Last week, Cautionary Tales described the rise and fall of General Magic, the startup with too much money and too few constraints. Was General Magic really doomed from the start, and what does its fate tell us about AI, drug discovery and our love lives? David Epstein, author of Inside The Box: How Constraints Make Us Better, joins Tim to make the case for loving limitations.

[Apple] [Spotify] [Stitcher] [Patreon]

Other accounts suggested that when the fruit ripened, the result was a full-fledged lamb, or something very like it, attached to the stem of the plant by an umbilical cord. The lamb would graze near its roots, but when all the grass within reach had been consumed, the Vegetable Lamb would die.

Readers with a solid grasp of botany may already be aware that there is no such thing as a Vegetable Lamb. Indeed, there was likely no such person as Sir John Mandeville — his celebrated travelogue was written by persons unknown, and based heavily on previous works, including that of Odoric.

Yet this did not stop stories about the Vegetable Lamb from circulating, and being believed, among the cognoscenti of Europe for centuries. Baron von Herberstein, an ambassador to Russia nearly two centuries after the Mandeville travelogue became popular, confessed to being sceptical before becoming a believer after many “people of credence” told him they’d seen and tasted the delicious plant-animal hybrid.

“It took nearly four centuries after Mandeville’s writings appeared for European botanists and biologists to recognise the Vegetable Lamb as a myth,” write Cailin O’Connor and James Owen Weatherall in The Misinformation Age (2019).

Does that sluggishness reflect the failure of a particularly medieval mindset, or does it have lessons to teach us today?

Consider some slightly more recent cases. James Lind, a Royal Navy surgeon, is celebrated today for discovering that citrus fruit cured scurvy, after running a primitive clinical trial in 1747. The only trouble is that he did no such thing.

Lind conducted a trial all right, but it made next to no difference to anyone’s perception of citrus fruits. Both before and after Lind’s trial, they were thought by some sailors to be helpful, but Lind’s work made little impression. His Treatise of the Scurvy (1753) is a mass of confusion, although the brief description of the trial nearly 200 pages in is clear enough. Lind ended up recommending a boiled citrus syrup, and this proposal was widely ignored — which is no great shame, since boiling lemon juice destroys vitamin C and thereby renders it useless as a cure for scurvy.

Citrus fruit was embraced by the Royal Navy half a century later, and not because of a clinical trial but because it was recommended by a high-status doctor, Sir Gilbert Blane. Still later, the “limeys” lost confidence in the cure. Stale, badly stored lime juice contains almost no vitamin C but, in the age of steam, ordinary sailors rarely got scurvy anyway because voyages were quick and regular refuelling stops were required. Polar expeditions were another matter, and in the late 1800s were frequently struck by scurvy despite the use of (stale) lime juice.

At the same time, germ theory was emerging and the idea that scurvy was something to do with citrus fruit began to seem quaint. As a result, Robert Falcon Scott was advised by medical experts that scurvy was caused by toxins called ptomaines, and his Antarctic expeditions were afflicted by scurvy more than 150 years after Lind’s “discovery”. Ptomaine poisoning is no more a real explanation for scurvy than the Vegetable Lamb is a real animal.

Ninety-nine years ago, the Korean newspaper Jungoe Ilbo published a story titled “Strange Harm from Electric Fans”, warning that having a fan running too close to you might cause death as you choked on your own carbon dioxide. (It won’t.) For decades, Korean newspapers continued to run articles about people dying in the presence of fans, whether or not there was any reason to believe the fan was to blame.

Why do we believe things that aren’t true? Two fashionable views are that we believe falsehoods because of political polarisation, or that we believe them because of emotional triggers such as fear, lust or greed. That explains why the political party faithful are willing to believe almost any smear about the hated enemy, while closing their eyes to wrongdoing on their own side — and why people will send money to strangers on the internet claiming to be Nigerian princes.

But neither tribalism nor greed explains the beliefs that your desk fan could kill you, that citrus fruit doesn’t prevent scurvy and that somewhere in Asia lambs grow on trees. Nor do they explain why these false beliefs persisted for so long.

For O’Connor and Weatherall, the point of the Vegetable Lamb story is that most of what we believe, we believe because everyone around us believes the same thing. From the belief that fresh vegetables are good for you to the idea that the Earth is basically spherical, our worldview reflects the worldview of the society in which we live.

The world can be a confusing place. Seeing white fluff bursting from cotton plants for the first time, it is not surprising that Europeans perceived a resemblance to lamb’s wool. Sailors on fast-moving steamships derived no antiscorbutic benefit from stale lime juice — and needed none.

When reality supplies us with signals that are mystifying (white fluff on a plant) or inconsistent (with or without lime juice, scurvy may or may not develop) or long-delayed (cigarettes dramatically raise your risk of lung cancer) it is hardly a surprise that we turn to each other for information.

Or we may turn to the experts, who will sometimes have a clearer view of what is going on. (Not always: Scott’s advisers were fretting about ptomaine poisoning.) Yet we still have to decide who we should regard as having relevant expertise. That might be a scientist, or a YouTube influencer, or Wikipedia, or the output of a large language model — but in any case we will take cues from others as to who counts as “persons worthy of credit” and who does not.

“Most of us get our false beliefs from the same places we get our true ones,” write O’Connor and Weatherall. It is a sobering thought.

I draw three lessons. First, if most of our beliefs are handed to us by the society in which we live, it is vital that our society itself values truth and truth-telling. Second, it is never a bad idea to stop and think for a moment about what other people tell you.

And third, contact with unfamiliar views is healthy. What eventually punctured the Vegetable Lamb myth was that every time Europeans journeyed to Asia in search of the creature, the locals found the whole business hilarious.

Written for and first published in the Financial Times on 19 Aug 2026.

I’m running the Oxford Half Marathon in October in support of a very good cause and in memory of my mother. If you felt able to contribute something, I’d be extremely grateful.

It’s 1990. A new company called General Magic is going to build its vision of the future: the smart phone. They have talent, almost unlimited cash, and a 15 year head start on the iPhone. What could go wrong?

[Apple] [Spotify] [Stitcher] [Patreon]

Further Reading

This episode was based, with permission, on David Epstein’s new book Inside The Box: How Constraints Make Us Better. The documentary General Magic is available to watch on various streaming services.

Other sources include the PBS Documentary The Information Society, New York Magazine’s oral history of General Magic, Michael Kanellos’s article in Forbes magazine, and the YouTube channel Retro Roadshow.

It remains a seductive question, and one of the answers he offered was to use more behavioural science: “The best way to get someone to cut their electricity bill is to show them their own spending, to show them what their neighbours are spending, and then show what an energy-conscious neighbour is spending.”

Nobody should object to a judicious nudge like this, particularly since such approaches are easily tested with randomised experiments. But I’ve long argued that Cameron was just wrong to say that it was “the best way”. A tax on carbon-intensive energy would be better still: it would help shift the energy system to cleaner sources, encourage consumers to conserve energy more assiduously if it was from dirtier fuels, and raise revenue into the bargain. I’ve also argued that while nudges can be quick, easy and effective, serious policy action usually requires a change to regulations or the tax system. We shouldn’t become so enamoured of the icing that we forget to bake the cake.

But were my protestations based more on instinct than on evidence? Is it really true that a carbon tax has a bigger impact on behaviour than a cleverly formatted energy bill? At last, we have some data to answer such questions.

A new working paper from the economists John List, Matthias Rodemeier, Sutanuka Roy and Gregory Sun reviews more than 600 studies of different nudge-style interventions and asks the question: how significant are behaviourally informed public policies?

Start with the reformatted electricity bill. After reviewing such interventions around the world, the researchers conclude that the kind of tweak Cameron was applauding might reduce electricity consumption by something between 4 and 7 per cent. That doesn’t sound like much.

But who is to say that 4 per cent is a small saving? Small compared with what? One natural comparator is my preferred policy: a tax increase. How much would the electricity price have to increase to persuade people to cut back consumption by 4 per cent?

“The answer is actually quite a lot,” says Rodemeier. “You would have to raise electricity prices by 11 per cent . . . Try to go to the public and say, ‘We’re going to raise electricity prices by 11 per cent.’”

Fair. Cameron’s latest successor as UK prime minister, Andy Burnham, made a cut in VAT on electricity bills one of the very first policies he announced. There is not much point in columnists arguing that taxes on energy use are better than mere nudges, if politicians disagree.

In other areas, the best behavioural approaches are even more effective and the equivalent price rise would be even higher. The details vary. In the case of cigarettes, the most effective nudges have proven to be information about the risks of smoking. For vaccination, the best tactic is simply a timely reminder. The researchers reckoned that cigarette prices would have to increase by 28 per cent to match the impact of the best nudges. Alcohol prices would have to rise even further, by 34 per cent. As for a flu vaccine, the researchers estimate that a 100 per cent subsidy — making the vaccine free — is still not as effective at improving uptake as a reminder to get vaccinated.

So nudges are not just politically convenient. They induce a change in behaviour that is more substantial than we might assume.

When estimating the value of such nudges, it is useful to distinguish two yardsticks. The first is the cost-benefit ratio: for every dollar we sow implementing some clever behavioural tweak, how much social benefit would we expect to reap? The second is the total value of a policy intervention.

As an analogy, I might try to insulate my home by fitting some draught excluders, or by installing double glazing. The draught excluders look good on a cost-benefit basis because I might save 50 times what I spent. Still, if I want a noticeably warmer home with lower bills, draught excluders won’t do the job — it will have to be those costly new windows.

The same story appears to be true for behavioural public policy. The cost-benefit ratios of such ideas are often large, with tiny expenditures yielding outsized rewards. But the biggest total benefits emerge not from nudges but from more traditional policies such as appropriate taxes, subsidies and regulations. List, Rodemeier, Roy and Sun estimate that the optimal energy tax “generates roughly seven times the total welfare of the optimal nudge, even though the nudge is more cost-effective”.

So which approach is best? That depends on where the bottleneck is. A government with a limited budget to invest would do best looking for clever nudges here and there. If the binding constraint is political capital, space in the legislative agenda, or the attention of the prime minister, then a better bet is to pick a handful of traditional policy reforms and push them hard.

A well-functioning bureaucracy should be able to do both, of course. While political leadership is focused on major reforms, a bit of delegation goes a long way in implementing the small stuff. Whether the UK and the US currently have well-functioning bureaucracies is a separate question.

In any case, the two approaches complement each other. Take vaccination: a government keen to maximise the uptake of a vaccine would do well to reduce the price (zero is a particularly attractive price point, it turns out) but should also send out well-timed and well-phrased reminders. There is nothing mutually exclusive about these options.

Or energy policy: a government might favour a carbon tax to push every element of the energy system towards lower-carbon options, but some of those elements are likely to prove sticky. Bills that provide well-formatted information and a bit of peer pressure can help shake up consumer behaviour, and reforms elsewhere can help the whole system respond to the tax incentive.

If the choice is between the icing and the cake, politicians should choose the cake. Nudges are far from trivial in their benefits, but there is no substitute for using taxes to get price signals right. List et al write: “In four of five markets, price instruments deliver substantially larger total surplus than nudges.”

All too often, politicians are turning their back on the most beneficial policies because they lack the political courage or the political skill to introduce them.

Written for and first published in the Financial Times on 12 August 2026.

I’m running the Oxford Half Marathon in October in support of a very good cause and in memory of my mother. If you felt able to contribute something, I’d be extremely grateful.

This episode is available exclusively to Pushkin+ subscribers and members of the Cautionary Club.

[Apple] [Spotify] [Stitcher] [Patreon]

Further Reading

Protagonists on the Great Donorshow reminisce about the programme on BNNVARA and Villamedia, clips of the show can be seen on YouTube, and contemporary commentary comes from the BBC, NYT and BMJ.

Nora Kenworthy’s book is Crowded Out: The True Costs of Crowdfunding Healthcare. Ted Closson’s story about Shane Patrick Boyle can be found here. You can read more about Eliza on the Cure Sanfilippo Foundation website.

Evel Knievel, the world-famous daredevil, is planning his greatest ever stunt. In 1974, he wants to leap over Snake River Canyon in Idaho. The chasm is almost a third of a mile wide, much bigger than any jump he’s done before. To make it work, he’ll need a custom-made rocket-powered “Skycycle,” TV camera crews and – most importantly – a very positive attitude.

[Apple] [Spotify] [Stitcher] [Patreon]

Further Reading

The primary sources for this story were Evel Knievel: Life of Evel by Stuart Barker, the article Evel Knievel, Insurance Salesman by Jake Rossen and Think and Grow Rich by Napoleon Hill – which you may or may not wish to approach as a work of fiction.

Some are practical: “Make an exhaustive list of everything you might do and do the last thing on the list.” Some are gnomic: “A line has two sides.” Some are guaranteed to annoy most professional musicians: “Change instrument roles.”

The prompts are aimed at triggering useful conversations between members of a band, reminding producers of options or perspectives they may have forgotten and, above all, forcing skilled artists to abandon their comfortable clichés and find a new approach. My 2016 book, Messy, praises the cards and explains why they, and related annoyances, are such powerful stimuli to solve problems. Strange, then, that I almost never use them.

I’m not alone in that reluctance. When Eno persuaded David Bowie to embrace the cards while recording his celebrated Berlin trilogy of albums, Low, “Heroes” and Lodger, the musicians working with him did not always appreciate the intervention.

Carlos Alomar, one of the world’s great guitarists, found himself playing the drums for the track “Boys Keep Swinging”. (“Change instrument roles”.) Alomar later reflected, “I mean, some of it worked, some of it didn’t . . . it did make me leave my frustration at what I was doing and totally look at it from another different point of view and although I didn’t like the point of view, when I came back, I was fresh.”

The poet Simon Armitage described the cards “as if you’re asking the blood in your brain to flow in another direction”, which does not sound like a comfortable experience. But the world is full of uncomfortable obstacles that prompt a problem-solving response.

A favourite example of mine is the 2014 strike by workers on the London Underground, which disrupted journeys for 48 hours — and prompted tens of thousands of commuters to change their travel habits after discovering that they had been commuting wrong.

A 2009 laboratory experiment explored the impact of awkward team interactions: the researchers gave small teams 20 minutes to solve a multiple-choice murder mystery problem. In some cases, the groups consisted of four people from the same college sorority; in others, three sorority buddies were joined by a stranger, leading to annoying social friction. The researchers found that the teams with the stranger did a much better job at solving the problem, but enjoyed the task less and were more likely to believe they had failed. In short: the awkward group did better but believed they did worse.

From anxiety-provoking social situations to travel chaos to randomly drawn mystifications, we might call these problem-solving prompts “desirable disruptions”. The trouble is that while they may be desirable, they are rarely desired. Few people relish a Tube strike, or the awkwardness of small talk with a stranger — and perhaps that explains why the Oblique Strategies deck in my cupboard is rarely touched.

Which brings us to the topic of games, which, according to the philosopher Bernard Suits, are “the voluntary attempt to overcome unnecessary obstacles”. The contrast between that definition and these desirable disruptions is striking.

Finding a better route to work when your usual one is blocked is no game — it’s a reluctant attempt to overcome inevitable obstacles. So is the awkward attempt to solve a problem in discussion with someone you’ve never met and with whom you have no obvious common ground. As for the Oblique Strategies, they might be something close to a game if it’s your creative project and your deck of cards, but if you’re a hard-working musician and Bowie and Eno keep forcing you to participate in their stupid experiment, the experience is unlikely to feel playful.

Are games the missing ingredient here? There is plenty of reason to believe that games can motivate us to push ourselves in ways we would otherwise resist. One survey of the experimental evidence concludes that physical activity can be boosted by gamifications such as “points, levels, rewards, leader boards, narratives and teams”, a finding that will be familiar to anyone who has been drawn in by step trackers and fitness watches. The irony is that gamification is a brittle imitation of an actual game.

A good game can push us to remarkable feats of concentration, the mastery of complex information or bursts of physical effort that leave us gasping (or, once we reach a certain age, regretting our lack of warm-up). My own gaming activities include learning not only the rules but also some of the best tactics for a bewildering series of board games, and pushing through my shyness and introversion to play in long role-playing game sessions that are nothing more than intense social interactions in disguise. I’ve even been trying to do handstands, although I can assure you that standing on two feet is easier. None of this needless striving would be happening without the pull of the game.

The same is true of many artistic pursuits. If you want to capture someone’s likeness, you could pull out a smartphone and snap a photograph, or you could ask them to sit while you draw or paint. There’s no doubt that the phone camera is the easier option, but as the writer and creative guru Austin Kleon declares, “a good hobby should look a little exhausting”.

It is tempting, then, to argue for embracing the spirit of play in more of what we do. If we played with different possible routes to work, we wouldn’t need to wait for a Tube strike to discover a swifter or more pleasant itinerary. Working with strangers is hard, which may be why so many icebreaker exercises are actually games. And if I could treat the Oblique Strategies deck as more of a game than a capricious engine of difficulty, I might enjoy its strange gifts more often.

But let’s be careful. “It’s part of mankind to play games,” the great board game designer Klaus Teuber told me nearly 20 years ago, adding, “It’s an escape from the everyday grind.” There is much to be said for playfulness at work, but there is also a risk of turning our pleasures into chores. After all, if a game takes on a broader purpose then it risks no longer being fun. And a game that is no fun isn’t much of a game.

Written for and first published in the Financial Times on 5 Aug 2026.

I’m running the Oxford Half Marathon in October in support of a very good cause and in memory of my mother. If you felt able to contribute something, I’d be extremely grateful.

In 1942, eight men are sent from Nazi Germany to infiltrate America. Their mission: to sabotage critical infrastructure. Their problem: they’re idiots. Operation Pastorius soon devolves into one of the most baffling failures of World War II.

This episode was available oroginally released for members of the Cautionary Club – along with a monthly bonus episode, club members also receive a monthly behind-the-scenes video episode, newsletter, and ad-free listening.

[Apple] [Spotify] [Stitcher] [Patreon]

Further Reading

The tale of Operation Pastorius is told in They Came To Kill: The Story of Eight Nazi Saboteurs in America, by Eugene Rachlis, and Saboteurs: The Nazi Raid on America, by Michael Dobbs. The classic experiment of the smoke-filled room is Bibb Latane and John M Darley’s Group inhibition of bystander intervention in emergencies. And if you think you might have seen Joseph Schmidt, please contact J. Edgar Hoover.

This is a special crossover episode with Business History.

John DeLorean was a visionary who wanted to create a futuristic new sports car – a stainless steel marvel with distinctive gull-wing doors. The British Government agreed to finance the project, provided DeLorean built his factory in a troubled part of Northern Ireland. Production stumbled, sales fizzled and DeLorean liked to fiddle with the numbers. As the company approached ruin, a neighbour offered DeLorean a lifeline, if he’d only come in on a cocaine deal.  

[Apple] [Spotify] [Stitcher] [Patreon]

It was a viral hit, as was the follow-up book, Thing Explainer (2015), and it produced an intriguing reaction from the science-communication community, who are among the most enthusiastic of Munroe’s many fans. First, there was a burst of people trying to explain their own fields using the same constraint. Then came the backlash.

“Condescending”, wrote one science blogger. “By talking down to our audience, we risk alienating them, and reinforcing the common preconception that scientists consider and hold themselves apart from non-scientists.” The great science writer Carl Zimmer criticised the craze (not the cartoon) for encouraging the idea that there was anything useful in the constraint. The craze was, he argued, a disheartening waste of time for any scientist trying to figure out how to explain their field to a lay audience, because they might wrongly conclude that it was hopeless to try.

Well, maybe. Up Goer Five is tremendous fun but more puzzling than pellucid. The typical three-year-old can recognise 1,000 words, and there is nothing particularly clarifying about trying to explain something complicated while limiting oneself to the vocabulary of a three-year-old.

Nevertheless, unreasonable-seeming constraints can produce surprisingly powerful results. The artist and writer Theodor Geisel was once given the challenge of writing a book for six-year-olds while deploying no more than 225 different words from a short list. At first, Geisel was frustrated at the lack of adjectives; it was “like trying to make a strudel without any strudels”. But then, as David Epstein explains in his book Inside the Box: How Constraints Make Us Better (2026), he grabbed the first two rhyming words and went for it. The result, The Cat in the Hat (1957), made Geisel — better known as Dr Seuss — a phenomenon. Green Eggs and Ham (1960) is even better, and while it is packed with crazy situations, it contains only 50 unique words, making Up Goer Five seem positively lexiphanic. (Dr Seuss took on that challenge to win a bet.)

In the arts, the value of constraints is not a new idea. Monet almost never used black paint. Bach tied himself in knots with self-imposed rules, and part of the joy of his music comes from hearing him untie and retie them with ease. Miles Davis recorded Kind of Blue (1959) with no rehearsal, minimal composition and used the first take of each piece.

But when it comes to a more practical problem, is it really tenable to impose needless constraints? Perhaps it is. The simplest constraint for anyone trying to explain a complex topic is to accept the request to “explain it to me like I’m 10 years old”. What’s packed into that request is, vividly exaggerated, the reminder that what seems obvious to an expert is often incomprehensible to the listener. Technical jargon and fusillades of acronyms are often used to convey an explanation that has been denuded of all the foundations and context that might allow it to make sense.

Worse, in a kind of inverted Dunning-Kruger effect, the very expertise that makes the expert worth consulting also makes it hard for the expert to imagine what it was like to know nothing about the topic.

Hence “explain it to me like I’m 10 years old”. If you are reading this column you definitely aren’t a typical 10-year-old, but the image of the bewildered 10-year-old gives the expert an anchor to help them slow down and simplify. My book, The Truth Detective, was supposed to be a version of How to Make the World Add Up for 10-year-olds. My wife tells me she prefers it to the one that was for the grown-ups.

What is true in technical communication may also be true in design. David Epstein describes the moment the US Army tried to adapt its body armour for female soldiers. As Caroline Criado Perez forcefully argues in Invisible Women, women are not merely scaled-down men and are not well served by equipment designed for men. But as the US Army started grappling with that challenge, they discovered some surprising benefits. They were able to replace 11 sizes of body armour with just eight sizes by making the armour a modular mix-and-match system. This simplified manufacture and logistics while greatly expanding the options available to an individual soldier.

Meanwhile, many of the men in the army found the “female” modules a better fit — a narrow vest gave room to soldier a rifle, while a notch in the back of the helmet designed to accommodate hair buns allowed all soldiers, regardless of hair-do, to raise their heads while prone. “A new, more meticulous sizing process that benefitted women benefitted everyone,” Epstein writes.

A similar story can be told about designing for people without a full range of mobility or senses. In 1972, design professor Marc Harrison and his students produced a demonstration house full of features that accommodated users with special needs — flat thresholds between rooms made life easy for wheelchair users and people on walking frames; levers on taps and door handles are easier to use for people with a weak grip. But, of course, levers are easier for anyone to use — try twisting an old-fashioned doorknob with a cup of coffee in each hand.

Subtitles help people who can’t hear what the actors are saying — and it turns out that, for some gritty dramas, that is all of us. Web pages that are designed with screen readers for the blind in mind are typically better structured, simpler and friendlier for mobile devices. The Oxo kitchen appliance company was founded by a retired cookware entrepreneur whose wife had arthritis, but the chunky rubber handles are far more widely enjoyed. And of course, the dropped kerb may be life-changing for wheelchair users but it is pretty convenient for anyone pushing a pram.

The challenges faced by users who are unusually small, large, old or young, or who have disabilities, are, writes Epstein, “more extreme versions of the challenges that many other users face”.

Of course, we should not make the mistake of arguing that the reason we should design for disabled people is to help the majority, or that we should design for women because of the benefits for men.

But in a world where it is often hard to put ourselves in the shoes of someone else, the challenge of explaining the world to a 10-year-old, or designing a jar that an octogenarian can open, is sometimes just a prompt to do a better job of what we were trying to do in the first place.

Written for and first published in the Financial Times on 15 July 2026.