• General Overview
  • The Central Thesis: The Exponential Gap
    • Exponential technologies: improve over 10 percent yearly for decades, while institutions adapt only incrementally.
    • The exponential gap: the growing chasm between fast technology and slow social institutions structures every modern problem.
    • Two cultures: technologists and policymakers cannot understand each other, so no society thinks wisely.
    • Technology is not neutral: tools encode their makers' preferences and power structures.
    • Linear minds: humans evolved for slow change and systematically misjudge compounding growth.
    • Stakes: today's gap is existential, mediating health, work, government, and production.
  • How Technology Went Exponential
    • Moore's Law: a social fact, not physics — the industry treated a doubling target as obligation.
    • S-curve adoption: technologies spread slowly, accelerate sharply, then plateau — and now spread faster each generation.
    • Three drivers: learning by doing, recombination of technologies, and networks of information and trade.
    • Wright's Law: unit costs fall a constant percentage each time cumulative production doubles.
    • Standards: common interfaces let innovations combine like Lego blocks into new capabilities.
    • Point B: around the 2010s cheap solar, smartphones, and tech giants tipped the world into a new era.
  • Four Domains of the Exponential Age
    • Beyond computing: exponential cost declines now span energy, biology, and manufacturing too.
    • Solar and wind: costs collapsed roughly 500-fold, making renewables the cheapest power in most of the world.
    • Batteries: prices fall 19 percent yearly, solving renewables' intermittency and mothballing fossil plants.
    • Biology: sequencing fell a million-fold in under 20 years; synthetic microbes become factories.
    • 3D printing: additive manufacturing builds layer by layer, allowing customization without waste.
    • General purpose technologies: these reshape all of society, but only after a long installation phase.
  • The Unlimited Company
    • Superstar firms: network effects, platforms, and intangible assets break the old ceiling on corporate scale.
    • Winner-take-all: top firms capture nearly half of sales; 10 percent of public companies earn 80 percent of profits.
    • Capital efficiency: platforms orchestrate ecosystems while owning almost none of the assets.
    • Intangibles: 84 percent of S&P 500 value is now intangible, scaling without new factories.
    • Data network effects: every click improves the product, drawing still more users and more data.
    • Rebuilt antitrust: infrastructure power, interoperability, and utility duties replace consumer-price tests.
  • Labor's Loves Lost
    • No robopocalypse: employment held up; the real threat is job quality, not job quantity.
    • Moravec's paradox: computers master Go yet stumble on a one-year-old's perception and mobility.
    • Platform labor: gig work adds millions of positions but concentrates bargaining power in the platforms.
    • The digital panopticon: algorithmic management monitors, paces, and auto-fires workers at scale.
    • Declining labor share: workers' slice of national income fell across advanced economies since 1980.
    • A fair work settlement: dignity, flexibility, security, equity — plus renewed collective bargaining.
  • The World Is Spiky
    • Globalization's backlash: free trade peaked and nationalism returned, partly enabled by its own technologies.
    • Relocalization: additive manufacturing, vertical farming, and renewables let production move near consumers.
    • Developing-world rupture: local production could strip poor countries of export manufacturing jobs.
    • Cities as engines: agglomeration and serendipity concentrate innovation in dense urban clusters.
    • Splinternet: data localization and firewalls fragment a once-borderless web into national blocs.
  • The New World Disorder
    • Expanding attack surface: every connected device opens a vulnerability as defense lags offense.
    • Cheap warfare: cyberattacks and drones cost a fraction of conventional weapons and spare attackers exposure.
    • Stuxnet precedent: software alone destroyed centrifuges that once required an airstrike.
    • Misinformation: falsehoods spread six times faster than truth and are weaponized by 70 governments.
    • Autonomy: a third of weapons systems could select and attack targets without human control.
    • Responsibility gap: no legal framework assigns blame for autonomous or deniable digital warfare.
  • Exponential Citizens and the Path Forward
    • Code is law: private platforms, not parliaments, now write the rules governing public conversation.
    • Data self commodified: our digital doppelgängers are extracted, profiled, and sold back to us.
    • Homophily engineered: engagement algorithms push users toward extremes and fracture shared reality.
    • Four-part answer: transparency, interoperability, data rights, and digital commons return power to citizens.
    • Technology is not destiny: the same tools yield different outcomes in different societies — we choose.
    • Principles: commonality, resilience, and flexibility build a social settlement for permanent change.
• Deep Dive
  • Preface - The Great Transition
    • Preface - The Great Transition · I
      • A sudden historical transformation
        • Interwar London: farmland became modern streets in decades, showing technology's sudden social reorganization.
        • Everyday life transformed: cars replaced horses; electricity and telephones rewired homes and work.
        • Work reshaped: production systems and transport brought full-time contracts, familiar benefits, and the commute.
        • Ripple effects: inventions spill across jobs, wars, politics, manners, and habits.
      • Technology in complex systems
        • Complex systems: society's countless interacting elements make small changes ripple chaotically.
        • Feedback loops: technological shifts spiral into major, often unforeseen repercussions.
        • Phase transitions: like water to steam, societies can abruptly reorganize; today is such an inflection.
        • Historical discontinuities: Columbus and the Berlin Wall mark world-changing societal phase changes.
      • The exponential present
        • Exponential pace: computing, AI, renewable energy, storage, biology, and manufacturing accelerate month by month.
        • Public unease: over 60% in Edelman's 2020 survey felt technology moves too fast, up a fifth in five years.
        • Metamorphosing industries: new innovations alter relations between firms, workers, cities, citizens, and markets.
        • A hopeful view: turmoil comes first, but humans adapt and eventually learn to thrive.
      • Two flaws in our tech conversation
        • Neutrality myth: Silicon Valley claims technology is neutral, but tools encode makers' preferences and power structures.
        • Design reflects power: phones fit men's hands; drugs often work less well on Black and Asian patients.
        • Entrenched bias: encoded power becomes infrastructure that is inscrutable and less accountable than humans.
        • Political ignorance: Brexit deal called defunct Netscape Communicator a modern email package.
        • Two-world literacy: C. P. Snow's divide; sound analysis must straddle technology and society's institutions.
    • Preface - The Great Transition · II
      • The Two Cultures Problem
        • Two cultures: technologists and everyone else, with a gulf of mutual incomprehension.
        • Stakes: when cultures grow apart, no society can think with wisdom.
        • Technological culture races ahead; humanities and policymakers cannot keep track.
        • Author's bridge: a career spanning journalism, startups, investing, and social science.
      • The Exponential Gap Thesis
        • Exponential technologies: invented and scaled faster while dropping rapidly in price.
        • Institutions: political norms, economics, relationships adapt slowly, along a linear path.
        • Exponential gap: chasm between fast technology and slow social institutions.
        • Consequence: questions of work, conflict, politics, and civil society demand new answers.
      • Roadmap of the Book
        • Part one: explains exponential technologies and the arrival of the Exponential Age.
        • Economy: superstar firms dominate winner-takes-all markets; employees face gig work.
        • Political economy: re-localization of production and energy reshapes global conflict.
        • Citizen and society: state-sized companies and data economy erode cherished values.
        • Solution: resilience, collective ownership, and decision-making can close the gap.
      • Lessons from the Pandemic
        • Exponential growth creeps up then explodes—humans struggle to conceptualize its speed.
        • Tech dependence: lockdowns, vaccines, and connectivity reveal embedded exponential tools.
        • Feedback loop: technology reshapes politics, economics, culture; each reshapes technology.
        • Provisional map: the book aims to reveal terrain rather than predict the future.
  • Chapter One - The Harbinger
    • From ZX81 to Moore's Law (Chapter One - The Harbinger · I)
      • A Childhood of Rapid Upgrades
        • First encounter: a build-it-yourself computer kit in 1979 Lusaka, Zambia, brought the digital revolution into reach.
        • ZX81: compact £69 machine of 1981, capable in principle of any computation, yet obsolete within a few years.
        • BBC Master: six years later, 128 times the memory, multicolor graphics, and room for spreadsheets and games.
        • PC clone: by 1991, Intel 80486 and 4MB RAM delivered millions of times more capability than the ZX81.
        • Microsoft-Intel symbiosis: Windows demanded more power, Intel supplied it; “What Andy giveth, Bill taketh away.”
      • The Machinery Underneath
        • Boolean logic: George Boole reduced reasoning to binary digits, representable as 1 and 0.
        • Electronic switching: Claude Shannon showed circuits could implement Boolean logic, enabling electronic computers.
        • Transistors: postwar Bell Labs semiconductors made switches smaller and more reliable than vacuum valves.
        • Integrated circuits: photolithography etched many tiny transistors onto silicon chips, lowering cost and raising speed.
      • Moore's Law as Social Fact
        • Moore's observation: in 1965, Gordon Moore predicted chip power would double for the same cost every 18–24 months.
        • Not a physical law: descriptive, not predictive; the industry treated it as a target and made it true.
        • Exponential scale: transistors per chip multiplied nearly ten-million-fold from 1971 to 2015.
        • Collapsing price: a transistor fell from $150 in 1958 to a few billionths of a dollar by 2014.
        • Computing ubiquity: from 264 machines in the 1950s to over five billion computers, including smartphones.
      • Defining Exponential Technologies
        • Core definition: a technology improving at over 10 percent per year for several decades.
        • Compounding power: 10 percent yearly yields 2.5x capability per decade and cuts cost by three-fifths.
        • Exclusion: diesel engines improved fast briefly, then stalled; chips sustained ~50 percent annual gains for fifty years.
        • Dual effect: falling prices put technologies everywhere while rising power enables genuinely new uses.
      • The Shape of Adoption
        • S-curve spread: technologies diffuse slowly, then sharply accelerate, then taper as markets saturate.
        • Market limits: saturation eventually flattens adoption; households cannot absorb unlimited copies of any product.
        • More than one per home: Bill Gates’ “computer on every desk” went far beyond that; households now hold many devices.
        • Dediu's data: two centuries of US technology adoption reveal repeated logistic curves across innovations.
    • Acceleration, Limits, and New Paradigms (Chapter One - The Harbinger · II)
      • The Accelerating Spread of Technology
        • Generational contrast: born in 1920 saw stable tech; Moore's Law generations experience constant acceleration.
        • Saturation speed: social media took 14% of an average lifespan; electricity took 62%.
        • Early-century baseline: telephone, power, and auto required 30+ years to reach 75% of US households.
        • Digital diffusion: exponential technologies now reach 75% of homes in 8–15 years.
        • Hyperdeflation: plunging computing costs make new products possible that spread faster.
        • Harbinger: computing's breakneck acceleration foreshadows the broader Exponential Age.
      • Proof in Consumer Tech
        • Social media lineage: SixDegrees, Friendster, LinkedIn, and MySpace preceded Facebook.
        • Facebook's trajectory: first million users in 15 months; 2.5 billion users by 2019.
        • Lime's speed: six months to one million rides, seven more to ten million.
        • KakaoBank's launch: 4% of Koreans joined within two weeks; 20% by 2019.
        • TikTok's surge: unheard-of app became world's most downloaded in months; revenue quintupled in two years.
        • Contagious acceleration: each new digital generation spreads faster than the last.
      • Moore's Law and Successive S-Curves
        • Social fact: Moore's Law persists because the semiconductor industry obsessively engineers it.
        • Physical limits: heat leakage and quantum effects threaten transistors only atoms wide.
        • Escalating effort: research outlay rose 18-fold; fab costs rose ~13% yearly.
        • S-curve arc: each technology moves through slow start, explosive growth, and plateau.
        • Accelerating returns: Kurzweil's interacting S-curves keep overall progress exponential.
        • Paradigm handoff: technologies nourish each other; as one S-curve flattens, adjacent innovations take over.
      • Artificial Intelligence and Quantum Leaps
        • AI inflection: enough data and computing power finally released machine learning's stalled promise.
        • ImageNet: 14 million hand-annotated images supplied deep learning's needed training data.
        • AlexNet: 2012 neural net with 60 million parameters lifted ImageNet accuracy to 87%.
        • Compute explosion: AI training compute rose 300,000-fold in six years, far outpacing Moore's Law.
        • Specialist hardware: graphics and AI-specific chips sustain exponential gains without miniaturization.
        • Quantum promise: qubits model values between 0 and 1; Google's prototype was a billion times faster.
  • Chapter Two - The Exponential Age
    • Energy, Biology, Manufacturing Go Exponential (Chapter Two - The Exponential Age · I)
      • The Exponential Age’s Four Domains
        • Beyond computing: exponential cost declines now span energy, biology, manufacturing, and computing.
        • Key technologies fall by a factor of six or more every decade.
        • Drivers: growing demand, innovation recombination, and expanding networks of trade and information.
        • Result: a wholly new era of human society.
      • Solar: From Fantasy to Cheapest Power
        • Oil shock: tripled prices; Bond’s Solex Agitator embodied solar hopes.
        • Solar in 1975: $100 per watt—only satellites, watches, calculators.
        • Price fell ~500x to under 23 cents per watt by 2019.
        • By 2020, large-scale solar undercut the cheapest fossil fuel generation.
        • Today, solar is the cheapest electricity in two-thirds of the world.
      • Wind and Batteries Complete Energy Revolution
        • Wind costs fell 70% in the decade to 2019, about 13% yearly.
        • Batteries fell 19% annually from 2010; large storage near grid parity.
        • Intermittency challenge: renewables generate but don’t store energy; batteries are the missing piece.
        • Fossil plants mothballed as cheap renewables accelerate displacement.
      • Biology’s Exponential Unfolding
        • First genome cost ~$500M–$1B; by 2020 BGI quoted $100.
        • Million-fold improvement in <20 years—halving every year, beating Moore’s Law.
        • Cheaper sequencing drove explosive use: hundreds of thousands of genomes yearly.
        • Synthetic biology: microbes engineered as factories for chemicals, materials, medicines.
        • By 2040, 60% of physical inputs could be biological.
      • Manufacturing: From Subtractive to Additive
        • Manufacturing unchanged since human ancestors: subtractive chiseling of matter.
        • Casting and molding avoid waste but can’t customize without new molds.
        • 3D printing builds layer by layer—customization without the waste.
        • Printing advances to steel, ceramics, organs, even a Dubai office.
        • Exponential trajectory: market grew 11x by 2019; performance to improve 14x next decade.
      • General Purpose Technologies Reshape Society
        • GPTs are special: broad utility across sectors, unlike stirrups or light bulbs.
        • Historic GPTs—printing, electricity, car—transformed economies and everyday life.
        • Current wave: multiple GPTs emerging simultaneously in four domains.
        • GPTs take time: electricity’s infrastructure began in 1881, productivity gains in the 1920s.
        • Installation phase: skills and organizational change must mature before impact.
    • Deployment, Learning, and Combinatorial Growth (Chapter Two - The Exponential Age · II)
      • Deployment and Cascading Technologies
        • Deployment phase: installation gives way to rollout once firms grasp a GPT's purpose.
        • Golden age: economies enjoy a GPT's benefits only after long installation and learning.
        • Accelerated rollout: cloud and smartphones are already deployed, so transformation may come faster.
        • Cascading technologies: computing enabled niche products and new corporate strategies; PCs, internet, and smartphones followed.
        • Economic ripples: smartphones displaced cameras and maps; mobile commerce reached billions in sales.
      • The First Driver: Learning by Doing
        • Why now: three forces drive exponentiality; the first is learning by doing.
        • Moore's Law's limit: time-based description cannot explain progress; if production stops, it stops working.
        • Wright's Law: unit costs fall by a constant percentage each time cumulative production doubles.
        • Learning by doing: engineers and workers find efficiencies through repeated production.
        • Demand spiral: greater demand drives production, learning, and lower costs, which fuels more demand.
        • Broad evidence: applies across technologies, including battery prices and chip prices.
      • The Vanishing Limits of Wright's Law
        • Old S-curve: Wright's Law once tapered as markets saturated, halting price declines.
        • Physical scaling: smaller chip components square efficiency; longer blades quadruple wind-turbine output.
        • Global markets: world trade grew sixty-fold, pushing saturation farther and extending learning effects.
        • USB storage: $50 bought 8 MB in 2000, then 2 TB in 2020—250,000x more for the same price.
        • Self-reinforcing loop: more production creates more demand, which drives more production.
      • The Second Driver: Combination
        • Combinatorial invention: GPTs amplify each other in novel, constantly shifting patterns.
        • Energy Vault: cranes, rubble blocks, generators, shipping, and machine vision combine into a gravity battery.
        • Machine vision's role: automated deep-learning control removes human operators, making storage competitive.
        • Old combinations: Ford joined electricity and automobiles; medieval pumps used flywheels and cranks.
        • Standard components: AA batteries and similar standards make complex products cheaper and faster to build.
      • Standardization Multiplies Combination
        • Standards as enablers: common interfaces let innovations combine like Lego blocks into new services.
        • Bottom-up standards: internet RFCs and web protocols came from researchers, not top-down bodies.
        • Interoperability payoff: standardized email and web systems let diverse products work together seamlessly.
        • Consensus standards: international bodies eliminate compatibility headaches for manufacturers.
    • Standardization, Networks, and Political Context (Chapter Two - The Exponential Age · III)
      • Standardization Unlocks New Capabilities
        • Cheap Earth observation: accurate satellite images now cost tens of dollars and guide hedge funds, traders, and insurers.
        • Satellite revolution: launches rose from about ten a year in the 1990s to 372 in 2018, thanks to standard CubeSats.
        • Software componentization: developers assemble ready-made code blocks, so apps build on components from many industries.
        • Combination effect: modern technologies combine and recombine, producing continuous waves of innovation.
      • Networks of Information and Open Science
        • Laura O'Sullivan: a 16-year-old built a cervical-smear anomaly detector that beat a human doctor.
        • GANs in reach: a four-year-old breakthrough, downloaded from GitHub, trained on her father’s consumer PC.
        • Preprint servers: arXiv, and later bioRxiv, PsyArXiv, and SocArXiv, spread cutting-edge research freely and quickly.
        • Open access shift: removing paywalls lets outsiders join science; COVID-19 produced over 84,000 open papers by late 2020.
        • Collaboration tools: GitHub, Behance, Wikipedia, and social networks connect millions of people sharing know-how.
      • Networks of Trade and Logistics
        • Containerization: standardized shipping containers, first loaded in 1956, transformed ports and made global trade seamless.
        • Cost plummets: sea freight fell 2.5 times after WWII, then halved again once containers dominated shipping.
        • Scale explosion: container-ship capacity grew 25-fold from 1980 to 2015; port traffic tripled between 2000 and 2018.
        • Just-in-time supply chains: internet ordering and container shipping let firms like Apple hold fewer than ten days of inventory.
      • The Political-Economic Turn
        • Three drivers: learning by doing, technology combination, and information/trade networks reinforce each other.
        • Friedmanite orthodoxy: from the 1970s, deregulation and profit maximization displaced interventionist postwar policy.
        • Reagan and Thatcher: their governments cut taxes and red tape, unleashing entrepreneurialism and new markets.
        • Globalization synergy: free-market economics expanded trade networks, helping kindle the Exponential Age.
      • Point B: The Exponential Age Begins
        • Origin at point A: the internet and microprocessor, around 1969–71, started a curve that stayed slow for decades.
        • Tipping point at point B: the 2010s brought mass smartphones, cheap solar, and tech giants overtaking oil firms.
        • Defining force: exponential technologies now rewire business, work, politics, and even our sense of self.
        • Human response: the disruptive power of the era lies in how people and institutions react, not only in tech change.
  • Chapter Three - The Exponential Gap
    • Exponential Technology Outruns Linear Institutions (Chapter Three - The Exponential Gap · I)
      • Amazon's Exponential Ambition
        • Amazon's scale: $213B retail plus $172B from cloud, logistics, media, and hardware.
        • R&D budget: $36B in 2019, grown from $1.2B a decade earlier at about 44% yearly.
        • Innovation mandate: CTO Werner Vogels says stopping innovation means "out of business in ten to fifteen years."
        • Old-world contrast: Tesco, with £50B annual sales, ran a research lab on a six-figure budget.
      • The Exponential Gap Defined
        • Exponential gap: Exponential technologies outpace institutions limited to incremental change.
        • Two trajectories: Adaptive firms harness accelerating tech; others fall behind fast.
        • The curve's crossing: Exponential line overtakes linear change and hits an inflection point.
        • Structural stability: Laws, norms, companies, and polities are built to adapt incrementally.
      • Deflationary Computation as Fuel
        • Computation deflation: Moore's Law halves cost every couple of years; ten years yields a hundredfold decline.
        • Strategic foresight: Organizations that plan for cheap future computation can experiment early.
        • Digital giants: Uber, Alibaba, Spotify, TikTok grew by recognizing computation deflation.
        • Forward-looking firms: Tesla, Impossible Foods, Spire, and Planet Labs bet on exponential cost declines.
        • Unadaptive victims: Newspaper publishing ignored the shift and had little chance.
      • Where Norms Lag New Technology
        • Economy: Industrial-era monopoly rules may miss how exponential firms undermine market dynamism.
        • Work: Gig platforms create vibrant task markets but challenge secure employment statuses.
        • Privacy: Companies mediate more of life; old safeguards suddenly look inadequate.
        • Everyday divergence: Consumers, workers, bosses, and citizens all face the exponential gap.
      • Why Human Intuition Fails Exponentials
        • Linear intuition: Minds evolved for slow, seasonal, pre-industrial change.
        • Exponential rain: Wembley's doubling raindrops show why linear reactions leave you drenched.
        • Growth bias: People underestimate compounding; Swedes guessed 410 rather than 761 kronor.
        • Anchoring bias: Kahneman and Tversky show judgments stick to small early numbers.
        • Long warning: Malthus, Bartlett, and The Limits to Growth cautioned against exponential resource use.
    • Failing to Predict, Struggling to Adapt (Chapter Three - The Exponential Gap · II)
      • Linear Minds in an Exponential World
        • Braudel’s medieval pace: life was “almost imperceptible” change — constant repetition of births, harvests, deaths
        • Evolved for linearity: humans adapted to a linear world, not to dramatic continuous change within a lifetime
        • Rare exponential exceptions: pandemics and hyperinflation were idiosyncratic departures from the normal rhythm
        • Exponential Age rupture: exponential technologies force minds shaped for linearity to cope with relentless acceleration
      • The Prediction Gap: Underestimation
        • Hockey-stick blindness: executives dismiss new products early because absolute numbers remain small despite exponential growth
        • McKinsey’s cell-phone miss: forecast 900,000 US subscribers by 2000; actual exceeded 100 million
        • IEA solar forecasts: predicted 5 GW for 2015, revised upward for six years, yet reality still hit 56 GW
        • Solar blind spot persists: IEA predicted flat capacity for 2019; actual output passed 105 GW
        • COVID bias: people underpredicted infection growth by 46% after one week and 66% after two
      • The Prediction Gap: Overestimation and Blind Spots
        • Kurzweil’s brain-equivalence: The Age of Spiritual Machines predicted a $1,000 computer would match the human brain by 2019; too optimistic
        • Self-driving overpromise: Musk predicted one million robo-taxis by 2020; actual was zero, and every autonomous-car firm missed targets
        • Complexity defeats extrapolation: slight errors in assumptions compound when extrapolating exponential curves into poorly understood systems
        • Unforeseen consequences: smartphones cut US chewing-gum sales 15% in a decade — a disruption no one forecast
        • Three prediction traps: underestimation, overestimation, and unforeseen effects all stem from the exponential gap
      • The Industrial Revolution Precedent
        • Engels’ pause: GDP grew for fifty years while workers’ wages flatlined — technology ran far ahead of social gains
        • Longer, harsher work: average British workweek rose from 41.5 hours in the 1760s to 57 hours by the 1870s
        • Victorian testimony: Dickens’ Hard Times and Southey’s 1814 account chronicle factory filth, noise, and misery
        • Institutional lag: Britain had a modern economy but a pre-modern political order, delaying factory regulation until 1833
        • Slow catch-up: labor movements and the welfare system took decades to match the new industrial reality
      • Institutions Built for Slower Times
        • Definition: institutions are enduring arrangements between groups that stabilize social life
        • Not just buildings: institutions include governments, firms, churches, and international bodies — formal and embodied
        • Invisible institutions: the Rule of Law and intellectual property structure society without being groups of people
        • Incremental by nature: institutions have an inbuilt tendency towards incremental change, while technology accelerates
        • Two-part gap: first failure to predict exponential cadence, then failure to adapt — today the lag grows bigger and faster
    • Institutions Lag Behind Exponential Technology (Chapter Three - The Exponential Gap · III)
      • Institutions Are Often Unwritten Rules
        • Institutions: formal laws plus informal habits and practices that govern behavior.
        • Unwritten norms: can command as high or higher adherence than written rules.
        • Monopoly: players add "Free Parking" house rules despite not being in the rulebook.
        • Road courtesy: UK drivers flash headlights to thank, though Highway Code advises against.
        • Shared practice: bakers follow common best practice for making cake, with local variations.
      • Formal Institutions Adapt Extremely Slowly
        • Catholic Church: took 346 years to overturn Galileo's condemnation, long after Sputnik.
        • Cultural lag: Ogburn found liability laws for workplace accidents lagged decades behind harm.
        • Social norms: smartphone-at-dinner etiquette and gaming values still unresolved after decades.
        • Pattern: even accurate predictions rarely trigger timely institutional response.
      • Corporate Case Studies: Kodak and Microsoft
        • Kodak: invented digital camera in 1975, predicted 15–20 years to compete, then failed to pivot.
        • Kodak's Ofoto: bought photo-sharing site but tied it to print sales, missing social sharing.
        • Kodak's end: institutional memory killed digital opportunity; bankruptcy in 2012.
        • Microsoft: missed internet until 1995 "tidal wave" memo, ceding search, e-commerce, messaging.
        • Microsoft's iPhone: Ballmer applied the PC paradigm, dismissed iPhone; mobile plans later shelved.
        • Microsoft rebound: Nadella later used cloud computing to restore the company's standing.
      • Why Institutions Stay Locked In
        • Path dependence: early choices confine later options through lock-in.
        • Family dinners: 18th-century dining-room norm still shapes parenting wisdom today.
        • Layering: new norms build on old ones, as NHS pagers persisted into 2020.
        • Drift: institutions keep old policies despite changed context, as BBC applied TV schedules to web.
        • Conversion: old practices redeployed in new contexts; all modes reinforce lock-in.
        • Exceptions: wars and catastrophes produce punctuated equilibrium, e.g., IMF/UN created rapidly after 1944.
      • The Exponential Gap and Its Stakes
        • Exponential gap: prediction difficulty plus institutional slowness leaves society static as tech takes off.
        • Trivial past: early computer-industry failures harmed companies, not society.
        • Existential present: exponential tech now mediates services, government interactions, health, education, production.
        • Two-tier society: those who harness exponentiality shape rules and norms; the rest fall behind.
        • Slowing tech: neither practical nor desirable, since climate and global development need exponential solutions.
        • Better path: accelerate institutional adaptation, resilience, and even distribution of exponential benefits.
  • Chapter Four - The Unlimited Company
    • Gravity, Networks, and Superstar Firms (Chapter Four - The Unlimited Company · I)
      • The Twentieth-Century Ceiling on Scale
        • Industrial giants: big firms grew via economies of scale but hit organizational costs as complexity rose.
        • Coase's gravity: management and information frictions slow expansion and erode scale advantages.
        • Diminishing returns: each extra dollar invested yielded less; market share rarely passed roughly 40 percent.
        • Competition as check: dispersed markets and rivals kept dominant firms honest; monopolies invited state action.
        • Standard Oil caution: 90 percent control of refined oil led to 1911 breakup into thirty-four pieces.
      • Superstar Companies Defy Gravity
        • Superstar firm: rises fast, unburdened by industrial limits; more productive, aggressive, innovative, and expansive.
        • Three forces: network effects, unleashed platforms, and capital efficiency drive and reinforce superstardom.
        • Dominance everywhere: Google search, Android/Apple phones, Facebook ads, Uber rides, Amazon retail concentrate in few names.
        • Winner-take-all economics: top 50 firms nearly half of top 500 sales; 10 percent of public companies earn 80 percent profits.
        • Digital intensity amplifies: IT productivity gap far exceeds gaps in shoes/cement; superstardom is global.
        • Exponential gap: existing norms, rules, and taxes cannot catch up with how superstar firms operate.
      • Why Networks Favor Winners
        • Network effect: each extra member raises network value for everyone; fax machines became useful as users multiplied.
        • Positive externality: Marshall and Pigou saw benefits beyond buyer-seller; network externalities drive winner-take-all markets.
        • Microsoft's flywheel: small lead attracted developers, more software brought users, larger base strengthened lead.
        • Spread across markets: Windows dominance helped Microsoft beat Lotus, WordPerfect, and rivals in productivity software.
        • Web and Wikipedia: network effects made Berners-Lee's web and Wikipedia the congregating points despite alternatives.
      • Platforms: Scaling Without Friction
        • Digital infrastructure: internet and smartphones pre-connect users, removing need to supply hardware for network effects.
        • Platform business model: firms connect producers and consumers instead of running a linear value chain.
        • Marketplace without limits: digital platforms avoid physical crowding and exhaustion; eBay, Alibaba, TikTok reach massive scale.
        • Choice abundance: GOAT can show 7,434 used sneakers; physical bazaars cannot match this range.
        • Capital efficiency: platforms scale without owning production assets, turbocharging growth and superstar dominance.
    • Platforms, Intangibles, and Superstar Growth (Chapter Four - The Unlimited Company · II)
      • Platform Economics: Own Nothing, Orchestrate Everything
        • Platform model: world’s largest cab company, hotelier, and showroom own no core assets.
        • Ecosystem absorbs costs: third parties supply stock, warehouses, logistics, and customers.
        • Platform ubiquity: Exponential Age platforms arise in Indonesia, India, Nigeria, and the Netherlands.
        • Ping An Good Doctor: AI answers 75% of 670,000 daily consultations; specialists work on demand.
      • The Shift to Intangible Value
        • Tangible-to-intangible: S&P 500 intangible share rose from 17% in 1975 to 84% in 2015.
        • Superstar book value: top five tech firms’ physical assets represent only ~6% of market value.
        • Product value migrates: worth lies in research, design, branding, and service, not manufacturing.
        • Two drivers: product complexity demands know-how; global markets amortize high first-copy costs.
        • Winner-takes-most: intangibles scale without new factories, accelerating market dominance.
      • Data Network Effects and AI
        • AI as intangible asset: acts like a perpetual motion machine, yielding more value as data feeds in.
        • Data network effect: product improves as usage generates data, drawing still more users.
        • Google: every click or “Back” refines search rankings — the ultimate network-effects machine.
        • Netflix: recommendation system turns viewing habits into better suggestions and more data.
      • Increasing Returns to Scale
        • Diminishing returns inverted: Exponential Age superstars grow faster as they get larger.
        • Stagnation broken: large mature companies once stopped growing; digital giants defy that law.
        • Salesforce: revenue per employee rose from $230,000 to $350,000 while scaling massively.
        • Netflix: revenue per employee grew from ~$1M to $2.7M while original content exploded.
      • The Superstar Mindset
        • Winner-takes-all economics: only one firm per sector dominates; second place is distant.
        • Growth obsession: managers seek the next big thing because stalling means irrelevance.
        • Blitzscaling: Reid Hoffman's doctrine prioritizes growth over efficiency, defying traditional rules.
        • Competition is for losers: Peter Thiel urges founders to dominate markets without rivals.
      • Expansion Playbooks
        • Horizontal expansion: firms move into adjacent markets to compound network effects.
        • Ant Financial: Alibaba's data-fed "Ant Brain" drives payments, wealth, credit, lending, insurance.
        • Vertical integration: companies pull supply-chain stages in-house, reviving Carnegie-era logic.
        • Google's stack: advertising targeting, auction, placement, and tracking all run within Google.
        • Custom chips: Apple and Google design their own silicon to control AI and hardware.
    • Monopoly Rethought for Exponential Age (Chapter Four - The Unlimited Company · III)
      • R&D Builds New Sectors
        • Exponential scale funds invention: biggest firms raise R&D spending faster than revenues.
        • Giants seed new markets: massive budgets let superstars invent whole sectors and later dominate them.
        • Alphabet's far-flung bets: cold fusion research, X-lab experiments, self-driving cars, and Malta salt-based energy storage.
        • Paradigm confirmed: dominant cross-sector superstars are a durable economic era, not a passing trend.
      • The Changing Face of Monopoly
        • Bork's consumer-welfare test: antitrust traditionally asks only whether behavior hurts consumers' pockets.
        • New monopolies exploit producers: modern giants squeeze suppliers and sellers more than shoppers.
        • Apple's App Store levy: 15–30 percent fees raise fairness questions where Apple is the market.
        • Google's ad dominance: self-dealing ad practices hurt advertisers and drew EU and UK scrutiny.
        • Digital consumers seem served: free search and rapid improvement mask the underlying monopoly problem.
      • Hidden Costs of Winner-Takes-All
        • Dynamism drains despite startups: record venture funding coexists with giants buying nascent rivals.
        • Small teams disrupt: large teams develop science; small teams produce breakthroughs — and get acquired early.
        • Narrow corporate research: commercial AI papers favor proven paths, shrinking the space of explored ideas.
        • Academic brain drain: AI professors leave universities for industry, reducing public-interest research.
        • Tax avoidance built into intangibles: tech giants exploit loopholes like the Double Irish, paying roughly 16 percent tax.
      • Rebuilding Antitrust for Exponential Age
        • Khan's infrastructure-power lens: conflicts of interest, not size alone, signal monopoly — as with Amazon and Ecobee.
        • Oversee acquisitions early: regulators should approve or unwind giant-firm purchases like Android and Instagram.
        • Interoperability cuts network effects: open standards let users switch and rivals compete, as SMS texting proved.
        • Open banking precedent: data portability between banks and fintechs reduces lock-in and boosts switching.
      • Treating Digital Giants as Utilities
        • Essential-service reality: modern life depends on Google, Apple, Amazon, Facebook, Microsoft, and Netflix.
        • Utility responsibilities apply: regulators can impose universal service and infrastructure-sharing duties.
        • Openreach model: splitting BT's network access enabled fair competition without duplicating infrastructure.
        • EU's Digital Services Act: very large gatekeepers face reporting, auditing, and data-sharing obligations.
        • Closing the exponential gap: industrial-age rules are finally yielding to policies built for winner-takes-all markets.
  • Chapter Five - Labor’s Loves Lost
    • The Robopocalypse Reconsidered (Chapter Five - Labor’s Loves Lost · I)
      • The Panic’s Long History
        • Automation fears: recurring since the Industrial Revolution; machines promised one-off costs instead of regular wages.
        • Machine-breaking revolts: Luddites smashed automated looms in 1810s England; GM workers torched assembly controls in 1970s.
        • Keynes’s warning: coined “technological unemployment” in 1928 as technical efficiency outpaced labor absorption.
        • Media resurgence: headlines like “Millions of UK workers at risk” framed AI as an imminent job killer.
      • AI Hype vs. Radiologist Reality
        • Narrow AI excels: face recognition, translation, and scheduling already beat humans at specific tasks.
        • Hinton’s forecast: in 2016, the AI pioneer urged people to stop training radiologists—deep learning would overtake them.
        • Actual outcome: US radiologist numbers rose to 28,025 by 2019; demand and shortage persisted.
        • Augmentation, not replacement: Rajesh Jena’s 3D tumor tool cut diagnosis time from hours to four minutes, aiding doctors.
      • The Robopocalypse Narrative
        • Alarmist books: The Rise of the Robots, The Second Machine Age, and Surviving AI predicted mass redundancy.
        • Frey–Osborne study: placed 47% of US jobs at high risk from machine learning; cited over 7,000 times.
        • Corporate automation: Deutsche Bank targeted “abacus” workers; UiPath rode the boom to a $35B valuation.
        • Manufacturing arithmetic: generating $1M in US output took 25 workers in 1980, but only 5 by 2016.
        • Giant firms, few workers: Alphabet’s $1.4M sales per employee dwarfed GM’s $74,000 at peak.
      • The Employment Paradox
        • Employment held up: OECD and global jobless rates were near record lows until COVID-19—no robopocalypse.
        • Moravec’s paradox: computers master Go, yet still struggle with a one-year-old’s perception and mobility.
        • Tacit knowledge: Polanyi’s “we know more than we can tell”; unrecorded workplace know-how gives AI only half the training picture.
        • Hidden care work: Graeber saw “unskilled” jobs like tube staff as closer to nursing than bricklaying.
        • Automation creates work: historically, new sectors and human skills emerge to replace waning jobs.
        • Real threat: job quality—income stability, skill-building, control—may fall even if job quantity stays high.
      • Gradual Automation in Practice
        • Slow automation: Frey says production became automatable only after tasks were subdivided and simplified; craftsmen resisted robots.
        • Wall Street example: algorithmic trading grew from 30% to 70% of share trades; pit trading vanished.
        • Passive funds milestone: rules-based automated funds took over half of global assets under management in 2019.
        • Limited self-driving: AVs debut on Phoenix’s wide roads, not rainy autobahns or city streets; full automation remains far off.
        • Three forces: gig platforms, automated management, and falling labor share—not robopocalypse—widen the exponential gap.
    • Automation, Competition, and the Gig Economy (Chapter Five - Labor’s Loves Lost · II)
      • Automation as Competitive Weapon
        • Automation: not a fringe or mass-unemployment force, but a tool for competitive advantage.
        • Amazon: relentless automation plus pandemic demand led to 308,000 new jobs in one year.
        • Netflix: AI may pick “Up Next,” yet the firm grew its workforce by 9.3 percent in 2020.
        • Automated growth: Ocado and JD.com expanded headcount while deploying massive warehouse robotics.
        • Blockbuster vs. Netflix: job losses came from failing to adapt, not from automating stores.
        • French study: robot-adopting firms became more profitable and hired; laggards lost employment.
      • Macroeconomic Effects and the Lump of Labor Fallacy
        • Mixed evidence: Acemoglu and Restrepo found 650,000 US manufacturing jobs displaced by robots.
        • European study: each additional robot per 1,000 workers raised overall employment by 1.3 percent.
        • Lump of labor fallacy: work is not fixed; new technologies create new sectors and skilled roles.
        • Long run: automation historically creates more jobs, but short-run pain can last years.
        • Quality not quantity: the real future concern is the quality of available work options.
      • Sid’s Story: Automating Your Own Job
        • Sid Karunaratne: managed servers full-time until exponential data growth forced him to build tools.
        • Self-automation: he wrote software to handle routine tasks, shrinking his old job to hours weekly.
        • Growth spiral: his effectiveness helped the business grow, creating more roles for the team.
        • Human + Machine: companies investing in AI generate entirely new job categories across the economy.
      • The Rise of Platform Labor
        • Uber: 3.9 million drivers with no employment contracts—roughly 200 gig workers per employee.
        • Mechanical Turk: launched 2005, hidden human labor performed tasks algorithms couldn’t handle.
        • Crowdsourcing: internet and smartphones let platform work scale through network effects and computing power.
        • Gig platforms: TaskRabbit, Talkspace, Wag! expanded from data chores to local services.
        • Gig economy: short-term freelance work allocated online, not automation, raises the hardest questions.
        • Scale: by 2017, Uber carried more NYC passengers than yellow cabs; UK had 2.8 million platform workers.
    • Platform Labor and Digital Taylorism (Chapter Five - Labor’s Loves Lost · III)
      • Gig Work’s Promise: Bigger Markets
        • Gig platforms: add ~72 million full-time positions to global labor by 2025; digital piecework grows the workforce ~2%
        • Market efficiency: databases and algorithms match demand and supply, cutting middlemen and expanding opportunity
        • Global reach: UpWork lets freelancers serve distant clients; billings grew twelvefold between 2011 and 2020
        • Emerging economies: platforms formalize casual labor and widen tax bases in India; Kobo360 helps Nigerian truckers
      • Gig Work’s Reality: Low Pay and Imbalance
        • Bargaining imbalance: platforms dominate armies of isolated workers with no collective voice or unions
        • Low pay: German and French platform workers earn 29–54% below minimum wage; nearly 40% of UK gig workers earn under £8.44/hour
        • Expenses squeeze: after costs, DoorDash workers earn $1.45/hour; Uber drivers net about $3.37/hour (MIT study)
        • Precariousness: sick pay and stable wages absent; platforms change pay and operating rules unilaterally
      • The Fight Over Worker Classification
        • Platform defense: companies claim they merely connect buyers and sellers, like modern temp agencies
        • California compromise: after AB5, Uber, Lyft, and DoorDash won Prop 22, keeping drivers as independent contractors
        • UK Supreme Court: Uber’s control over trips, pricing, and communication proves “subordination”; drivers count as workers
        • Exponential gap: twentieth-century labor laws fail to cover smartphone-enabled freelance labor, so courts decide instead
      • From Time Clocks to Tech Perks
        • Bundy’s time recorder: automatic clocking-in enabled scientific, regulated people management from the 1880s
        • Taylorism: time-and-motion studies quantified output but treated workers like machines; the unobserved worker is inefficient
        • Twentieth-century deal: stable wages and benefits in exchange for autonomy; Gilbreth added psychological testing
        • Tech-era perks: bean bags, yoga, and egg freezing replaced cubicles; Netflix offers unlimited vacation, Google generous death benefits
      • The Digital Panopticon
        • Workplace surveillance: logged emails, Slack messages, and documents make every employee visible — Bentham’s panopticon realized
        • Mood tracking: facial recognition, Hitachi sensor badges, Microsoft productivity scores, and brainwave hats assess engagement
        • Algorithmic management: gig workers face quantified pacing; too many ride refusals trigger deactivation from platforms
        • Ultra-Taylorism: Amazon’s metrics auto-fire ~10% of warehouse staff yearly; Deliveroo punishes late arrivals regardless of weather
        • Janus face: the same firms grant elite perks while algorithmically controlling others; workers lack collective power
    • Labor's Shrinking Share and Renewal (Chapter Five - Labor’s Loves Lost · IV)
      • The Declining Labor Share
        • Labor vs capital: workers' share of national income fell 6.5% across 34 advanced economies from 1980 to 2014
        • US divergence: workers took 65% of national income in 1947, only 56.7% by 2018; most decline came since 2000
        • Pay vs productivity: US productivity rose 255% since 1948 while pay rose only 125%; wages stalled after 1973
        • Four causes: globalization, union decline, intangibles, and superstar firms; intangibles and superstar firms explain over two-thirds of loss
      • The Bifurcated Workforce
        • Growing chasm: high-quality work for some, insecure gig work for others, with algorithmic management treating employees as fungible
        • Uber model: engineers averaged $147,603; drivers made about $30,390 before expenses
        • Facebook contrast: half of employees earn $240,000+; content moderators are contracted at about $28,000; users are unpaid
        • Hollowing out: Zymergen shows demand for PhD scientists and low-paid support staff, with little middle-wage work
      • Historical Precedent and Warning
        • Long-run optimism: plough, electricity, and indoor lighting eventually made labor safer and raised living standards
        • Century of lag: wages rose 12% from 1790 to 1840 while GDP per worker rose 50%; catch-up took until 1900
        • Winners and losers: British handloom weavers were immiserated even as industrial wages rose overall
        • Keynes warning: "In the long run we are all dead"—decades of lag mean real human cost
      • The Fair Work Settlement
        • Dignity: consent and autonomy matter; BP's voluntary Fitbit wellness scheme shows monitoring need not exploit
        • Flexibility: volatile industries demand constant reskilling; digital micro-learning on TikTok hints at future education
        • Security: UBI trials in Stockton and Helsinki improved wellbeing and re-employment; Denmark's flexicurity guarantees benefits despite easy firing
        • Equity: wage transparency, minimum-wage bargaining tied to living costs, and employer investment can close gaps
        • Entrepreneurial fixes: Portify builds gig workers' credit histories through small monthly repayment loans
      • Organizing for Renewal
        • Union decline: fifty-year political assault; UK membership fell 20% from 1979–1988; Reagan fired 11,000 controllers
        • Tech anti-unionism: Intel's Noyce called non-union survival; Alphabet union started with 200 of 135,000 employees
        • Digital organizing: WhatsApp and Facebook groups fueled West Virginia teachers' strikes; app drivers formed a global alliance
        • Collective bargaining: unions remain the best route to fair work and to adapting employment laws to Exponential Age
  • Chapter Six - The World is Spiky
    • The Spiky Return to Local (Chapter Six - The World is Spiky · I)
      • Flat-World Globalization Unravels
        • Pix Moving: autonomous-vehicle chassis designed in Guiyang, printed in California, defying US-China tariffs.
        • Friedman's thesis: The World is Flat predicted endless globalization through ten flatteners in a third phase.
        • Globalization's rise: trade grew from a quarter of global GDP in 1970 to nearly 60 percent by 2019.
        • Backlash: financial crisis, offshoring, Brexit, and Trump nationalism turned the West against openness.
        • Exponential irony: globalization's own technologies create reasons and tools for borders, making geography matter again.
        • Spiky map: dense, face-to-face tech clusters empower cities while nation-state structures creak.
      • From Comparative Advantage to Local Production
        • Classical trade theory: Smith's specialization and Ricardo's comparative advantage justified free exchange of goods.
        • Hyper-global supply chains: phones source lithium, aluminium, palladium; aircraft wings cross borders six times.
        • Food globalization: Spanish tomatoes reach British homes through the Golden Triangle distribution hub in three days.
        • Dematerialized production: export blueprints, not goods; additive manufacturing prints components wherever customers are.
        • Local inversion: instead of shipping food or fuel, you can transform what is nearby using technology.
      • Vertical Farming Makes Food Hyperlocal
        • Vertical farms: stack fields twelve stories high; AI controls light, water, heat to boost productivity.
        • Resource efficiency: hydroponics or aeroponics use up to 80 times less water and no pesticides.
        • Photon precision: tune LED wavelengths to crops so no light energy is wasted.
        • Urban locations: Lufa Farms' rooftop greenhouse in Montreal supplies fresh produce from atop a warehouse.
        • Exponential growth: high-intensity vertical-farm market expands over 20 percent annually.
      • Energy Localism and the Dematerializing Grid
        • Renewables' effect: wind and solar reduce shipping of coal, oil, and gas, enabling energy independence.
        • Coal's collapse: UK coal use for electricity fell 94 percent by 2019, imports down 70 percent.
        • Decoupling growth: UK GDP rose 75 percent while electricity use fell 15 percent—wealth per kilowatt-hour doubled.
        • Solar equity: sun exposure varies only ~4x across nations, versus oil reserves a million-to-one disparity.
        • Vehicle-to-grid: EVs' stored electricity can power homes; 11m UK cars could cover national demand.
    • Localization, Divergence, and Urban Power (Chapter Six - The World is Spiky · II)
      • Energy and Materials Localize
        • Virtual power plants: Moixa networks idle car batteries, replacing a physical power station with parked EVs.
        • Microbial production: Zymergen engineers microbes to grow materials like Hyaline, reducing oil use in manufacturing.
        • New dependencies: green economy still leans on lithium and rare earth metals, not just hydrocarbons.
        • Slow transition: a fully green economy may not arrive until the late twenty-first century.
        • Less stuff: exponential economy moves toward localized commodities and lower material intensity.
      • Manufacturing Shifts from Goods to Ideas
        • Shipping ideas: designs cross borders while fabrication happens close to consumers.
        • Intangibles dominate: a $1,000 iPhone holds less than $400 in parts; rest is design and brand.
        • Automation reshoring: robots make high-wage countries viable for complex production again.
        • Adidas Speedfactory: German automated plant made personalized shoes, but closed before machines matured.
        • Material savings: printed parts cut weight and waste, as in aircraft components.
      • The 3D-Printing Wave
        • Wright's Law: 3D printing costs fall around 30 percent annually as volume grows.
        • Pandemic supply: schools' 3D printers produced masks when global supply chains failed.
        • Small but growing: 3D printing was a $10 billion market in 2019, now reaching industrial users.
        • BMW i8: a 3D-printed bracket lets the roof fold—tens of thousands of such parts planned.
      • Rupture for the Developing World
        • Commodity dependence: African economies rise and fall with raw-material prices, from boom to bust.
        • Oil-price shock: the 2020 collapse showed petrostates like Saudi Arabia need high prices to stay stable.
        • Trade losses: ING estimates 3D printing could eliminate 40 percent of world imports by 2040.
        • Manufacturing incomes: rich-world local production may strip poor countries of export jobs.
        • Growing chasm: poor states lack capital and high-skill labor to share in exponential gains.
      • Global Order Under Strain
        • Supply-chain revolution: UPS is investing in on-demand printing, not cross-border package delivery.
        • Institutional erosion: localized production weakens the trading systems and multilateral bodies that sustain order.
        • US disengagement: fracking-based energy security removes a central reason for Middle East intervention.
        • Uneven future: high-tech rich economies thrive while others get left further behind.
      • Cities Are the Real Engines
        • Urban milestone: more than half of humanity has lived in cities since 2007.
        • Underestimated role: cities drove wealth, science, and trade, yet national metrics obscure them.
        • Remote-work myth: exponential tools seem to disperse work, but intangible economy intensifies urban clustering.
        • Agglomeration: knowledge workers cluster around universities, labs, and skilled labor pools.
        • Specialist hubs: Delhi for Bollywood, Tel Aviv for cyber security, Hsinchu for chips.
        • Serendipity: dense mixing of strangers sparks innovation that remote work cannot replicate.
    • Cities and Splinternet Tensions (Chapter Six - The World is Spiky · III)
      • Cities as Exponential Magnets
        • Strangers as fuel: city life is millions of variables, generating collisions of ideas
        • Perpetual motion: bigger cities attract top talent; top talent makes them bigger
        • Geoffrey West: positive feedback loop—size boosts innovation, income, and infrastructure efficiency
        • Megacity future: 41 cities will house 9% of humanity by 2030, led by Asia and Africa
        • Pandemic resilience: COVID flight was temporary; cities reasserted their pull
      • The Urban-Rural Divide
        • City premium: urban citizens are more educated and earn double or more rural incomes
        • Tension: liberal, rich, high-tech cities clash with poorer national governments
        • Regulatory fights: London vs Uber showed national leaders taking firms' side
        • Immigration: cities welcome labor while nations restrict—"nations talk, cities act"
        • Power shift: Exponential Age favors cities, eroding the nation-state's dominance
      • From Global Internet to Splinternet
        • Barlow's declaration: cyberspace announced independence from governments in 1996
        • Globalist early web: borderless ideas undermined sovereignty and enabled protest
        • American roots: early internet protocols and governance had a distinctly US liberal tint
        • Government reassertion: firewalls, surveillance, and app bans splinter the web
        • Data localization: territorial data rules spread to 84 countries by 2016
      • Institutional Gaps and Conflict Risks
        • Outdated institutions: IMF/WTO globalist assumptions fit neither relocalization nor city power
        • Global threats remain: climate, pandemics, cybersecurity still need global cooperation
        • World Data Organization: proposed to keep data flowing despite digital walls
        • Digital minilateralism: small agile groups like Digital Nations set shared AI principles
        • Urban federalism: devolving power to cities may reduce national-urban divides
        • War risk: economic independence and urban-rural tensions may breed conflict; warfare is transforming
  • Chapter Seven - The New World Disorder
    • Cyberwar, Cheap Attacks, Fractured Order (Chapter Seven - The New World Disorder · I)
      • Estonia 2007: The First Digital State Under Siege
        • Estonia 2007: coordinated botnet and denial-of-service attacks disabled banking, media, and government, forcing a national internet shutdown.
        • Mob to malware: riots over a Soviet statue and Russian minority heritage preceded digital salvoes on national infrastructure.
        • Misinformation layer: pro-Russian hackers flooded news sites with fake stories, amplifying hysteria.
        • Article 5 ambiguity: NATO's mutual-defense clause could not confidently cover cyberattacks or prove Kremlin culpability.
        • Precedent: the attack showed governments that cheap digital warfare could paralyze a connected nation.
      • The Expanding Attack Surface
        • Attack surface: total exploitable vulnerabilities grows as every connected device opens a new security loophole.
        • From castle to cloud: medieval defenses with one drawbridge faced limited weapons; modern states face proliferating digital vectors.
        • Civilian exposure: smart bulbs, cars, and corporate Wi-Fi let hackers reach targets far beyond military systems.
        • Scale of leaks: 37 billion data records leaked in 2020, a 46-fold jump in five years; British firms hit 96% in 2019.
        • Exponential gap: offensive technologies race ahead while defense structures and security resources lag.
      • War Gets Cheaper and More Remote
        • Cost collapse: malicious code follows Moore's-law-like price dynamics; drones are a thousand times cheaper than a decade ago.
        • No front line: digital and drone attacks spare attackers from immediate fire, lowering the barrier to conflict.
        • Osirak to Stuxnet: Israel's 1981 F-16 raid needed millions in planes; Stuxnet destroyed Iranian centrifuges with software alone.
        • Civilian blowback: NotPetya malware spread from Ukraine to 65 countries, causing up to $10 billion in damage.
        • New state actors: North Korea's 5,000–7,000-strong cyber army projects force beyond its borders without physical operations.
      • Deglobalization and New Wars
        • Trade-peace link: economic interdependence reduces military conflict, so deglobalization raises war risk.
        • Kaldor's new wars: post-Soviet conflicts involve non-state actors, identity politics, civilians as targets, and internal fractures.
        • Exponential contribution: urban-rural divides, weak intergovernmental bodies, and cheap military tech feed non-state conflicts, which doubled by 2018.
        • Grinding future: Sir Richard Barrons says technology is "blowing away" established ideas of how war is fought.
        • Norms gap: Geneva-era rules don't cover hacker armies, drone swarms, or cyber-to-missile retaliation.
    • Cyber, Disinformation, and Cheap Drones (Chapter Seven - The New World Disorder · II)
      • Cyberattacks as Asymmetric Power
        • Cybercrime revenue: North Korean cyber experts allegedly raised up to $2 billion for weapons programs.
        • Cheap havoc: A computer and a talented hacker let relatively poor states project disproportionate force.
        • Shamoon strikes: Iran-linked malware erased 30,000 Saudi Aramco computers and disrupted output for nearly two weeks.
        • Deniable edges: Cyber ops give Iran cost-effective, deniable leverage against wealthier rivals.
        • Sandworm chaos: GRU-linked hackers used NotPetya to wreck Maersk globally while seeking disorder in Ukraine.
      • Outdated Defenses and Legal Order
        • Westphalia legacy: sovereignty-based treaties cannot regulate digital subterfuge by unofficial actors.
        • Geneva Conventions: old rules on targets don’t map onto cyberattacks harming private firms and civilians.
        • Late cyber defenses: US Cyber Command adopted “defending forward” only in 2018; UK’s National Cyber Force began in 2020.
        • SolarWinds breach: foreign hackers spied on US Treasury and Homeland Security for months via a trusted software vendor.
        • Unclear acts of war: NotPetya’s route through American software leaves no easy legal answer.
      • Misinformation and Disinformation Campaigns
        • Fake news factories: Veles teens earned up to $2,000 a day from viral pro-Trump stories in 2016.
        • Viral falsehoods: MIT study found false claims rose thirtyfold and spread six times faster than truth.
        • Platform accelerants: cheap bots and algorithmic rewards for outrage make misinformation easy to amplify.
        • State disinformation: actively malicious information is used as attack; 70 countries ran online ops by 2019, Russia led.
        • Real-world harms: COVID infodemic killed Iranians via methanol and threatened UK engineers over 5G; Russia weaponized US divisions.
        • Slow governments: UK built a disinformation unit only in 2018; US had no coordinated strategy by 2020.
      • Exponential Rise of Cheap Drones
        • Integrated warfare: disinformation and cyberattacks soften an adversary; drones deliver the physical blow.
        • US drone expansion: Obama launched 542 strikes, killing 3,797 people and normalizing remote killing.
        • Exponential price drops: smartphone accelerometers, chips, batteries, and AI made drones cheap.
        • Cost asymmetry: Turkish Bayraktar TB2 costs about $5 million; a US Global Hawk costs $130 million.
        • New drone actors: PKK packed Amazon drones with explosives; Houthis hit Saudi oil, halving output for a day.
    • Autonomy, Asymmetry, and Fragile Order (Chapter Seven - The New World Disorder · III)
      • The New Asymmetric Battlefield
        • Asymmetric war: cheap exponential drones let weaker actors strike powerful states with devastating effect.
        • Nagorno-Karabakh, 2020: Azerbaijan's Turkish- and Israeli-made drones destroyed over half of Armenia's heavy materiel.
        • Drone agility: small platforms launch from low-key sites, reach hard areas, and sharpen targeting precision.
        • Swarm decoys: light-vehicle and helicopter-launched swarms can blind radars and lure enemy forces.
        • Illegal warfare risk: precise drones may make chemical or biological attacks temptingly viable for rogue states.
      • The Autonomy Continuum
        • Autonomy continuum: remote-controlled drones are one stage toward weapons that decide targets themselves.
        • Human in the loop: Tomahawk navigates autonomously but needs a person to specify target and timing.
        • Autonomous munitions in use: Israel's Harpy and Harop hunt with electromagnetic, visual, and infrared sensors.
        • Cheap civilian autonomy: Skydio R1 tracks its owner with thirteen onboard cameras and sells for under $2,500.
        • Automated targeting: SIPRI found more than one-third of 154 systems could select targets and attack without humans.
        • Swarm intelligence: onboard computers let synchronized drones act faster and remove communication-delay vulnerabilities.
      • The Responsibility Gap
        • Dystopian prospect: fully autonomous systems could decide whether, who, when, and how to attack without a human trigger.
        • Legal vacuum: war norms, the International Criminal Court, and Nuremberg all assume individual human responsibility.
        • Responsibility gap: scholars see no clarity on who is liable—engineers, programmers, generals, or politicians.
        • No new Geneva Convention: absent a legal framework for autonomous warfare, accountability remains unresolved.
      • The Exponential World Is Disordered
        • More actors, more surfaces: drones, cyberattacks, and disinformation give more actors new ways to strike.
        • Cheap war: exponential weapons make conflict cheap in money and lives, from nuisance misinformation to full-blown attacks.
        • Front-line citizens: civilians and businesses, not just armies, face the crossfire of fragmented conflict.
        • Unstable order: re-localization and low-level conflict make escalation unpredictable and hard to stop.
      • Rebuilding National Defences
        • Three state tasks: strengthen defences, set communication norms, and curb proliferation of attack surfaces and weapons.
        • Weak Western castles: expensive fighters like the F-35 cannot stop cyberattacks; investment must shift to cyber defence.
        • Corporate attack surface: private firms own digital infrastructure, so states need public accountability and security duties.
        • Digital Geneva Convention: Microsoft's initiative treats companies as guardians shielding civilians from cyberattacks.
        • Citizen digital hygiene: passwords, multi-factor authentication, and media literacy are frontline defences.
        • Finland and Taiwan models: public-private programs teach critical thinking and fake-news recognition in schools.
      • De-escalation and Non-Proliferation
        • Cold War hotlines: direct leader-to-leader communication reduces the risk of small miscommunications escalating.
        • Blurred red lines: NATO only began weighing cyberattacks under Article 5 in 2019—rules lag technology.
        • Treaty precedents: bans on chemical arms, landmines, and cluster bombs made unacceptable tactics rare.
        • Intrusion as a service: CyberPeace Institute urges bans on commercial hacking tools like NSO's Pegasus.
        • Autonomous weapons law: a wider consensus favours mandatory human control over weapons design and use.
  • Chapter Eight - Exponential Citizens
    • The Privatized Public Sphere (Chapter Eight - Exponential Citizens · I)
      • The Facebook Controversy That Opened an Era
        • Censorship by fiat: Facebook removed the iconic napalm-girl photo as child pornography, then suspended the Norwegian journalist who posted it.
        • Even Norway's PM silenced: Erna Solberg’s reposted image was also removed until global outcry forced Facebook to reverse course.
        • Unaccountable power: Mark Zuckerberg controls roughly 60% of voting shares; Facebook’s board advises rather than oversees.
        • Exponential scale: increasing returns let tech giants expand horizontally into sectors once considered beyond the market.
      • Three Frontiers of Market Encroachment
        • Public sphere privatized: private platforms, not elected parliaments, increasingly make the rules governing public conversation.
        • Private selves commodified: health, relationships, and intimate data are monitored and sold, creating digital doppelgängers of every user.
        • Social bonds remade: Facebook, Twitter, and TikTok increasingly shape who we meet, how communities form, and how polarized we become.
      • Code Is Law: The New Rule-Makers
        • Coders govern cyberspace: Lessig’s Code is Law warned that software choices constrain behavior as powerfully as formal legislation.
        • Dorm-room values go global: Facebook’s “It’s complicated” relationship status became a universal standard, exporting one campus’s worldview.
        • Apple and Google set pandemic policy: their privacy-first contact-tracing design determined what governments and scientists could do, without democratic mandate.
      • Censorship and Its Failures
        • Arbitrary moderation: Facebook removed the napalm photo yet allowed messages that fueled murderous mobs in India and Myanmar.
        • Missed warnings: a Facebook page urged arming before Kenosha; 455 user reports left it visible before Rittenhouse killed two protesters.
        • Trump’s ban exposed inconsistency: platforms drew a random line after years of racist conspiracy theories and false voter-fraud claims.
    • Private Power, Data Selves, and Engineered Divides (Chapter Eight - Exponential Citizens · II)
      • Private Companies Govern Public Speech
        • Platform moderation: Trump's post-Capitol ban followed a decade of tolerated misinformation and unchecked Rohingya hate speech.
        • Public conversation privatized: our shared speech runs on terms of service and commercial incentives of exponential platforms.
        • The real issue: not whether platform calls are right, but whether private companies should be making them at all.
        • Facebook's Oversight Board: expert review panel is a rushed "shim," not a substitute for democratic governance.
        • Nick Clegg's defense: without agreed democratic frameworks, companies must make real-time content decisions.
      • The Data Self Is Privatized
        • Digital doppelgängers: numeric data from internet trails outdoes self-knowledge; health sensors and brain interfaces will add more.
        • Privacy is power: Carissa Véliz says knowing personal details is the quintessential digital-age power.
        • GDPR's double meaning: "data subjects" are both protected people and subjects of data.
        • Market encroachment: private lives have been taken into the private sector, sold back as targeted profiles.
        • Tracking explosion: from Guardian-era log files to hundreds of organizations tracking each visitor's every move.
      • Data Economies Promise and Threaten
        • Data gold rush: Google's AdWords proved behavioral data's value, triggering mass extraction by thousands of firms.
        • Data benefits: rich datasets improve medical, economic, and credit models; credit bureaus broaden access to loans.
        • Facebook's data deals: 500 advertisers held a reporter's contact info; device makers got friends' data without consent.
        • 23andMe: sold consenting customers' genomic data to GlaxoSmithKline for $300 million, outside HIPAA and re-shareable.
        • Credit bureau flaws: inaccurate data and breaches turn consumer-credit tools against users.
      • Profiling and Prejudice
        • Social credit: China's data citizen scores target a trust crisis but carry dystopian implications.
        • Demographic profiling: banks and insurers sort us by behavior and demography, making prejudice inevitable.
        • Algorithmic racism: hospital algorithm referred fewer Black patients with the same illness, reflecting a racist society.
        • What gets measured gets managed: companies profile, target, and manipulate us from our own private information.
      • Homophily Turns Extreme Online
        • Homophily: we cluster with similar people; unchecked, it breeds pernicious polarization and democratic breakdown; exponential technologies will worsen it.
        • Engineered cliques: networks connect like-minded users for engagement and easier ad targeting.
        • Algorithmic push: recommendation engines promote progressively more extreme content to hold attention.
        • YouTube radicalization: Tufekci calls YouTube "one of the most powerful radicalizing instruments"; studies confirm migration to extremes.
        • Extremist networks: ISIS recruiting via Facebook recommendations; leaked memo: "Maybe someone dies... And still we connect people."
        • Facebook's own research: 64% of extremist-group joiners came via recommendations; algorithms exploit attraction to divisiveness.
    • Rebalancing Power in the Exponential Age (Chapter Eight - Exponential Citizens · III)
      • The Second Enclosure
        • Enclosure returns: life’s common spaces—conversation, data, intimacy—are now fenced off as private property
        • Genomic selection: IVF embryo screening could normalize choosing children, deepening homophily and eroding shared values
        • Homophily as fragmentation: sameness-seeking atomizes us into smaller, more extreme, walled-off groups
        • Private rule-making: platforms dictate social norms; if democracy doesn't set limits, code becomes law
      • Transparency and Oversight
        • Transparency principle: algorithms’ decisions about censorship and amplification must be open to public scrutiny
        • Inconsistent moderation: Norway’s PM silenced, an American president tolerated—the rules are made in private
        • External inspection: algorithms need independent oversight, like FAA certification of aircraft or bank stress tests
      • Interoperability as Power-Shifting
        • Interoperability principle: users should carry data and reach friends across rival platforms, restoring choice
        • A closed history: open APIs once allowed tools like FriendFeed; Zuckerberg bought it, then walled the networks
        • Mandate at scale: governments could require interoperability for services above 10–15 percent market share
        • Beyond social media: gig workers could carry reputations across apps; patients could own portable medical records
        • Kryptonite effect: interoperability dismantles winner-take-all network effects and corporate control
      • Data Rights, Not Data Sales
        • Reject data-as-property: selling our digital doppelgängers only deepens the marketization of private life
        • The payoff is trivial: Facebook’s profit per user in 2019 was about seven dollars—half buys a coffee a year
        • Three inviolable rights: no unreasonable surveillance, no surreptitious manipulation, no data-based discrimination
        • Bill of rights limits: declarations protect against abuse but don't enable the shared benefits of aggregate data
      • Commonality and Digital Commons
        • Ostrom, not Hardin: shared resources can be self-governed through community norms, not privatization or nationalization
        • Data commons work: UK Biobank pools 500,000 people’s medical and genomic data for legitimate researchers
        • Open-source infrastructure: volunteers build Linux, Firefox, and Wikipedia—managed collectively, owned by no one
        • Comedy of the commons: digital resources grow with use, making commons more viable than commercial hoarding
        • The four-part answer: transparency, interoperability, rights, and commons return power to exponential citizens
  • Conclusion - Abundance and Equity
    • Exponential Abundance and Its Costs (Conclusion - Abundance and Equity · I)
      • COVID-19: The Exponential Age Made Visible
        • Vaccine speed: from viral genome to approved vaccines in under a year, versus decades for earlier diseases.
        • Genome sequencers: twenty thousand machines worldwide turned the virus code into open-access data for instant collaboration.
        • Machine learning: Moderna tapped data feedback loops to compress vaccine design from years to days.
        • Digital rollout: online trial recruitment, databases, and text reminders made mass vaccination possible.
        • Lockdown tech: Zoom, streaming, and delivery apps — all post-2010 — turned isolation into networked life.
      • Pandemic's Exponential Shadows
        • Viral spread: a hyperconnected economy propelled SARS-CoV-2; reported deaths topped 2.7 million within a year.
        • Misinformation R0: fake news on social networks spread with infection rates from 1.46 to 2.24 — itself exponential.
        • Economic shock: lockdowns triggered market turmoil on par with the 1929 crash and disrupted global supply chains.
      • The Next Exponential Decade
        • Computing power: a dollar will buy at least 100x more computation in a decade, a discontinuity like electricity after 1920.
        • Energy transition: solar and wind five times cheaper, EV batteries a third the price, fossil fuel plants winding down.
        • Biological medicine: genome sequences near $1, enabling routine sequencing, personalized drugs, and early detection.
        • Planetary sensing: nanosatellites and tree-by-tree tracking could monitor forests, reefs, and ocean ecosystems.
        • New food systems: vertical farms can cut water and resource use while supplying healthier urban food.
      • Abundance Comes with Costs
        • More from less: exponential tech delivers computation, energy, biology, and materials at ever-lower cost.
        • Virtuous cycle: demand drives production, efficiency, and further innovation via Wright's Law.
        • Demand side: supply catalyzes demand — cheaper things invite more use, not less.
        • Material limits: cheap solar panels, chips, batteries, and buildings still require sand, metals, and land.
        • Rebound effect: a 100x cheaper genome test that is used a million times more raises total resource use.
    • Agency, Principles, and New Institutions (Conclusion - Abundance and Equity · II)
      • Risks of Ungoverned Exponential Growth
        • Overconsumption: falling costs can trigger gross increases in consumption, risking environmental catastrophe.
        • Destabilization: exponential speed disrupts institutions, with the heaviest burdens falling on vulnerable firms and less-trained workers.
        • Power concentration: superstar companies and early adopters gain unaccountable influence over society.
        • Exponential gap: systems change slower than technology, opening dystopian futures—unless we act.
      • Technology Is Not Destiny
        • Human agency: technology’s shape, direction, and impact are not preordained; we decide what we want from our tools.
        • Context matters: DDT and gig platforms show the same technology yields different outcomes in different societies.
        • Technology transforms: Kranzberg’s insight: technology is neither good, bad, nor neutral—it always brings change.
        • Managed disruption: embrace experimentation and direct change where possible; manage surprises when not.
      • Responses Across Every Domain
        • Monopoly: update competition rules and societal expectations for firms that defy old definitions of growth and market power.
        • Workers: guarantee dignity, flexibility, and security through new forms of collective action for employees and gig workers.
        • Place and power: decentralize governance to agile local units while building international bodies to prevent exclusion.
        • Security: strengthen defences and resilience; create new norms to de-escalate conflict and curb weapons proliferation.
        • Market limits: protect public life from commodification through transparency, digital rights, and common ownership.
      • Principles for the Exponential Age
        • Commonality: build cooperative institutions across states, businesses, workers, and communities.
        • Resilience: design sturdy systems—like flexicurity and digital rights—that withstand constant change.
        • Flexibility: keep institutions agile enough to adapt, avoiding the rigidity that leads to being outpaced.
      • A New Social Settlement
        • Historical precedent: universal suffrage, permanent contracts, and global supply chains once seemed impossible—yet were built.
        • Constant disorder: the Exponential Age demands institutions fit for permanent destabilization, not gradual evolution.
        • Human ingenuity: modern history is powered by technological change and people’s ability to shape its direction.
        • Harnessing steam: water turns to steam, but with new tools and systems we can capture its power instead of being scalded.
  • Acknowledgments
    • The Collective Behind a Solo Craft
      • Book as collective work: the writer travels alone, but never without a network of supporters
      • Research team: Marija Gavrilov led, with Sanjana Varghese, Emily Judson, Joseph Dana assisting
      • Breadth of inquiry: from turn-of-the-century labor relations to photolithography limits and quantum supremacy
    • Editorial and Publishing Partnership
      • Rowan Borchers: editor and instigator who taught a first-time author how to tell his story
      • Ruthless cuts, delivered gently: wrestling the manuscript into shape
      • Gemma Wain: meticulous and sensitive copy editing
      • Jeff Shreve: agent and sounding board as the project took shape
      • US partnership: Scott Waxman, Keith Wallman and Diversion Books brought dynamism
    • Early Readers and Critics
      • Full-draft readers: Mark Bunting, Kevin Werbach and Tom Glocer braved preliminary iterations
      • Chapter shapers: dozens of readers whose comments reshaped key arguments
      • Reader mix: technologists, policymakers, economists and strategists stress-tested the thesis
      • Debt to critics: sharp challenge is what turns a draft into an argument
    • The Exponential View Ecosystem
      • Podcast and projects teams: kept guests—and insights—flowing
      • Newsletter readers: gave the ideas their early momentum
      • Exponential Do community: vigorous ongoing discussion illuminated many issues
      • Origins: Zavain Dar and Tuan Pham drew out the first presentation in December 2016
      • Shamil Chandaria: first argued the thesis should become a book
      • Unreturned calls: founders and clients who tolerated the crunch periods
    • Intellectual Foundations
      • Carlota Perez: technological revolutions and their relationship to economic paradigms
      • W. Brian Arthur: increasing returns to scale and technology's ecological nature
      • Vaclav Smil: energy's role in human and societal development
      • Unknowing contributors: countless conversations sharpened the thinking
    • Teachers, Family, Personal Debt
      • Writing mentors: journalists who taught clearer thinking and prose
      • Early teachers: foundations in computing, politics and economics
      • Parents: introduced economics and its development impact; his mother brought computers home in the early 1980s
      • Extended family: sister, in-laws and children as steady encouragement
      • Shehnaz Suterwalla: challenged and sharpened a nebulous argument while moving family and projects forward
  • Select Bibliography
    • Reference Section (Select Bibliography · II)
      • Source List
        • Bibliography: reference apparatus — no chapter content to distill
• Core Conclusion and Practical Takeaways
  • The Core Diagnosis
    • Exponential gap: technologies compound at over 10 percent yearly while institutions adapt only incrementally.
    • Point B: the 2010s mark the tipping point where exponential tech rewires business, work, politics, and selfhood.
    • Technology is not neutral: tools encode the preferences and power structures of those who build them.
    • Prediction fails both ways: we underestimate compounding and overestimate extrapolation into complex systems.
    • Human agency governs outcomes: the same technology produces different results in different societies.
  • Mindset Shifts: Thinking Exponentially
    • Wright's Law: costs fall a constant percentage with every doubling of cumulative production.
    • Expect handoffs: as one technology's S-curve flattens, adjacent innovations take over the curve.
    • Hockey-stick blindness: small absolute numbers mask explosive growth, so forecasters systematically miss adoption.
    • Beware overestimation: exponentials extrapolated into complex systems compound errors; self-driving cars proved it.
    • Four domains, not one: computing, energy, biology, and manufacturing now decline in cost exponentially together.
  • Rebuilding the Economy and Work
    • Superstars defy gravity: network effects, platforms, and intangibles produce winner-takes-all markets with increasing returns.
    • Modern antitrust: police conflicts of interest and infrastructure power, not just consumer prices; scrutinize acquisitions early.
    • Interoperability as kryptonite: data portability dismantles network effects and lets users and rivals switch.
    • Treat digital giants as utilities: impose universal service and infrastructure-sharing duties on essential platforms.
    • Fair work settlement: guarantee dignity, flexibility, security, and equity as work splits into elite and gig tiers.
    • Organize collectively: unions and digital organizing remain the most effective route to fair work.
  • Reclaiming Citizen Power
    • Private rule-making: platforms, not parliaments, now set the terms of public conversation.
    • Transparency and inspection: algorithms governing censorship and amplification need independent external oversight.
    • Data rights, not data sales: prohibit unreasonable surveillance, surreptitious manipulation, and data-based discrimination.
    • Digital commons: Ostrom-style shared governance, like UK Biobank, pools data without privatizing it.
    • Resist the second enclosure: defend conversation, data, and intimacy from becoming private property.
  • A New Settlement: Security, Place, and Principles
    • The world is spiky: relocalization and city power erode flat globalization and nation-state dominance.
    • Cheap asymmetric warfare: drones, cyberattacks, and disinformation let more actors strike at low cost.
    • Rebuild defenses: shift spending from legacy platforms toward cyber resilience, norms, and corporate accountability.
    • Commonality: build cooperative institutions spanning states, businesses, workers, and communities.
    • Resilience and flexibility: design systems like flexicurity and digital rights that absorb constant change.
    • Abundance has costs: rebound effects and material limits mean exponential growth still needs governing.