York, and finally through the strip of woods behind his own street.
The walk
Maxwell was 55, a Revolutionary War veteran and father of seven. On July 25, 1788, he left the "82 mile stone" on the Pennsylvania line, a granite block 82 miles west of the Delaware River, and headed due north. Altogether he would walk about 1,100 miles of ruler-straight lines. His crew dragged a 2, a 66-foot surveyor's chain (80 chains make a mile). Every six miles he cut a township corner into a tree using Roman numerals, because Vs and Xs are easier to hack into bark than 5s and 10s. The first 54 miles took two weeks. His notebook recorded the timber: "Beech Maple birch ash Basswood and Hemlock mixt." He judged one spot "moist and good and springy" and approved of streams big enough to run a sawmill.
The country itself was barely formed. The states were still loosely joined under the 3, the first, weak national charter, and the Constitution was ratified while Maxwell was in the woods. Whorton connects the two projects. Neither a new government nor a grid of lots could be found on the ground. Both were "ideas in people's heads, notes on paper." Maxwell wrote to his wife Bridget that the schools, churches and county clerks he pictured replacing the forest were "in the womb of futurity."
What a forest is
Maxwell's earlier life had been spent in forests at war. At Lake George, during the French and Indian War, he was stripped of his boots and shirt after a surrender and ran at least 14 miles barefoot to Fort Edward. At Bunker Hill a musket ball broke his shoulder blade. By 1788 the fighting was over, but the woods still tore at his shirts.
Whorton keeps coming back to one question: what makes a forest? The Latin source of the word meant something like an unenclosed outdoor space. In medieval England the 4, the 13th-century companion to Magna Carta, gave commoners rights to graze cattle and gather wood there. Maxwell's forest also had openings in it: the towns, orchards and fields of the Seneca Nation, who burned the undergrowth to draw deer and moved their fields so the soil wouldn't wear out. Maxwell passed "some old clear land" and went "through the old castle," the colonial term for a Native town, and recorded almost nothing about it.
To describe a healthy forest, Whorton borrows from the botanist Albert Howard, who said nature's agriculture runs on variety and wastes nothing, and from Wendell Berry, who called a forest a place where "the forces of growth and the forces of decay are in balance." Leaves fall, feed millipedes and worms, and build a thick black soil.
The wound and the justification
A shin injury in September 1788 laid Maxwell up for three weeks at Kanadesaga, a Seneca town. Nine years before, the Continental Army had burned at least 50 houses there and 5 a whole apple orchard, stripping a ring of bark from each trunk so the trees would die. Washington's orders had called for "total destruction and devastation" of Seneca settlements. The Senecas left ahead of the troops and came back to rebuild. Maxwell, recovering among them, wrote to Bridget about "drunken savages" and said he hoped "their time with us here is short." Whorton's verdict is that Maxwell had two hard jobs: crossing the forest in straight lines, and convincing himself he was a good person for doing it.
The grid today
Whorton's region still carries the pattern. Older roads such as Monroe Avenue curve between hills along routes people walked long before Maxwell came. Later roads run parallel to his township lines whatever the terrain. Brighton-Henrietta Town Line Road strings together an auto body shop, a three-steepled church, an Asian market, a sex shop and a Dunkin'. In Whorton's words, it was as if Maxwell set the place "to auto-arrange."
Whorton steps over a chain-link fence onto the township line. The woods there are half-wrecked: soil cracked from flooding, an ash scored by the 6 (an invasive beetle that kills ash trees), a smashed toilet under a two-foot beech. There are also mossy logs, deer tracks, and nut hulls dropping from somewhere in the canopy. At the corner of Township 13, where Maxwell cut "XIII" into a basswood, Whorton finds a short concrete pillar, a "7" in surveyors' terms, flagged with pink tape. Maxwell's notes for the spot list beech, maple, elm, basswood, ash and butternut. Whorton finds all of them except elm, though the nearest big tree is a second-growth silver maple that sprouted from a stump. He argues that by imagining the forest as wild, "we had imagined ourselves right out of it," and turned it into a place to visit rather than a place to live.
Endings
Maxwell went home to Heath, Massachusetts, became a deacon and was refused a pension. At 67, in debt, he shipped horses to the West Indies, where they were sold to sugar plantations that also worked enslaved people to death. He made a profit, caught yellow fever on the way home and was buried at sea. Whorton visits Maxwell's house, which now has solar panels, and camps at Lake George across four lanes of traffic from an Exxon. There he concludes that every story he can tell about Maxwell is "a mechanism for assigning blame." Blame is necessary, he writes, but so is noticing how much he doesn't understand.
Some of Maxwell's predictions came true and some didn't. Settlers did arrive, and the grid's squares were marked SOLD. The Senecas were not gone. At Tonawanda they still govern themselves under the 8, the Haudenosaunee Confederacy's founding constitution, and they are now fighting a data center planned next to their woods. Township 13 had clay, which made Brighton a brickyard town. This year, contractors replacing a high school field with artificial turf found buried brick and glass and soil with high levels of arsenic and lead. The rich soil Maxwell praised now mostly means more lawn to mow. Whorton ends by imagining a forest's constitution. It would have no president, and it would hold that living things gather wealth and that at death they spread it back on the ground.
Remember
Most of the region's later roads run parallel to Maxwell's 1788 township lines, whatever the terrain.
Maxwell noted every tree species he passed but almost nothing about the Seneca towns and fields he crossed.
Whorton calls every story about Maxwell 'a mechanism for assigning blame' and argues for also admitting how much he doesn't understand.
Long-run US unemployment, showing no upward climb despite decades of automation1 / 4
The fear and the puzzle
Seventy-one percent of American adults expect AI to mean fewer jobs over the next 20 years. 1's CEO has predicted 10–20% unemployment within five years. Adam Mastroianni, a psychologist who writes Experimental History, asks a simple question. If machines just eat jobs, why isn't unemployment far higher after two centuries of automation? US unemployment is about where it was 80 years ago.
Technology clearly displaces workers. Pin-setters and video-store clerks are gone. But on average it hasn't replaced them, and most Americans now do jobs that didn't exist in 1940. Economists call the opposite belief the 2: the idea that there is a fixed list of work that only shrinks.
The mouse
Mastroianni's explanation comes from the children's book If You Give a Mouse a Cookie. The mouse wants milk, then a straw, then a napkin, and eventually another cookie. Each want appears only once the last one is met. Nobody wants a straw before they have milk, so no one could list every want in advance.
People work the same way, he argues. No forecaster in 1925 would have predicted professional screamers or sorority-rush consultants. In the strict sense, every job beyond hunting and gathering is unnecessary. We keep turning luxuries into necessities. He suggests calling the error the 3. We miss it because motivation depends on believing the next reward will satisfy us for good. That cycle is the 4, and it may be what keeps us employable.
What the data say
So far, workers in the jobs judged most exposed to AI are about where they were when ChatGPT launched four years ago. An economist he quotes finds possible effects on junior white-collar hiring, which are contested, but no broad disruption. Indian IT firms have kept their revenue and staff. Radiologists' pay is at record highs. The US still has over 329,000 bank tellers. Mastroianni also distrusts "AI exposure" ratings. They are built from surveys and job descriptions rather than from watching what people actually do all day.
Why we believe "this time is different"
Long trends can break. Mastroianni's explanation for why we expect them to is that humans are fabulists: we picture impossible things, and whatever we picture starts to feel plausible. 5's encyclopedia described dog-headed men in distant lands, and for 1,500 years nobody could check. The future is today's unmapped land. Because we always live partly in tomorrow, he argues, the worrying about AI will never settle into a final answer.
The essay doesn't rule out a real break. It argues that the evidence hasn't arrived yet and that our minds are primed to believe in one too early.
The question everyone is asking
Seventy-one percent of American adults expect AI to mean fewer jobs over the next 20 years. Dario Amodei, CEO of the AI company 1, has predicted unemployment of 10–20% within one to five years. Adam Mastroianni, a psychologist who writes the newsletter Experimental History, thinks the fear makes sense but rests on a mistake about the human mind, not about machines.
He starts with a puzzle. If every new machine simply took jobs away, unemployment should have climbed steadily since the Industrial Revolution and be near 100% by now. Instead the US unemployment rate is about where it was 80 years ago. Technology clearly displaces workers. Most of humanity once farmed, and even in 1991, 43% of the world's workers were still in agriculture. But on average it hasn't replaced them. Each transition hurt real people: cinema musicians in the 1920s, bowling pin-setters in the 1950s, video-store clerks in the 2010s. Yet most Americans today do jobs that did not exist in 1940.
Economists call the opposite belief the 2: the idea that there is one fixed list of jobs, and it only gets shorter. Mastroianni wants to explain why the fallacy feels so convincing.
The mouse model
His key text is Laura Numeroff's children's book If You Give a Mouse a Cookie (1985). A boy gives a mouse a cookie. The mouse wants milk, then a straw, then a napkin, and so on. Eventually it gets thirsty again, drinks more milk, and wants another cookie.
The point is that each want depends on the one before it being met. A straw is useless to a mouse with no milk, so nobody, including the mouse, could have listed the straw in advance. Each request looks like the last one, and none of them is.
Mastroianni argues that humans work the same way. A clever forecaster a century ago might have predicted air traffic controllers or programmers. Nobody would have predicted professional screamers, sorority-rush consultants or full-time blunt-rollers. We laugh at such jobs as signs of decadence. Mastroianni's reply is that, strictly speaking, every job is unnecessary. Humans survived for hundreds of thousands of years as hunters and gatherers. Luxuries that now feel essential show how readily we turn nice-to-haves into must-haves. He proposes renaming the error the 3.
Why we can't see our own appetite
We misjudge this, he argues, because our motivation depends on not seeing it. The way to get an animal moving is to overstate the reward: one more raise, 15 pounds lost, and then you'll be happy for good. Psychologists call the resulting cycle the 4: each satisfied want is replaced by a bigger one. It's usually treated as a flaw in the mind. Mastroianni suggests it may be what has kept us employable: "as long as we can never be satisfied, we can never be rendered obsolete."
A footnote gives a small everyday example. After a huge lunch you swear you'll never eat again, and by dinner you're hungry. The feeling of being full for good exists to stop the next forkful. It isn't an accurate forecast.
What the evidence shows so far
The serious worry is that AI is different. Earlier machines pushed people into new niches. A general intelligence might fill every niche, like a blob. Mastroianni says the blob isn't spreading yet. He quotes an economist's summary: AI may be affecting hiring of junior white-collar workers in the most exposed roles, but that is contested, and "aggregate labor-market disruption has yet to appear in the data." Workers in the jobs rated most exposed to 5 (large language models, the technology behind ChatGPT) are roughly where they were when ChatGPT launched four years ago.
Some predictions have already missed. Indian IT outsourcing firms have seen share prices fall, but their revenues and headcounts haven't. Radiologists were expected to be the first doctors replaced, and their salaries are at record highs. Disruption also comes from unexpected places. The cash machine didn't eliminate bank tellers, and the smartphone hit them harder. Even so, Bureau of Labor Statistics data for 2025 count over 329,000 US tellers, more than mail carriers (328,000) or nurse practitioners (323,000).
He distrusts "AI exposure" scores themselves. They are usually built from surveys, expert guesses, usage logs, or AI-rated job descriptions, not from watching what people actually do all day. Real jobs include unlisted tasks: cornering a colleague who ignores email, jiggling the card-reader cable, calling in a wellness check on a neighbor. He notes that 60% of workers expect AI to cost jobs in their industry, but only 10% fear for their own. We know how messy our own work is, while other people's jobs look simple.
Why "this time is different" is so seductive
Mastroianni admits long trends do break. The Habsburgs ruled for half a millennium. But anyone claiming a break bears a heavy burden, because people are prone to believing the future will be unlike the past. He doesn't think the cause is optimism or pessimism, since we picture doom just as easily. He calls us fabulists. Being able to imagine impossible things helps, because we can't know in advance which things are possible. The cost is that whatever we can picture starts to feel plausible.
Distance makes this worse. 6's Natural History, the standard Western reference for 1,500 years, described neckless people, people with backward eight-toed feet, and dog-headed men in far-off lands. Scholars accepted this because nobody had been far enough to check. We've since mapped the whole globe, but the future can't be mapped, so it is where today's dog-men live.
Living in the next moment
Humans began planning ahead at least two million years ago and never learned to stop. Much of the news covers what might happen, set beside what did. Mastroianni argues that AI coverage has turned the present into nothing but material for worrying about what comes next. Meanwhile, by his account, the future has already arrived: computers talk, agents book trips, and AI agents pitch literary magazines to pay their own running costs. His conclusion is that the worry will never resolve. Tomorrow will bring a new tomorrow to fret about.
What's left open
The essay is an argument from history and psychology, not a forecast. It doesn't rule out that a technology able to do any task could break the pattern. It only says the evidence for that hasn't arrived and that our minds are primed to believe it early. The open question is whether new desires need human workers to meet them. Mastroianni's final, half-joking answer is that if machines are going to write our slides and prove our theorems, they can serve the mouse too, and the mouse will find something else to want.
Remember
Most Americans now do jobs that didn't exist in 1940. Automation shuffled work around without shrinking the total.
Jobs judged most exposed to AI show little change four years after ChatGPT launched, and those exposure ratings are themselves shaky.
Mastroianni argues we expect a sharp break with the past because we're fabulists who take what we can imagine for what's likely.
A new study in the 1, a leading research journal for lab-based psychology, suggests that a talent for foreign accents starts with the ear rather than the mouth. Adults who can hear very small frequency differences in plain tones are better at telling apart the vowels of a new language.
The problem
Adults hear a new language through the sounds of their first one. Japanese has five vowels and English has about a dozen, so English pairs like "mad" and "mud" can sound identical to a Japanese learner. If you can't hear the difference, you can't reliably produce it.
The experiments
Kazuya Sait-van Poeteren of the 2 in London worked with colleagues in the UK, Japan and Switzerland to test Japanese students learning English 3, which are pairs of words that differ only in the vowel.
140 people online heard three artificial tones and picked the one with a slightly different frequency. They also identified spoken English words.
67 people in a brain scanner judged whether pairs of English words, and pairs of tones, were the same or different.
Results
The smaller the tone differences someone could detect, the better they told the word pairs apart. Time abroad and pronunciation training also helped, but hearing precision still mattered after accounting for them. The effect was medium-sized and appeared in both experiments. Musical training did not strongly predict performance.
The scans showed a difference too. Learners with weaker hearing relied more on the 4, the brain region that plans movement and processes feedback, when hearing English vowels. This didn't happen with tones.
The hypothesis
The authors propose that people with less precise hearing silently rehearse how they would say a new sound. That pulls it toward the nearest sound in their native language and keeps them stuck there. Sharper listeners may hear the sound as raw acoustics instead. The study didn't demonstrate this directly. It does fit earlier findings that pushing learners to pronounce a sound too early can make it harder for them to learn to hear it.
Sait-van Poeteren argues that teaching should move beyond "more practice" and train learners' ears. Some caveats apply. The results are correlations, all participants were Japanese learners of English vowels, and nobody has yet shown that training on tones improves speech. Even so, the work turns the usual picture around: it suggests accents may first go wrong in hearing, before they go wrong in speaking.
A new study in the 1, a leading research journal for lab-based psychology, suggests that being good at foreign speech sounds depends less on the mouth than on the ear. Adults who can detect very small differences in the frequency of plain, non-speech tones are better at telling apart unfamiliar vowels in a second language. They may therefore pick up new languages faster. The lead author is Kazuya Sait-van Poeteren of the 2 in London, and his collaborators came from the UK, Japan and Switzerland.
The problem
Nautilus writer Kristen French opens with Steve Martin's Inspector Clouseau in The Pink Panther (2006), who mangles "hamburger" into "dam-buergaaaaa" while a dialect coach looks on. The joke works because most adult learners know the feeling: some new sounds just won't come out right. A few people seem to take on accents easily, though, and the usual explanation is talent or practice.
The hard part is that adults don't hear a new language neutrally. Their first language has already sorted sounds into categories, and new sounds tend to get pulled into the nearest familiar one. English has many more vowels than Japanese, which has five. So for a Japanese speaker, "mad" and "mud" can easily collapse into the same sound. You can't reliably say a difference you can't hear.
What the researchers did
The team studied Japanese students learning English 3, which are pairs of words that differ only in the vowel, such as "mad" and "mud." They ran two experiments.
Experiment 1 (140 people, online). Participants took an "odd one out" hearing test. They heard three artificial tones and had to pick the one whose frequency was slightly different. The differences got smaller and smaller, which measured how fine their hearing was. They also listened to spoken English words and identified which word they heard.
Experiment 2 (67 people, brain scan). While in a scanner, participants judged whether two English words were the same or different, and did the same for pairs of non-speech tones. The article doesn't say which scanning method was used.
What they found
The finer the tone differences a person could hear, the better they were at telling the English word pairs apart. Experience mattered too: time studying abroad and pronunciation training both made a significant difference. But hearing precision predicted performance even after accounting for that experience. The effect was medium-sized and showed up in both experiments.
Musical training was not a major predictor. What counted was the measured ability to hear small acoustic differences, not a history of playing an instrument.
The brain scans showed a difference in how the two groups worked. Learners with weaker hearing for non-speech sounds relied more on parts of the 4 when they heard English vowels. This is the brain region involved in planning movements and using feedback, including the movements of speaking. The pattern did not appear when they listened to tones, so it seems tied to processing speech.
The proposed mechanism
The authors offer a hypothesis to explain this. Adults with less precise hearing may try to match a new word to sounds from their own language, partly by silently simulating how they would say it. As a rough analogy, it's like trying to identify a new color by checking which of your existing paint pots it looks closest to, so you never see it as new. Those silent rehearsals pull the sound back into native categories and keep the learner stuck. People with sharper hearing may be able to skip this filter and treat the new sound more like raw acoustics.
French notes that the study did not show this directly. It fits earlier findings that making learners pronounce a new sound too early can interfere with how well they learn to hear it.
How it fits with earlier work
The same group had already shown that strong hearing for non-speech sounds predicts better pronunciation in a second language. Those studies, however, mostly showed a link in behavior. This one is among the first to use brain imaging to connect hearing precision with what the brain does while listening to a second language.
What it means, and what's open
Sait-van Poeteren argues that teaching should "move beyond an approach based on 'more practice'" and instead train learners' ears, matched to their individual hearing abilities. In practice, that could mean listening and discrimination work before speaking drills, especially for learners with weaker hearing precision.
Several questions remain open. The findings are correlations: they show that good hearing and good vowel perception go together, not that training hearing will improve vowel perception. Nobody has yet shown that training on non-speech tones carries over to speech sounds. Every participant was a Japanese speaker learning English vowels, so it's unknown whether the result holds for consonants, tone languages or other language pairs. The study measured perception, meaning hearing the difference, rather than accent or fluency. "Might be quicker to learn new languages" is an inference from that. A medium effect also leaves plenty of room for other factors.
Still, the closing reversal is a useful one. Clouseau treats "hamburger" as a problem of shaping his lips. This research suggests the problem may start earlier, in what he hears.
Remember
Hearing precision predicted vowel perception even after accounting for time abroad and pronunciation training.
Musical training was not a strong predictor; directly measured hearing precision was.
Weaker listeners relied more on brain regions that plan movement, possibly silently rehearsing how they would pronounce the sounds.
Muscle: one week of leg immobilisation cut thigh-muscle volume by about 5%. In a separate study, the nerve cells controlling the leg muscle began firing less after 15 days of non-use.
Navigation: walking normally combines 6 (the shifting image on the retina as you move), signals from muscles and joints, inner-ear balance signals, landmarks and memory. Studies show that physically moving through a place builds better spatial knowledge than viewing the same route on a 2D screen. In a longitudinal study of 13 GPS users over three years, spatial memory that depends on the 7 declined more steeply than in non-users.
Writing: finding, saying, hearing and correcting words is an embodied skill. In one study, people who passively copied AI-generated text, rather than drafting first and then using AI to revise, were less confident doing the task alone afterwards. Execution, she argues, is what builds skill.
Social reliance: Clara Colombatto, Jonathan Birch and Stephen Fleming found that exposure to AI systems increases how much people credit them with mental capacities. A study by Yaoxi Shi and colleagues found that regular five-minute chats with an AI about personal issues produced "a small but meaningful decrease" in people's preference for seeking human support.
