What We Owe The Future

People matter if they live thousands of miles away — and thousands of years away too.

That’s among the primary arguments from What We Owe the Future, a 2022 book by the Scottish philosopher and ethicist William MacAskill that popularized a concept of longtermism (which has coincided with effective altruism).

Below I share my notes for future reference.

My notes:

• “Future people count. There could be a lot of them we can make their lives go better.”
• We operate with “Tyranny of the present” over future people
• People matter if they live thousands of miles away and thousands of years away too
• Oren Lyons of Iroquois Confederacy: we hold a “seventh-generation” principle
• Why do we care more about present? It’s special relationships and reciprocity
• If homo sapiens lasted 700k more years to reach 1m years of existence (the lifespan of a typical mammal species) and our population remained flat, there would be 80 trillion people to come — 10k times more then people today (Roughy 100b people have lived so far)
• Thucydides was perhaps oldest surviving writer to have actively written for the future
• In 1784, Ben Franklin put a thousand pounds ($135k today) in a fund for Boston and Philadelphia because he was teased about being so optimistic for the country’s future; by 1990, Boston’s was $5m and Philadelphia was $2.3m (as has been written)
• The Founding Fathers were influenced by Locke (1632-1704) and Montesquieu (1689-1755), who were in turn influenced by Polybius’s 2nd-century BCE analysis of Roman governance (especially separation of powers)
• Decarbonization is the ideal case study for long term thinking; It benefits us now but the costs are really justified by longterm benefits
• From 10k BCE onward it took centuries for the global economy to double; the most recent doubling happened in 19 years. It can’t keep doing that; if global growth slowed to 2%, in 10k years there’s 10m trillion times as much output as accessible atoms within 10k light years
• African megafauna evolved alongside humans so they were more prepared for predatory Homo sapiens and resulting environmental change than other megafauna like dire wolf and glpytodonts who died off in South America after human contact (Reminds ya of Guns, Germs and Steel, or 1491)
• Author’s framework for importance: Significance (does it matter), persistence (will it last) and contingency (was it inevitable, or did something really change?).
• Change the duration of humanity and change its trajectory
• Expectancy Value Theory (Vroom, 1964): Motivation is determined by (a)how likely what you’re doing will result in the stated outcome and (b) how much you want that outcome
• “Early plasticity, later rigidity” for ideas and technologies: We can shape the trajectory of shifts when they’re new but seem inconsequential, but by the time it feels real it’s too late and entrenched. The lesson is that even if something has a small chance of being a big problem we should take early measures.
• Why did abolition happen in England in early 1800s and not sooner? Was it contingent (specific actors or events made it so) or inevitable? Benjamin Lay, a Quaker who later lived in Philadelphia, played a major role. Would someone else have filled his role if he hadn’t?
• In 1681-1705, an estimated 70% of leaders at the Quakers Yearly Meeting owned slaves; in 1754-1780, that figure was 10% and falling. Why then and there?
• Promote a goal not a policy: this is value change author wants (ie abolition of slavery)
• This is the “dead hand problem” in philanthropy in which an old mandate makes no sense in the new world. (Ie ScotsCare is a charity promoting welfare of Scots in London which made sense in 1611 but less so today; in Philadelphia, Girard College was founded to assist poor white boys, but times changed)
• Culture is “any socially transmitted information “
• Cultural evolution like Darwinian evolution: variation, differential fitness and inheritance
• How contingent was the abolition of slavery?
• In the 1944 book Capitalism and Slavery, Eric Williams makes the influential argument that the abolition of slavery was just ending an already dying to institution. This author think it’s more complex. A 1977 book Econocide shifted historians to believing it was more complex too
• The shelf price of sugar increased 50% in Britain after slavery: it had economic consequence (5% British expenditure over 7 years)
• “Even though Britain had abolished the slave trade in 1807, more Africans were taken to the transatlantic slave trade between 1821 and 1830 than in any other decade except the 1780s.”
• Political scientists Robert Pape and Chaim Kaufmann described Britain’s anti-abolition naval and military crusade between 1807-1867 as “the most expensive international moral effort in modern history”
• By the 12th century, slavery was largely replaced in France and England by serfdom, but following a wave of 14-century Black Death ,the institution returned as a response to labor shortages.
• Christopher Leslie Broen in Moral Capital wrote that “anti-slavery organizing was odd rather than inevitable”
• Confucius emerged in The Hundred Schools of Thought era (6th century BC to 221 BC), when various philosophies emerged like legalism and Daoists and the Mohists too, which had similarities to John Stuart Mill and Bentham
• William Nordhaus: artificial general intelligence (AGI) would do at least two things for growth: a country’s economic growth would grow at the rate at which productive AGI could be produced and AGI could automate the process of technological innovation. We’d likely run into limitations to that growth though, such as the mining of material; developing computer chips or sourcing energy
• Ajeya Cotra says AGI would be like a boy king trying to pick which adult should serve as surrogate without knowing if they’re scheming or genuine (alignment)
• Nick Bostrom’s Superintelligence and Brian Christian’s The Alignment Problem
• Values lock in: This relates to the “later rigidity” idea, since AGI might follow the values we initially establish
• JBS Haldane’s 1927 The Last Judgement essay was perhaps first time someone explored humans living on other planets; He predicted it would take 8m years for a return trip to the moon
• Ajeya Cotra: AI systems today are about as powerful as insect brains, 1% of human brains
• “Based on extrapolations of these trends, and our best guess is from neuroscience, Coltra found that we are likely to train AI system that uses much computation as a human brain within roughly the next decade, and that we may well have enough computing power to essentially simulate the complete history of biological evolution by the end of this century.”
• Cotra: 10% chance of AGI by 2036, 50% by 2050
• In 2016, Katia Grace polled 350 machine learning researchers and there was great range in when they thought AGI would arrive, if at all. AI researchers did predict that AI would take over all human tasks sooner than they’d take over AI research (which is of course impossible but funny how we assume our work is hardest)
• The non religious has shifted from 16% to 13% of global population (trends)
• Author argues we did not see enough country-level diversity in approaches to the Covid vaccines to allow for genuine learning: Why did no country try “challenge trials” or allow people to buy the vaccine early? These could have gotten us to vaccines early through experimentation. But we have a value lock in of some of these values globally. Many think alternatives are abhorrent so we don’t ry
• John Stuart Mill is credited in his 1859 classic On Liberty for arguing for the marketplace of ideas
• Charter cities could be used for any experiment (Deng Xiaoping started Shenzen and its success changed China’s economic pathway
• For every 250 full time lab workers working for a year we should expect 1 accidental infection
• Metaculus estimates a 0.6% probability of an extinction-level event by an engineered pandemic killing 95%+ people by 2100; Tony Ord in the 2020 book Precipice puts it at 3%
• Did Russia and Ukraine end the Long Peace?
• Would a different intelligent technologically capable species evolve over millions of years if human went extinct? Maybe not, it seems rare
• “In the accounts of several leading economic historians, the comparative political fragmentation of Europe, after the fall of Rome, partly explains why the scientific and industrial revolution occurred there rather than in China.” 124
• The General Crisis period (17th and 18th-century Europe) was brutal but global population still climbed
• Extinction and civilizational collapse are two different levels of thread
• Human shadow of death from Hiroshima: Hiroshima had limited rail service the day after it was bombed (lots of other examples of its rapid rebound)
• Stanislav Petrov: the famous Soviet analyst that overrode a false alarm to avoid initiating a nuclear attack
• Ought we plan to leave easier-to-use excess carbon fuel like coal for possible reindudrialization for humans or human-like species who rebound after a catastrophic collapse?
• Efflorescences and Economic Growth in World History: Rethinking the “Rise of the West” and the Industrial Revolution – a 2002 paper by Jack Goldstone
• Robert Gordon shout out on his analysis of economic slowdowns — if it’s just slowing that’s inevitable but going to zero of stagflation is bad as Tyler Cowen popularized in his 2011 book
• Are ideas getting harder to find paper ?: we’ve thrown more researchers at problems to overcome the lowest hanging fruit problem (which is a bigger problem than the opportunities of growing on past research and tools)
• Excerpt from p 152: “Over the past century, we’ve seen relatively steady, though slowing, technological progress. Sustaining this progress is the result of a balancing act: every year, further progress gets harder, bur every year we exponentially increase the number of researchers and engineers. For instance, in the United States, research effort is over twenty times higher today than in the 1930. The number of scientists in the world is doubling every couple of decades, such that at least three-quarters of all scientists who have ever lived are alive today. So far, exponential growth in the number of researchers has compensated for progress becoming harder over time. So to think about whether we can sustain technological progress, we have to think about whether we can keep exponentially growing the number of researchers. Consider that there are two ways to do this. First, you can increase the share of the population that is devoted to research. Indeed, we’ve been doing a lot of that, so that’s been the source of most of US technological progress in the last few decades. Technology-driven growth of US per-capita incomes has averaged about 1.3 percent per year. A full percentage point of that comes from increasing the fraction of the population doing R&D and from improving the allocation of talent, such as by reducing gender and racial discrimination. The second way by which you can increase the number of researchers is by increasing the total size of the labour force: that is, you can grow the population. Over the last few decades, population growth has contributed about 0.3 percentage points to the United States’ technologically driven per-capita growth rate. Historically, increasing population sizes have been a major factor in rates of technological progress. As Nobel Prize-winning economist Michael Kremer has noted, sheer population size seems to explain a big part of the very long-run comparative development of different geographic regions. With the end of the most recent ice age in 10,000 BC, five regions of the world became mutually isolated from one another: the Eurasian and African continents, the Americas, Australia, Tasmania, and Flinders Island. By AD 1500, they had dramatically diverged technologically. The more populous a region was in 10,000 BC, the more complex their technology was. By AD 1500, Eurasia had the most complex technology; the Americas followed with cities, agriculture, and the Aztec and Mayan civilizations; Australia was in an intermediate position; while Tasmania had seen little technological development and the population of Flinders island had died out completely”
• Today 5% of US GDP is dedicated to research , it can go up but it has obvious limits (both population growth and percentage of those working in R&D
• Africa can get us another century of growth and innovation (Read more here)
• “If we get to AGI before we stagnate, then long term stagnation is not an issue “ and instead the focus should be on values lock in
• Human cloning values lock in: It is not socially accepted (yet?) to clone a bunch of Einsteins to lead research and advancement
• English productivity growth from 1250-1600 was literally zero but stagnation didn’t last (Islamic Golden Age research spilled into Europe to kickstart the Enlightenment): we need a diversity of societies
• One theory for why the Islamic Golden Age (8th-13th century) ended was that anti-scientific religious ideology set in to entrench power
• “At 1.5 children per women (roughly the average in Europe), within 500 years, the world population would fall from 10 billion to below 100 million; at one child per woman (roughly the fertility rate in South Korea), the world population would be under 100 million within 200 years.”
• Benjamin Friedman: we make moral progress when our countries grow richer; we get worse when we stagnate
• Derek Parfit’s (1942-2017) introduction of population ethics: “What now matters most is that we avoid ending human history.” Which is better: Higher average well-being or higher total well-being? He popularized this in his influential 1984 book Reasons and Persons
• Jan Narveson: “We are in favor of making people happy, but neutral about making happy people”
• John Broome: “intuition of neutrality” (if you could support a happy child but are not doing so then you’re neutral to creating happy people) (Intuition of neutrality seems easily answered by the point that we make decisions for ourselves not for the future children)
• Population ethics creates an alternative to the “intuition of neutrality,” which Parfit called Theory X: average view (disputed); total view (Parfit called it the Repugnant Conclusion), and Critical Level View (that more should be brought into the world above some level of well-being)
• Author basically argues for a low but Critical Level View, though he acknowledges uncertainty
• “If future people will be sufficiently well off then a civilization that is twice as long or twice as large is twice as good.” 189
• “Non welfarist goods “
• What is well being? Preference satisfaction view (wishes met even if not were) hedonistic (conscious experiences) and objective
• Life satisfaction surveys: only 47% between 2005 and 2015 had average scores over 5 on a scale of 1-10; a small study in UK put the “neutral” mark between 1-2 and so between 5-10% of people in the world would have existence worse than positive
• World Values Survey: more than half of the world is happy; 91% of Americans
• “Experience sampling” is a new way to get at happiness
• Across all these versions and others, no more than 10% of people today believe their life is not worth living
• Easterlin paradox: it is relative income within a country (rather than absolute wealth) that makes people happy. This remains true, though his 1974 argument has weakened because we do know now that as countries get richer they do get happier (but relative wealth decides how happy you are)
• Life satisfaction and GDP per capita do correlate
• Lottery winners are happier despite media narratives (though many regress back to a mean of happiness as other research shows)
• In the US, the black-white happiness divide has shrunk but persists; male and female happiness divide has shrunk too but because women’s level of happiness has declined
• In 2015, 18k people surveyed only 10% thought world getting better
• Hunter gatherers had shorter working days than industrial-era workers but “around half of children, born in pre-agricultural, hunter gatherer societies, died before the age of 15, compared to one and 200 in Europe today. If a hunter gatherers made it to 15, they could expect to live until 53, whereas the average Brit, who makes it to 15 today can expect to live until 89.”
• Number of neurons by animal is a way to compare pain: beetle 50k, honeybee 1m, 200m chickens, 500m for dogs, 80b for humans : by neuron counts, humans outweigh all farmed animals 30 to 1
• But wild fish outweigh humans by a factor of 17
• Aymara is a rare language exception where future metaphor is in back and past is in front: though we travel forward, we also know what is in front of us
• “When thinking about how to improve the world, the first step is to decide which problem to work on” 227
• We don’t do enough prioritizing: reducing plastic bags in rich counties has little impact but vegetarianism has more and direct cash donation to effective climate change research nonprofits (Giving What You Can) does even more
• Donations are often more effective than personal consumption changes: you are not optimized for harm; but they are optimized for good
• To change the world other than donations, he recommends: political activism (voting), spreading good ideas and having children
• 80,000 hours in your career: think critically about your career by seeking upside and capping downside
• Moral philosophy: consequentialists believe the ends justify the means, vs nonconsequentialists