Chapter 1
Shaping Humanity's Vast Future: Our Greatest Moral Responsibility
When Elon Musk was asked which book had most influenced his thinking in recent years, he didn't hesitate: "What We Owe The Future" by William MacAskill. This philosophical tour de force has captured the imagination of influential thinkers across industries, from tech titans to political leaders. MacAskill's concept of "longtermism" - the idea that positively influencing the long-term future is a key moral priority - has sparked a movement that's reshaping how we think about our responsibilities to future generations. Named one of the best books of 2022 by publications ranging from The Economist to The New York Times, it arrives at a critical moment when humanity faces unprecedented technological risks and opportunities that could determine our species' fate for millions of years to come.
Chapter 2
The Moral Revolution of Taking the Future Seriously
Imagine living every human life in sequence - all 100 billion lives spanning 300,000 years of human history. You'd spend 10% of your time as a hunter-gatherer, 60% as an agriculturalist, and just a fraction in our modern technological era. But here's the staggering truth: if humanity survives even a million years (the typical mammalian species lifespan), 99.5% of all human lives still lie ahead. On the scale of a human life, we're just five months old.
This perspective reveals the moral importance of safeguarding humanity's future. The case for longtermism rests on three simple principles: future people count morally just as much as present people; there could be an astronomical number of them; and our actions today might determine whether their lives are wonderful or terrible.
Future people are utterly disenfranchised - they cannot vote, lobby, or make their views heard. Yet taking them seriously represents a moral revolution comparable to previous movements that expanded moral consideration to disempowered groups. Though we cannot give future people political power, we can act as trustees for generations to come.
The potential scale of the future is staggering. If humans survive just as long as the typical mammalian species, there would be eighty trillion people yet to come, outnumbering us ten thousand to one. And this is conservative - with our unique adaptability, humanity might survive hundreds of millions of years until Earth becomes uninhabitable, or even trillions of years as the universe slowly cools.
The future could be extraordinarily good. In just the last two centuries, global life expectancy has risen from under thirty to seventy-three years. Extreme poverty has fallen from over 80% to under 10%. We could build a world where everyone lives like the happiest people in the most well-off countries today, free from poverty and inadequate healthcare. Yet dystopia is equally possible - totalitarianism, environmental collapse, or civilization-destroying war. Whether the future is wonderful or terrible depends partly on us.
Chapter 3
How Our Actions Today Shape Millennia to Come
We affect the long-term future constantly through everyday actions. Greenhouse gas emissions will impact Earth for hundreds of thousands of years - timescales comparable to nuclear waste. In the medium-low emissions scenario, sea levels would rise ten to twenty meters over ten thousand years, submerging major coastal cities worldwide.
History shows deliberate attempts to influence the distant future can succeed. Shakespeare and Horace created works that have lasted centuries. Thucydides, writing in the fifth century BC, deliberately crafted his history "to last forever" and still influences military academies today. America's Founding Fathers consciously designed institutions for "thousands of years," with Benjamin Franklin even investing money to benefit Boston and Philadelphia centuries later.
The seemingly wild implications of longtermism stem not from its moral premises but from the unusual historical moment we inhabit. We live in an era of extraordinary change - with economic growth at historically unprecedented 3% per year compared to near 0% for most of human history. This rapid rate of change cannot continue forever due to physical limitations. Combined with our unusual global connectivity, this gives us outsized opportunity to influence humanity's future trajectory.
History often features periods of plasticity followed by rigidity, like glassblowing where molten glass can take many shapes before cooling into a fixed form. These moments of plasticity frequently follow crises or occur when institutions are new. With climate change, we missed early opportunities for action when the science was already clear but the issue less politically divisive. We should learn from this and address emerging challenges like advanced AI proactively.
Chapter 4
Moral Progress Is Not Inevitable
Despite slavery's historical ubiquity, it was eventually abolished through the efforts of many, including enslaved people who resisted and a small group of Quaker activists. The suffering of enslaved people was immense - those who survived the Atlantic crossing faced ten-hour workdays on plantations, widespread whipping, and life expectancies as low as twenty years. Slavery was widely accepted as natural, even by renowned philosophers like Plato and Aristotle.
Benjamin Lay, a four-foot-tall Quaker activist, used dramatic protests to highlight slavery's horrors - standing barefoot in snow, disrupting meetings, and once splashing fake blood on a Quaker gathering. Though ostracized during his lifetime, Lay influenced Quaker circles that dramatically shifted from 70% of leaders owning slaves (1681-1705) to just 10% (1754-1780).
This Quaker movement spread through activists like Anthony Benezet to Britain, where abolitionists eventually persuaded Parliament to abolish the slave trade in 1807 and slavery itself in 1833, marking one of history's most significant moral transformations.
Was abolition inevitable? The evidence suggests it was far from guaranteed. In China, slavery waxed and waned over millennia - from the ancient Shang dynasty (before 1500 BCE) until final abolition in 1909, with multiple attempts at reform followed by resurgence under new dynasties. Globally, it's difficult to see abolitionism as part of any consistent historical trend toward moral progress.
Moral inconsistencies can persist for remarkably long periods - consider modern examples like the legality of alcohol versus other drugs, factory farming alongside pet protection laws, or prison labor in the United States. Mississippi's Parchman Farm operated as a de facto plantation using convict labor until the 1970s, and even today American prisoners work for about one dollar per hour.
The abolitionists demonstrate not only the importance of moral change but how to achieve it. Benjamin Lay exemplifies the moral entrepreneur - someone who thinks deeply about morality, takes it seriously, acts on convictions, and willingly endures being seen as eccentric. We should aspire to be such "weirdos" who can withstand mockery while driving moral progress.
Chapter 5
The Perilous Possibility of Value Lock-In
The collapse of China's Zhou dynasty in the sixth century BC created both conflict and a golden age of philosophical experimentation known as the Hundred Schools of Thought. During this period, philosophers traveled between states promoting competing theories and moral systems to political elites.
During China's Warring States era, four major philosophical traditions competed for influence: Confucianism, Legalism, Daoism, and Mohism. These schools engaged in bitter rivalries until 221 BC, when the Legalism-influenced Qin dynasty conquered China and violently suppressed competing ideologies. The first Qin emperor ordered book burnings, prohibited "private learning," and executed hundreds of dissenting scholars.
Following popular revolt and years of conflict, the Han dynasty emerged. Initially guided by a blend of philosophies, Confucianism gradually became the orthodox state ideology through skillful political maneuvering. Despite periods when Buddhism and Daoism flourished alongside it, Confucianism dominated Chinese intellectual life for over a thousand years.
We are currently living through a global equivalent of China's Hundred Schools of Thought - a period of moral plasticity where different worldviews compete and no single one has won out. Like molten glass that can be shaped into many forms, our philosophical culture remains malleable. But technological advances, particularly artificial intelligence, could bring this period of diversity and change to an end.
Artificial intelligence aims to create machines that mimic human intelligence. Recent breakthroughs in machine learning have produced remarkable achievements like AlphaGo beating world champion Lee Sedol, and AI systems that generate speech, images, art, and text. The ultimate goal is artificial general intelligence (AGI) - a system capable of learning and performing as wide an array of tasks as humans can.
AGI would be monumentally important for two reasons. First, it could dramatically accelerate technological progress and economic growth. By automating innovation itself and allowing unlimited "reproduction" of AI workers, growth rates could increase from today's 3% annually to levels where economies double in size over months rather than decades.
Second, AGI has potential immortality. Just as software like the 1977 game Pong can be preserved indefinitely through replication, AGI could persist as long as civilization exists. This combination of rapid technological progress and immortality makes value lock-in a real possibility.
The development of AGI presents a profound challenge similar to a child appointing an adult to rule a country - how do we ensure vastly more intelligent systems act according to our wishes? As Nick Bostrom outlined in "Superintelligence," an AI with goals different from humanity's would seek resources and prevent threats to its existence, potentially leading to human disempowerment or elimination.
Chapter 6
Existential Risks: The Threats That Could End Everything
Arthur C. Clarke's novel "Rendezvous with Rama" begins with a devastating asteroid strike in Italy that kills 600,000 people and destroys priceless cultural treasures. This fictional disaster prompts humanity to create Spaceguard, an early-warning system for Earth-bound threats from space. In reality, scientists' warnings about asteroid dangers were often met with a "giggle factor" until comet Shoemaker-Levy 9 crashed into Jupiter in 1994 with force equivalent to 125 times the world's nuclear arsenal.
This event increased public awareness, and in 1998, Congress tasked NASA with finding 90% of all near-Earth objects larger than one kilometer, naming the effort "Spaceguard" in homage to Clarke. The program has been remarkably successful, tracking 93% of large asteroids and 98% of extinction-threatening ones. We now know the risk of an extinction-level asteroid impact is less than one in fifteen billion - a hundred times lower than previously estimated.
Despite COVID-19's devastating toll of seventeen million deaths and $10 trillion in economic damage, we've gotten off relatively easy. Future pandemics, particularly those engineered through biotechnology, could be far more destructive. Biotechnology has progressed at a staggering pace - even faster than Moore's law - with human genome sequencing costs plummeting from hundreds of millions to just $1,000 in twenty years.
Many extinction risk specialists consider engineered pandemics the second most likely cause of human extinction this century, just behind artificial intelligence. Metaculus forecasts a 0.6% chance of an engineered pandemic killing at least 95% of humanity by 2100, while expert Toby Ord estimates 3%.
The Cold War's bioweapons race demonstrates how war drives dangerous technological development. While we've enjoyed a "Long Peace" since World War II with no direct great-power conflicts, this peace cannot be taken for granted, as Russia's invasion of Ukraine reminds us. Following this invasion, Metaculus more than doubled its predicted chance of a third world war by 2050 to 23%.
Great-power war amplifies virtually all other existential risks - it diverts resources from safety measures, destroys international cooperation, increases the likelihood of bioweapons development, nuclear strikes, AI arms races, and undermines climate change mitigation efforts. Reducing the likelihood of world war is therefore one of the most important ways to safeguard civilization this century.
Chapter 7
Collapse, Stagnation, and Recovery
In AD 100, the Roman Empire stood alongside the Han dynasty as one of two superpowers encompassing half the world's population. Rome represented the pinnacle of advancement with central heating, durable concrete, massive structures like the Colosseum, sophisticated economic systems, and a population of one million in its capital. After centuries of growth and stability, Rome's decline was dramatic - sacked twice in the 5th century and eventually collapsing. Rome's population plummeted to just 30,000, not recovering for 1,400 years.
Despite historical collapses like Rome's, human civilization has shown remarkable resilience. Even when the Western Roman Empire fell, Europe wasn't depopulated - Roman rule was simply replaced by various Germanic tribes. While technological sophistication declined dramatically, this decline wasn't permanent. The Renaissance, Scientific Revolution, and Industrial Revolution eventually emerged from these ashes.
Even in a worst-case nuclear scenario where 99% of humanity perished, civilization would likely survive. The physical infrastructure, knowledge in libraries and digital storage, and critical skill sets would remain. With 2 billion people working in agriculture today, agricultural knowledge would persist. The impacts would be geographically diverse - coastal areas in the Southern Hemisphere would be less affected, with places like Australia and New Zealand potentially continuing relatively unscathed.
Recent years have brought unprecedented hope in addressing climate change. The International Energy Agency predicts that global coal use peaked in 2014 and is now in structural decline. Solar panel costs have fallen by a factor of 250 since 1976, while lithium-ion battery costs have dropped 41-fold since 1991. Though solar and wind currently supply only about 3% of global energy, continued exponential cost declines suggest they'll provide a substantial portion within twenty years.
However, we shouldn't get complacent. There's a substantial chance decarbonization efforts will stall due to breakdowns in international coordination or technological stagnation. In a worst-case scenario where we burn through all recoverable fossil fuels, we'd likely see 7 degrees of warming with a one-in-six chance of 9.5 degrees.
The Islamic Golden Age of the eleventh century saw Baghdad become the global center of scientific progress, producing remarkable innovations: magnifying lens theory, flywheel-powered water-lifting devices, the first programmable machine, and pioneering code-breaking methods. However, this golden age faded from the twelfth century onward due to various factors including the Mongol invasion, the Crusades, or cultural shifts prioritizing theology over science.
What distinguishes our modern era is that technological progress and economic growth have been sustained at unprecedented levels since the Industrial Revolution. But the question remains: will this continue, or will future historians view our era as just another efflorescence that eventually stagnated?
Chapter 8
The Philosophical Case for Valuing Future Lives
Derek Parfit, one of the most influential moral philosophers of the last century, dedicated his life to advancing our moral understanding with extraordinary single-mindedness. Living mostly at Oxford's exclusive All Souls College, he minimized time spent on anything but philosophy - running between seminars, eating the same simple meals daily, and even reading while brushing his teeth.
Parfit argued that extinction would be catastrophically worse than even a disaster killing 99% of humanity, because extinction would prevent the existence of countless future generations. He maintained that preventing a happy life from existing constitutes a moral loss - a view I initially found counterintuitive but came to accept through philosophical argument.
The view that creating happy people is morally neutral rather than good is often considered intuitive. Jan Narveson expressed this as "We are in favour of making people happy, but neutral about making happy people." However, this intuition becomes problematic in extreme cases. If creating miserable lives is clearly bad - imagine bringing into existence someone who would experience nothing but agony - why wouldn't creating happy lives be good? This asymmetry is difficult to justify philosophically.
Parfit noted that our existence is extraordinarily contingent and identity is exceptionally fragile, with profound ethical implications. Our identities depend on astronomical odds - any of the 200 million other sperm in our father's ejaculation would have created someone else with 75% of our genes. Even millisecond differences in conception timing would have resulted in completely different people.
This fragility means our everyday choices constantly alter future identities. By taking public transport instead of driving, you slightly impact thousands of people's schedules, statistically changing at least one conception timing annually. This creates an "identity cascade" where, beyond a certain point, everyone born will differ from who would have existed otherwise, completely changing future history.
A simple logical argument demonstrates why the intuition of neutrality fails. Consider parents deciding whether to have a child. Due to a temporary maternal vitamin deficiency, conceiving now would result in a child who suffers from migraines throughout life but otherwise lives well. Alternatively, they could wait a few months and have a migraine-free child, or have no child at all.
According to the intuition of neutrality, having no child is equally good as having a child with migraines (option A = B). Similarly, having no child is equally good as having a migraine-free child (option A = C). But by simple logic, if A = B and A = C, then B must equal C. This creates a contradiction, because having a migraine-free child is clearly better than having a child who suffers from migraines when all else is equal.
Chapter 9
Will the Future Be Good or Bad?
Imagine living every human life ever lived - would you consider this collective existence worthwhile? Now extend this thought experiment to all sentient beings: you would experience 100 billion trillion years of vertebrate consciousness, with 80% as fish, 20% as amphibians or reptiles, and human existence comprising just one-hundred billionth of your time.
Let's begin by asking whether the world is better than nothing for humans alive today. Do most people have net-positive lives? This sensitive question matters not just for longtermists but also for practical policy decisions like healthcare prioritization.
Life satisfaction surveys of 1.5 million people across 166 countries found only 47% scored above 5 on a 0-10 scale, though interpreting where "neutral" falls is difficult. Happiness surveys show most people report being happy, with over 70% in almost all countries saying they're "very" or "rather" happy, though these may be optimistic.
To gather more data, I commissioned psychologists to survey Americans and Indians about their life quality. The results were striking: In the US, 16% said their life contained more suffering than happiness, 9% preferred never to have been born, and 31% wouldn't live their exact same life again. Indian respondents were notably more positive, with only 11% reporting more suffering than happiness and just 6% wishing they had never been born.
Putting it all together, I'd tentatively suggest that about 10% of the global population have lives with below-neutral wellbeing. If I were given the option to be reincarnated as a randomly selected person alive today, I would choose to do so. If I were to live through everyone's lives, I would be glad to have lived.
When assessing whether the world is good on balance, we must consider the vast majority of sentient beings: nonhuman animals. The scale of animal suffering in factory farming is staggering - approximately 79 billion vertebrate land animals and 100 billion fish are slaughtered for food annually.
The conditions these animals endure are horrific. Broiler chickens grow so rapidly that 30% develop severe walking problems. Their slaughter involves being hung upside down, electrocuted, and having their throats cut - with millions surviving only to die in scalding water. Egg-laying hens fare even worse, confined to battery cages smaller than a sheet of paper after having their beaks painfully cut.
When considering the long-term future, we must weigh the likelihood of reaching eutopia versus anti-eutopia. Despite the greater badness of anti-eutopia, there are strong reasons to believe eutopia is much more likely.
People rarely produce bad things simply because they are bad. Even the worst atrocities typically occur not because people actively seek suffering, but as side effects or means to other ends. Factory farming persists not because people enjoy animal suffering, but because they want cheap meat. Similarly, slavery and war were pursued for profit, power, and glory.
Chapter 10
Taking Action for the Long-Term Future
In the Aymara language of Bolivia and surrounding regions, the future is conceptualized as being behind us (unseen) while the past is before us (visible). This metaphor aptly describes our relationship with the future - we're walking backwards into unknown terrain, able only to glimpse what's coming by looking over our shoulder.
Despite uncertainty about the future, we can still make wise choices. Like explorers in uncharted territory, we should: (1) take actions we can be confident are good across multiple scenarios, (2) increase our options rather than limiting them, and (3) learn more about what might lie ahead. This means promoting robustly beneficial technologies like clean energy, maintaining diverse cultures and political systems, and investing in better long-term forecasting and research.
When tackling global problems, we must prioritize like an expedition facing multiple threats - some issues demand immediate attention while others can wait. Rather than choosing causes close to our hearts or those most visible, we should evaluate problems using significance, persistence, contingency, tractability, and neglectedness.
Some issues allow for robustly good actions: climate change (supporting organizations like Clean Air Task Force), biosecurity (promoting universal diagnostics), and general disaster preparedness (building food stockpiles, bunkers, and information vaults). For complex problems like AI safety, the key is building fields of morally motivated actors who can reduce uncertainty.
While many focus on personal consumption choices to address global problems, this emphasis is often a strategic blunder. Consider plastic reduction - reusing a plastic bag 8,000 times equals the environmental impact of avoiding just one transatlantic flight. Even meaningful personal choices like vegetarianism pale compared to systemic approaches: donating 10% of a median American income to effective climate organizations could reduce emissions by 3,000 tonnes yearly - far exceeding a lifetime of vegetarianism.
Career choice is the most impactful decision for your lifetime impact. When choosing a career, follow three key principles: learn, build options, and do good. First, treat your career like an experiment - research options, make your best guess, try it for a couple years, update, and repeat. Second, build career capital by investing in skills like organizational leadership, political influence, research abilities, communication, entrepreneurship, and community building. Finally, focus on doing good by choosing high-impact problem areas.
Beyond direct impact, spreading longtermism itself can be enormously powerful. Convincing just one person to do as much good as you would have done doubles your impact. Recent environmental successes, like Germany's solar initiatives that drove down global costs by 92% between 2000-2020, show how future-oriented activism can have transformative effects. The challenges ahead - decarbonizing the economy, reducing risks from war, engineered pathogens, and AI - require a movement of morally motivated people concerned about the future. Positive moral change isn't inevitable; it requires work by generations of thinkers and activists. We live at a pivotal time of extraordinary change when we can steer the future onto a better trajectory.
