In 1952, a distinguished guest stole a solid gold George III salt shaker at a Buckingham Palace banquet. Churchill pocketed the matching pepper pot, approached the thief, and said "I think we've both been spotted. We probably better put these back." By presenting himself as a fellow offender rather than accuser, he changed the emotional response completely. This demonstrates how achieving an emotional response through psychological means is superior to rational confrontation.

The world is dominated by engineers, finance people, and accountants who focus on quantities—things with measurable amounts like money, speed, time, price, capacity. We spend vastly more effort solving problems through objective means than subjective means. The ratio is ridiculous.

Example: Britain spent £6 billion building a railway to reduce London-Paris train time from 3h 10m to 2h 40m. Could have achieved same desirability with £50 million of excellent Wi-Fi, changing the quality of time rather than quantity. Or employ supermodels serving champagne for £1 billion—people would ask for trains to be slowed down.

People love narrow engineering questions ("how can we make this faster?") because they lead to single right answers. They avoid creative open-ended questions ("how do you make the train journey so enjoyable people feel stupid going by plane?") because there are multiple answers and no one can claim to be definitively right.

Range anxiety contains two words: range (engineering/physics) and anxiety (psychology). Billions are spent increasing range through battery density. We should also spend millions reducing anxiety:

• Don't show battery going from 87% to 83%—just show "80%+" (specificity frightens people)
• Make charging stations highly visible (more neon, not just more lithium)
• Petrol gauges never showed that granularity

Electric car owners never go back, like air fryers—joining is painful but once done, you've joined a cult.

The US uses "feels like temperature"—a brilliant concept the UK lacks. How you feel is more important than ambient temperature. You're not performing chemistry experiments; you need to know what to wear. Human evolution makes us feel hot proportional to health risk. Heat plus humidity is more dangerous than heat alone; wind chill reduces risk. We've evolved an internal thermometer sensitive to survival, yet we use chemical thermometers that only tell part of the story. This metaphor applies to business—we measure what's numerically available rather than how people actually feel.

James Watt was selling steam engines to mine owners who used horses to drain mines. Engineers discussed boiler size, calorific capacity, bore length—nobody bought. Watt realized mine owners wanted to know how many horses they could get rid of. He invented "horsepower" as a unit. A 25-horsepower engine does the work of 25 horses, so you can get rid of 75 horses (they worked in shifts). Once presented in compelling, easily understood terms, behavior changed completely. Horsepower isn't named after a scientist because it's a marketing contrivance.

Multiple railway lines existed for decades but nobody used them—you can't navigate by looking at multiple maps simultaneously. They added these lines to the metro map and suddenly everyone understood the value. Carries as many people daily as a new line that cost £20 billion, but only cost £200 million (1% of cost). Achieved with pixels and ink, not rails and rolling stock. On first day when added to map, usage went up 4x. We can't buy something if we don't understand it.

Standard hotel booking: rooms graded by size (standard, deluxe, suite, etc.). Someone realized people want to be close to gym or pool. Created "pool access" designation—guests pay $20 premium. Didn't build walls, corridors, or elevators—just described rooms in different context. A billion-dollar idea for the hotel industry achieved with pixels, defining something differently. We don't know who came up with it because it's a marketing idea, not technological.

Airlines asked class preference on first page. If you said economy, you only saw economy price. This made buying premium tickets difficult even when not much more expensive. Perception of value is relative, not objective. You won't pay £200 extra for £500 total, but you will for £1,500 total—it's relative, not absolute. Showing prices side-by-side made £10 million incremental premium revenue annually, sustained for 15 years. High-margin revenue from a £25,000 UX change.

Business is dominated by chess-players who think rationally, but business is actually poker. Four or five hands determine whether you do well or badly all evening—you pick your moment and hit pay dirt. Jeff Bezos: "In baseball you can only hit four. In business you can hit 500." We aren't spending enough time looking for outsized effects. At American Express (15 years), four things made all the difference. "Member Since" on cards (suggested in the 60s) has been worth billions because nobody wants to cancel and go back to "Member Since 2025."

Marketers are held accountable for every penny spent but can only claim credit for revenue in the last 3 months. Like telling J.K. Rowling she only gets royalties on the first edition. Not proportionate evaluation of marketing's cost-benefit.

Speedometers show distance per time (mph). Two behavioral economists created a "pacometer" showing time per distance (minutes per 10 miles). This reveals something completely non-obvious: time saved by going faster decreases dramatically as you're already going fast. If going really slowly, extra 10 mph makes huge difference. If already going quite fast, don't bother—risk of accident, fatality, braking distance, energy consumption all going up exponentially while time saved is flatlining. Mathematically trivial but counterintuitive. How you show information completely affects how people respond.

Aldi's middle aisle (random items—go in for milk, come out with wetsuit) has discovered a fundamental truth: successful life requires two things. Walmart shopping (know what you want, pursue efficiently) and TJ Maxx shopping (swan around hoping to get lucky). You don't go to TJ Maxx looking for yellow cardigan—you go like a shark, swim around aisles until you find something. The ratio of effort toward exploiting what you know versus exploring what you don't know (increasing surface area exposure to positive upside optionality) is critical. Being famous increases chances of getting lucky—people you've never heard of bring business to you.

Bees do a "waggle dance" telling others where reliable pollen/nectar is. Most bees follow this information, but 20% ignore it and go randomly. Evolution doesn't tolerate inefficiency, so why hasn't this been eliminated? When modeled as complex system: without random bees, hives get trapped in local maximum and starve. Three problems: can't grow (reliant on existing sources), gets over-optimized on past (when environment changes, don't know where to go), can't get lucky (no upside). This is the "explore-exploit tradeoff" in algorithm design—not really a trade-off, but two complementary parts of same system (yin and yang).

Explore: High uncertainty, high payoff, fat-tailed distribution, many small bets with few outsized winners, poker-like, iterative experimentation, excel in uncertainty. Marketing and R&D are explore—shouldn't expect everything to succeed.

Exploit: Predictable returns, no downside risk, chess-like.

Problem: Exploit dominates. Creative people must present ideas to rational people for approval, but never the reverse. Complete asymmetry in corporate decision-making—rational people have veto power, creative people only have suggestion power.

What we notice is very small subset of what we see. Andy Clark's book "The Experience Machine" builds on William James's idea: most of what's in our brains is prediction. We reserve limited bandwidth for things that surprise us, things we weren't expecting. Like MPEG compression: every pixel has expectation value based on adjacent pixels; only use data for differences from expectation. Human brain evolved same data architecture as Samsung TV. Things you don't expect carry much more weight than things you do.

Will Guidara took Eleven Madison Park from #50 to #1 in world (San Pellegrino awards). Took team to #1 restaurant. Team looked for things to copy. Guidara: "I don't want to copy what they do well. Find something they're not doing well, then hit it out of the park." Two things: coffee and beer. Appointed coffee fanatic as coffee sommelier, kitchen guy as beer sommelier. In high-end restaurant, you expect "Yeah, we got Sam Adams on draft." Instead: craft beer list with tasting notes and pairing suggestions. Remembered for years because it's unexpected.

Water has zero taste so we're better at noticing anything that shouldn't be there. If water tasted like cherry Dr. Pepper, we wouldn't have heightened awareness of pollutants. Our taste buds calibrated for survival.

Find what others do badly, do it really well. Change frame of reference. Avis: "We're number two, so we try harder"—reframes small size as better customer care. Natural tendency to benchmark against competition doesn't help market (less diversity), businesses (direct competition), or consumers (lose differentiation). When businesses become less differentiated, everybody suffers.

Walt Disney: "The more you are like yourself, the less you are like anyone else, which makes you unique. Most people spend their lives trying to emulate others. Lots of copies but few originals."

John Maynard Keynes: "Worldly wisdom teaches that it is better for reputation to fail conventionally than to succeed unconventionally." No individual lemming gets singled out for blame. Do same dumb things as competitors, never get blamed.

Computing nerds optimized for processing power, RAM, clock speed, completely neglecting usability and aesthetics. In early 80s, putting a beige/gray computer in any room made it look like an office. Realized you didn't have to do that—things could be lovely to use and beautiful to look at. Classic reverse benchmarking.

Family-owned (most interesting companies are family-owned—don't justify to finance people, focus on customers). Entire focus: what's bad about gas stations? Restrooms. Who's the decider? Passenger, probably female. Strategy: get people willing to hold on 20 extra miles to go to Buc-ee's. Reverse benchmarking—take what everybody's terrible at, make it brilliant.

Sunday Times food hotel of year. Did something unprecedented: latest breakfast time is 11am (not 8am like most hotels). From that moment: "These are my people." Reverse benchmarking by spotting something bad nobody else noticed.

TRIZ: Soviet-era problem-solving toolkit. Mantra: "Your problem has already been solved by someone else in a different field." Car rental example: taxi home from airport costs £60—driver meets you, pushes trolley, loads luggage, drives you home. Rent car for week for £600: you find terminal, get keys, find car in unnavigable car park in 100-degree heat, no instructions on how to open fuel cap or operate interface. Often resort to YouTube next day. Car rental companies benchmark against each other, missing that many people would pay for escort to car.

Business assumes scarcity—can only do one thing. Must do both: exploit (efficient use of resources) and explore (continue seeking new value). Mike Smith at Decca Records turned down Beatles, signed Brian Poole and the Tremeloes instead. Tremeloes had two #1 hits, went on 15-20 years—not a bad band. The mistake: someone said "you can only sign one band." Tremeloes lived near London, Beatles in Liverpool (higher travel expenses—procurement's mistake). Nothing irrational about signing Tremeloes—the error was being allowed only one. When allowed only one, we always do the safe bet, never the risky wildcard.

AI will assault human employment in service industries. As person-to-person interaction becomes scarcer, its importance becomes disproportionately higher. Really good face-to-face experience is like a "brand quake." There are businesses he considers brilliant because of one phone call eight months ago—knows nothing else about them, but good telephone service places them high in the pantheon. Low expectations make it easy to impress.

Customer satisfaction figures didn't correlate with service performance. Areas with unreliable, late, hopeless mail—people loved Royal Mail. Areas with spectacular service—people indifferent. Tried correlating service quality with appreciation—made no sense. Discovery: vast majority of perception driven by one thing only—whether they like their postal worker (posty). Guy who leaves something round back because he knew you were away on holiday is total hero. Billions spent on service made no difference; one person with discretion to take intelligent contextual action is worth a thousand algorithms.

Rather than trying to change the world, change how people see the world. When people see the world differently, they behave differently, and that changes the world.