第1章
The World Through Fossil-Colored Glasses
When Bill Gates says he waits for a new Vaclav Smil book "the way some people wait for the next Star Wars movie," you know you're dealing with something extraordinary. "How the World Really Works" has become required reading among tech leaders, policymakers, and environmentalists alike for its unflinching examination of our civilization's material foundations. What makes this book particularly powerful is Smil's refusal to indulge in either apocalyptic doom-mongering or techno-utopian fantasies. Instead, he offers something far more valuable: a clear-eyed assessment of our dependence on fossil fuels and the physical realities that will shape our future.
第2章
The Fossil-Fueled Foundation of Modern Life
The story of human civilization can be told through revolutionary shifts in energy conversion. For billions of years, Earth witnessed only natural phenomena-volcanic eruptions, earthquakes, storms-before the first significant shift occurred 3.5 billion years ago with photosynthetic microbes. The pace accelerated with cyanobacteria creating our oxygenated atmosphere, followed by increasingly complex life forms. But the most consequential shift for humanity began several hundred thousand years ago when our ancestors mastered fire-the first use of energy external to their bodies.
This extrasomatic energy use expanded dramatically with agriculture 10,000 years ago, followed by animal domestication. Yet progress remained slow for millennia. Even by 1500 CE, over 90% of mechanical energy still came from human and animal muscles. The revolutionary shift began around 1600 when England started burning coal extensively. By 1700, fossil carbon provided 75% of England's heat.
The scale of this energy transformation is staggering: a 60-fold increase in fossil fuel use during the 19th century, a 16-fold gain in the 20th, and approximately 1,500-fold over the past 220 years. When accounting for improved conversion efficiencies, the actual gain in useful energy is around 3,500-fold. In per capita terms, useful energy availability rose from a mere 0.05 gigajoules in 1800 to 34 GJ by 2020. Today's average human has access to nearly 700 times more useful energy than their early 19th century ancestors-equivalent to 60 adults working non-stop for each person globally, and 200-240 for inhabitants of affluent countries.
Despite its fundamental importance, modern economics has largely ignored energy, treating it as a minor cost factor rather than the essential foundation of all production and progress. As physicist Ludwig Boltzmann recognized in 1886, free energy is the "object of struggle" for life. Erwin Schrodinger later observed that organisms feed upon "negative entropy" (free energy), while ecologist Howard Odum noted that "all progress is due to special power subsidies."
The advantages of electricity-our most abstract energy form-have driven ever-greater electrification. Its use is effortless, clean, and exceptionally efficient. For lighting, electricity reigns supreme, with modern electric lights producing about 500 times more light per unit of energy than candles. Electric motors revolutionized both industry and daily life, doubling American manufacturing productivity by 1930 and again by the late 1960s.
Despite advances in renewable energy, decarbonization outside electricity generation has progressed slowly. Germany's Energiewende has reduced fossil fuels in its primary energy mix only from 84% to 78% over two decades. Global fossil carbon demand now exceeds 10 billion tons annually-five times more than all staple grain harvests. This massive scale makes rapid substitution impossible, especially when considering the 5+ billion people in developing regions who need more energy-not less-to improve their living standards.
第3章
Eating Fossil Fuels
Securing sufficient food is humanity's existential imperative. Our evolution from foraging to agriculture represents a fundamental transformation in how we feed ourselves. Global undernourishment has plummeted from 65% in 1950 to just 8.9% by 2019-meaning we now adequately feed 7 billion people compared to 890 million in 1950, an eightfold increase despite tripling population.
This achievement depends on a peculiar hybrid energy system combining solar energy with fossil fuels. While photosynthesis remains the fundamental process converting sunlight into food, modern agriculture relies heavily on fossil fuel inputs. The transformation of farming practices over two centuries illustrates this dependency.
In 1801 in New York's Genesee Valley, American wheat farming remained virtually unchanged from ancient Egyptian methods. Farmers used oxen pulling wooden plows, hand-sowed saved seeds, and harvested with sickles-requiring about 27 hours of human labor per hectare just for planting and at least 120 hours per hectare for harvesting.
By 1901 in North Dakota's Red River Valley, mechanization had transformed farming. Horse-drawn gang plows, mechanical seed drills, and steam-powered threshing machines reduced labor to just 22 hours per hectare. Though yields remained at 1 ton per hectare, human labor dropped to just 1.5 minutes per kilogram of grain.
In 2021 Kansas, modern farms are 3-4 times larger than a century ago yet operated by just one or two people using powerful tractors, combines, and optimized fertilizers. Human labor has plummeted to less than two hours per hectare, with yields rising to 3.5 tons. The labor required to produce a kilogram of wheat has fallen from 10 minutes in 1801 to less than two seconds today.
The nitrogen paradox is particularly striking: though nitrogen makes up 80% of our atmosphere, it exists as non-reactive N2 molecules that plants cannot use. This nitrogen barrier was decisively broken with Fritz Haber's 1909 invention of ammonia synthesis, commercialized by 1913. The Green Revolution of the 1960s, combining high-yielding crop varieties with synthetic fertilizers, transformed global food production but created unprecedented dependence on fossil fuels.
Modern food production requires substantial fossil fuel inputs, even for seemingly simple foods. Wheat cultivation needs about 4 megajoules per kilogram (equivalent to 100mL of diesel fuel), while producing a 1kg sourdough loaf requires at least 250mL of diesel equivalent. Chicken production requires 300-350mL of diesel equivalent per kilogram of meat in optimal conditions. Most shocking are greenhouse tomatoes shipped to Scandinavia, which embody approximately 650mL of diesel equivalent per kilogram-meaning a medium tomato (125g) represents over 5 tablespoons of fossil fuel input.
Returning to purely organic farming without synthetic fertilizers and machinery would require radical societal restructuring. Global croplands currently receive about 110 megatons of nitrogen from synthetic fertilizers-making it impossible to feed even half of today's population without them. As Howard Odum noted fifty years ago, "industrial man no longer eats potatoes made from solar energy; now he eats potatoes partly made of oil." Even with the fastest realistic changes, we will be eating transformed fossil fuels for decades to come.
第4章
The Four Pillars of Civilization
Modern civilization depends on materials that receive far less attention than intangible phenomena like GDP growth or digital technologies. Four materials form the indispensable pillars of modern civilization: cement, steel, plastics, and ammonia. These materials are distinguished by their massive scale of production, indispensability, and heavy dependence on fossil fuels-claiming about 17% of the world's primary energy and generating 25% of all CO2 emissions from fossil fuel combustion.
The Haber-Bosch process for ammonia synthesis may be history's most momentous technical advance. While other inventions provide comfort or save lives from disease, without ammonia synthesis, we couldn't ensure survival for nearly half of today's population. After experiencing history's deadliest famine under Mao, China ordered advanced ammonia-urea plants following Nixon's 1972 Beijing visit. Today, about 60% of nutrients available to Chinese crops come from synthetic ammonia-essentially feeding three out of five Chinese people. Globally, synthetic nitrogen feeds approximately 50% of humanity.
Plastics are synthetic organic materials designed for moldability. Their synthesis begins with monomers like ethylene and propylene (produced by steam cracking hydrocarbon feedstocks) that bond into polymer chains. Global production skyrocketed from just 20,000 tons in 1925 to 370 million tons by 2019. Their most crucial role may be in healthcare, where modern life begins and ends surrounded by plastic items-a dependency dramatically highlighted during the COVID-19 pandemic when shortages of personal protective equipment created crises in Western hospitals.
Steel defines modern civilization's appearance and enables its fundamental functions. It forms critical components of buildings, bridges, vehicles, and infrastructure. By weight, steel dominates most transportation equipment-about 900kg in the average car. With nearly 100 million vehicles produced yearly, this represents 90 million tons of steel. Even more critically, the machines that make all other machines are largely steel. Modern steelmaking combines blast furnaces with basic oxygen furnaces in an energy-intensive process consuming about 25 GJ/t, totaling 34 exajoules annually (6% of global primary energy).
Cement, the essential binding agent in concrete, is produced by heating limestone and clay to at least 1,450C in massive kilns. Since 2007, most humans have lived in concrete-made cities, where the material forms not just foundations and basements but walls, ceilings, and virtually all urban infrastructure. China's concrete consumption has exploded from less than 80 million tons in 1980 to 2.2 billion tons in 2019-over half the global total. Astonishingly, China produced in just two years (2018-2019) nearly as much cement (4.4 billion tons) as the United States did during the entire 20th century.
The transition to renewable energy actually increases material demands. Wind turbines, symbols of "green" electricity, embody fossil fuels in their massive components: reinforced concrete foundations, nearly 200 tons of steel per megawatt capacity, and approximately 15 tons of difficult-to-recycle plastic resin blades for mid-sized turbines. Electric vehicles exemplify the new material dependencies-a typical lithium car battery weighing 450 kilograms contains 11kg of lithium, 14kg of cobalt, 27kg of nickel, 40kg of copper, 50kg of graphite, and 181kg of steel, aluminum, and plastics.
Modern economies will always require massive material flows. Until all energy used in extracting and processing these materials comes from renewable sources, civilization will remain fundamentally dependent on fossil fuels. No digital technology, artificial intelligence, or electronic communication can change this material reality.
第5章
The Reality of Globalization
Globalization manifests in countless everyday phenomena: container ships delivering goods worldwide, complex manufacturing networks spanning dozens of countries, international migration patterns, and massive financial flows. By 2020, global merchandise trade approached $20 trillion annually, commercial services reached $6 trillion, foreign direct investment neared $1.5 trillion yearly, and daily currency trading totaled nearly $7 trillion.
Despite popular belief, globalization isn't new, inevitable, or primarily driven by labor cost arbitrage. At its most fundamental level, globalization involves moving physical mass (raw materials, products, people) and transmitting information across continents. While human and animal power could accomplish limited transfers, only technological innovations enabled large-scale, affordable, reliable global exchange.
The earliest phase of globalization faced severe limitations. Human porters could carry only 40-50kg and animal caravans 100-150kg per animal. Dutch shipping records illustrate these constraints perfectly: voyages to Batavia (Jakarta) averaged 238 days in the 17th century-an effective speed of just 4.7km/hour, comparable to slow walking. Only 1,450 ships were built over two centuries, averaging just 700-1,000 tons capacity, making bulk commodity trade uneconomical except for high-value goods like spices and tea.
Globalization's first quantitative leap came through three crucial developments: reliable navigation (with John Harrison's accurate chronometer in 1765), steam power replacing sails, and telegraph enabling near-instant communication. These innovations quadrupled global trade between 1870-1913, increasing trade's share of worldwide economic product from 5% in 1850 to 14% by 1913.
The next wave emerged through diesel engines, airplanes, and radio. Diesel engines, designed by Rudolf Diesel by 1897, achieved 30% efficiency-double that of steam engines. Aviation progressed from the Wright brothers' first flights in 1903 to commercial airlines by the 1920s, dramatically reducing journey times from weeks to days. Radio communication transformed both navigation and information dissemination.
The post-1950 era of intensive globalization was powered by revolutionary technologies like Boeing's 747 wide-body jetliner, which made its first commercial flight in January 1970. Over five decades, Boeing delivered 1,548 of these aircraft, carrying an estimated 5.9 billion passengers-about 75% of the world's population. Computing advances were equally transformative, with Intel's groundbreaking 4044 microprocessor in 1969 placing over 2,000 transistors on a single silicon wafer.
The unprecedented expansion of globalization after the 1970s became possible when three previously isolated economic giants joined the global market system. China's opening began with Nixon's 1972 Beijing visit, accelerated with Deng Xiaoping's economic reforms after 1978, and reached full force when China joined the WTO in 2001. The Soviet Union's collapse in 1991 opened another vast market, while India's economic liberalization produced sustained growth above 4% for 18 of the 19 years between 2000-2019.
While globalization shows a long-term trend toward greater international economic integration, history proves it isn't inevitable. The first half of the 20th century saw significant retreat from economic integration amid unprecedented tragedies-World War I, the Bolshevik Revolution, the Great Depression, World War II, and the Cold War's beginning. Trade's share in global GDP plummeted from 14% in 1913 to just 4% by 1945.
Growing concerns about globalization's impact on national sovereignty, culture, inequality, and economic security-intensified by COVID-19-have prompted calls for reshoring manufacturing. The pandemic exposed critical vulnerabilities when 70% of rubber gloves came from a single factory and similar concentrations existed for medications and protective equipment. While much of today's globalization seems permanent, it appears increasingly likely we may have seen peak globalization, with its ebb potentially lasting not just years but decades.
第6章
Understanding Risk in a Complex World
Modern civilization can be viewed as a series of quests to reduce the risks inherent in human existence. We've successfully improved food security, eliminated recurrent famines, enhanced housing conditions, implemented public health measures, and reduced industrial and transportation accidents. Despite media impressions, violent conflicts have actually declined globally over decades. Yet risks persist due to our biological complexity, unpredictable natural processes, and inevitable human error in designing and operating intricate systems.
The fundamental difference between voluntary and involuntary risks shapes our risk tolerance. People willingly engage in activities with thousand-fold higher risks than dreaded involuntary exposures-rock climbing without ropes versus terrorist attacks. Nuclear power exemplifies this risk misperception. Many who smoke, drive recklessly, and overeat fear living near nuclear plants, despite evidence that nuclear power prevents more deaths than it causes by displacing air pollution from fossil fuels.
Using fatalities as our universal metric allows meaningful risk comparisons. The baseline mortality risk in affluent countries is approximately 1x106 per person-hour-meaning one person among a million dies every hour. Driving carries a fatality risk of approximately 5x107 per hour of exposure, increasing one's overall death risk by about 50% while behind the wheel compared to staying home. Commercial flying has become remarkably safer, with a risk of merely 2.8x108 per person-hour, increasing one's baseline mortality risk by only 3% while aloft.
Despite their dramatic media coverage, natural disasters pose surprisingly minimal statistical risks. Tornadoes in America's most affected states killed 1,994 people from 1984-2017, translating to just 3x109 fatalities per hour of exposure-1,000 times lower than simply being alive. Japanese earthquakes killed about 33,000 people between 1945-2020, representing only 5x1010 fatalities per hour-adding a mere 0.0001 to baseline mortality risk.
When considering existential threats, the Yellowstone supervolcano presents a tangible American concern. A Yellowstone eruption would release approximately 1,000 cubic kilometers of ash, potentially burying areas under meters of ash and destroying housing, infrastructure, and agricultural land. Another civilization-threatening risk is a catastrophic geomagnetic storm from a coronal mass ejection. The 1859 Carrington Event disrupted telegraph systems, but today's electronic infrastructure is vastly more vulnerable. A similar event now would potentially disable global electricity grids, communications networks, and navigation systems for years.
Viral pandemics represent a more frequent existential threat. People over 62 in 2020 had already lived through three pandemics: 1957-1959 (H2N2, 38 deaths per 100,000), 1968-1970 (H3N2, 28 deaths per 100,000), and 2009 (H1N1, 3 deaths per 100,000). Yet the World Economic Forum consistently failed to rank pandemics among top global risks between 2007-2015, demonstrating our collective failure to prepare for predictable threats.
第7章
Our Environmental Reality
This chapter addresses our relationship with Earth's only biosphere. For our species to survive another 5,000 years, we must ensure our interventions don't imperil Earth's long-term habitability by transgressing safe planetary boundaries. These critical boundaries include climate change, ocean acidification, stratospheric ozone depletion, atmospheric aerosols, interference in nitrogen and phosphorus cycles, freshwater overuse, land use changes, biodiversity loss, and chemical pollution.
Despite concerns about forest fires, the atmosphere's oxygen content remains remarkably stable. Terrestrial and marine photosynthesis annually absorb and release about 300 billion tons of oxygen in a nearly balanced cycle. Images of burning forests in the Amazon, Australia, California, or Siberia are destructive and harmful in many ways, but they won't suffocate us through oxygen depletion.
Water provision should be high on our environmental priority list-not because of absolute shortage but due to uneven distribution and mismanagement. While we don't need to drink as frequently as we breathe, adequate drinking water (1.5-3 liters daily) is basic survival. Beyond this existential need (about 750 liters annually), we require much more water for hygiene, cooking, laundry, productive activities, and especially food production. With 2C warming, between 500 million and 3.1 billion people may face climate-induced water shortages. However, demand management offers solutions, as demonstrated by the US, which reduced per capita water use by 40% over 50 years despite population growth and economic expansion.
The greenhouse effect is an existential imperative-without it, Earth would be permanently frozen at -18C. Trace gases like carbon dioxide and methane, though present in negligible amounts compared to nitrogen and oxygen, regulate Earth's temperature by absorbing outgoing infrared radiation, raising surface temperature by 33C to a life-supporting global mean of 15C. Human activities began affecting trace gas concentrations thousands of years ago through farming, forest clearing, and wood burning. These anthropogenic emissions became significant only with industrialization. CO2 levels rose from a pre-industrial 270 ppm to over 420 ppm by 2020-a 50% increase.
Though "global warming" only entered common parlance after 1980, our scientific understanding of the greenhouse effect dates back more than 150 years. In 1856, Eunice Foote first linked CO2 with warming, followed by John Tyndall in 1861 who explained that water vapor and "carbonic acid" in the air must produce climate change when their concentrations vary. By the century's end, Svante Arrhenius calculated that doubling atmospheric CO2 would raise Earth's surface temperature by 4C-remarkably close to modern estimates.
Despite three decades of climate discourse, global greenhouse gas emissions increased 65% between 1989-2019. While affluent nations reduced emissions by about 4%, Indian emissions quadrupled and Chinese emissions rose 4.5 times. The 2015 Paris accord, despite being labeled "landmark" and "ambitious," contained no specific reduction targets and would still result in a 50% increase in emissions by 2050 if all voluntary pledges were honored.
Climate scenarios promising rapid decarbonization consistently fail to account for real-world constraints. One EU-produced scenario assumes global per capita energy demand will drop 52% by 2050-an astonishing reversal given that it rose 20% over the previous three decades. This projection ignores developing nations' aspirations, allowing only a factor-of-two increase in mobility and factor-of-three increase in consumer goods ownership, while China's car ownership grew 100-fold and air conditioner ownership 400-fold in just two decades.
第8章
Between Apocalypse and Singularity
The final chapter explores the etymology of "apocalypse"-from Greek meaning "uncovering," which in Christian context shifted toward prophesied revelation of the second coming, and in modern usage has become synonymous with the end of life on Earth. In contrast, techno-optimists believe in salvation through technology, with some like Ray Kurzweil predicting a "Singularity" in 2045 when machine intelligence will surpass human intelligence. Modern anticipations of the future gravitate toward these two extremes despite evidence suggesting neither is likely.
The history of forecasting is littered with spectacular failures. Nuclear power's grand promises fell dramatically short-while it provides 10% of global electricity, this represents just a fraction of what was expected. Beyond electricity generation, nuclear ambitions included outlandish projects like nuclear-powered aircraft (abandoned in 1961 after billions in spending) and natural gas production via nuclear explosions (including a 29-kiloton bomb detonated underground in New Mexico in 1967).
Today's forecasting landscape remains similarly flawed. Environmental catastrophism dominates, with warnings of climate apocalypse and civilization collapse within a decade. Meanwhile, techno-optimists predict a near-future "Singularity" by 2030-2045 that will solve all problems through machine intelligence. History suggests neither extreme will materialize-we'll neither face civilization's end nor achieve godlike omnipotence.
Despite humanity's remarkable adaptability, fundamental constraints persist in our physical systems. Large, complex systems resist rapid change due to their basic energetic and material requirements. Steel production has theoretical energy minimums of 18 gigajoules per ton, and ammonia synthesis requires at least 21 gigajoules per ton. The scale of modern production further constrains transitions-what could replace 4 billion tons of cement or 2 billion tons of steel annually?
COVID-19 provided a costly global reminder of our limited capacity to predict the future. Despite a decade of celebrating "disruptive" advances in artificial intelligence and genetic engineering, none of these much-touted capabilities prevented the pandemic's spread. We resorted to medieval methods-isolation for quaranta giorni (forty days). While vaccines came relatively quickly, they neither cured the infected nor prevent future pandemics.
Addressing climate change requires a truly global, substantial, and prolonged commitment unprecedented in human history. Any effective climate commitments will be expensive and must last at least two generations to achieve meaningful reductions in greenhouse gases. Even with drastic action beyond what seems politically feasible, benefits won't materialize for decades due to the long atmospheric lifetime of greenhouse gases. Climate-economy models indicate that mitigation efforts launched in the early 2020s would only begin producing net economic benefits around 2080-meaning the generation born mid-century would be the first to experience benefits from our sacrifices.
The reality is that effective climate solutions will be non-magical, gradual, and costly. Affluent countries could significantly reduce energy use while maintaining comfort through simple measures like triple windows and more durable vehicles. Halving food waste and changing meat consumption would help more than promised "revolutions" dependent on unavailable mass-scale electricity storage. The question remains whether we'll act with foresight, respond only when forced, or fail to act meaningfully at all.
