What We Owe the Future
William MacAskill
Fundamental Books, $32 (fabric)
“House is massive,” wrote Douglas Adams in The Hitchhiker’s Information to the Galaxy (1979). “You simply gained’t imagine how vastly, massively, mind-bogglingly massive it’s. I imply, you might suppose it’s a good distance down the highway to the chemist’s, however that’s simply peanuts to area.”
What we do now impacts future folks in dramatic methods—above all, whether or not they’ll exist in any respect.
Time is massive, too—even when we simply suppose on the timescale of a species. We’ve been round for about 300,000 years. There are actually about 8 billion of us, roughly 15 % of all people who’ve ever lived. You might suppose that’s lots, nevertheless it’s simply peanuts to the long run. If we survive for one more million years—the longevity of a typical mammalian species—at even a tenth of our present inhabitants, there can be 8 trillion extra of us. We’ll be outnumbered by future folks on the size of a thousand to 1.
What we do now impacts these future folks in dramatic methods: whether or not they’ll exist in any respect and in what numbers; what values they embrace; what kind of planet they inherit; what types of lives they lead. It’s as if we’re trapped on a tiny island whereas our actions decide the habitability of an enormous continent and the life prospects of the numerous who might, or might not, inhabit it. What an terrible duty.
That is the attitude of the “longtermist,” for whom the historical past of human life to this point stands to the way forward for humanity as a visit to the chemist’s stands to a mission to Mars.
Oxford philosophers William MacAskill and Toby Ord, each affiliated with the college’s Way forward for Humanity Institute, coined the phrase “longtermism” 5 years in the past. Their outlook attracts on utilitarian excited about morality. In response to utilitarianism—an ethical idea developed by Jeremy Bentham and John Stuart Mill within the nineteenth century—we’re morally required to maximise anticipated mixture well-being, including factors for each second of happiness, subtracting factors for struggling, and discounting for chance. While you do that, you discover that tiny probabilities of extinction swamp the ethical arithmetic. In the event you may save 1,000,000 lives immediately or shave 0.0001 % off the chance of untimely human extinction—a one in 1,000,000 probability of saving not less than 8 trillion lives—it is best to do the latter, permitting 1,000,000 folks to die.
Now, as many have famous since its origin, utilitarianism is a radically counterintuitive ethical view. It tells us that we can not give extra weight to our personal pursuits or the pursuits of these we love than the pursuits of excellent strangers. We should sacrifice all the things for the larger good. Worse, it tells us that we should always accomplish that by any efficient means: if we will shave 0.0001 % off the chance of human extinction by killing 1,000,000 folks, we should always—as long as there aren’t any different adversarial results.
However even should you suppose we’re allowed to prioritize ourselves and people we love, and never allowed to violate the rights of some with a view to assist others, shouldn’t you continue to care in regards to the destiny of strangers, even those that don’t but exist? The ethical arithmetic of mixture well-being is probably not the entire of ethics, however isn’t it a significant half? It belongs to the area of morality we name “altruism” or “charity.” Once we ask what we should always do to learn others, we will’t ignore the disquieting incontrovertible fact that the others who occupy the long run might vastly outnumber those that occupy the current, and that their very existence depends upon us.
From this standpoint, it’s an pressing query how what we do immediately will have an effect on the additional future—pressing particularly in relation to what Nick Bostrom, the thinker who directs the Way forward for Humanity Institute, calls the “existential threat” of human extinction. That is the query MacAskill takes up in his new guide, What We Owe the Future, a densely researched however surprisingly mild learn that ranges from omnicidal pandemics to our new AI overlords with out ever changing into bleak.
Like Bostrom, MacAskill has an enormous viewers—unusually massive for a tutorial thinker. Invoice Gates has referred to as him “an information nerd after my very own coronary heart.” In 2009 he and Ord helped discovered Giving What We Can, a company that encourages folks to pledge not less than 10 % of their earnings to charitable causes. With our tithe, MacAskill holds, we needs to be utilitarian, aggregating advantages, subtracting harms, and weighing odds: our 10 % needs to be directed to the simplest charities, gauged by ruthless empirical measures. Thus the motion referred to as Efficient Altruism (EA), by which MacAskill is a number one determine. (Peter Singer is another.) By one estimate, about $46 billion is now committed to EA. The motion counts amongst its acolytes such distinguished billionaires as Peter Thiel, who gave a keynote address on the 2013 EA Summit, and cryptocurrency alternate pioneer Sam Bankman-Fried, who grew to become a convert as an undergraduate at MIT.
Efficient Altruists needn’t be utilitarians about morality (although some are). Theirs is a bounded altruism, one which respects the rights of others. However they’re inveterate quantifiers, and once they do the altruistic math, they’re led to longtermism and to the quietly radical arguments of MacAskill’s guide. “Future folks depend,” MacAskill writes:
There might be lots of them. We are able to make their lives go higher. That is the case for longtermism in a nutshell. The premises are easy, and I don’t suppose they’re significantly controversial. But taking them significantly quantities to an ethical revolution.
The premises are certainly easy. Most individuals involved with the consequences of local weather change would settle for them. But MacAskill pursues these premises to surprising ends. If the premises are true, he argues, we should always do what we will to make sure that “future civilization can be massive.”
MacAskill spends lots of effort and time asking how to learn future folks. What I’ll come again to is the ethical query whether or not they matter in the best way he thinks they do, and why. Because it seems, MacAskill’s ethical revolution rests on contentious, counterintuitive claims in “inhabitants ethics.”
In terms of the how, MacAskill is fascinating—if, at instances, alarming. Since having a optimistic affect on the long-term future is “a key ethical precedence of our time,” he writes, we have to estimate what affect our actions can have. It’s troublesome to foretell what’s going to occur over many 1000’s of years, after all, and MacAskill doesn’t strategy the duty alone: his guide, he tells us, depends on a decade of analysis, together with two years of fact-checking, in session with quite a few “area specialists.”
The long-term worth of working for a given end result is a perform of that end result’s significance (what MacAskill calls its “common worth added”), its persistence or longevity, and its contingency—the extent to which it depends upon us and wouldn’t occur anyway.
MacAskill’s ethical revolution rests on contentious claims in “inhabitants ethics.”
Among the many most important and chronic determinants of life for future generations, MacAskill argues, are the values we move on to them. And values are sometimes contingent. MacAskill takes as a case examine the abolition of slavery within the nineteenth century. Was it, he asks, “extra like the usage of electrical energy—a roughly inevitable growth as soon as the concept was there?” or “just like the sporting of neckties: a cultural contingency that grew to become almost common globally however which may fairly simply have been totally different?” Slavery had been abolished in Europe as soon as earlier than, within the late center ages, solely to return with a vengeance. Was it destined to say no once more? MacAskill cites historian Christopher Leslie Brown, writing in Ethical Capital (2006): “In key respects the British antislavery motion was a historic accident, a contingent occasion that simply as simply may by no means have occurred.” Values matter to the long-term future, and they’re topic to intentional change.
From right here we lurch to the alarming: MacAskill is fearful in regards to the growth of synthetic normal intelligence (AGI), able to performing as huge a variety of duties as we do, not less than as nicely or higher. He charges the probabilities of AGI arriving within the subsequent fifty years no decrease than one in ten. The chance is that, if AGI takes over the world, its creators’ imaginative and prescient could also be locked in for a really very long time. “If we don’t design our establishments to control this transition nicely—preserving a plurality of values and the opportunity of fascinating ethical progress,” MacAskill writes, “then a single set of values may emerge dominant.” The outcomes may be dystopian: What if the AGI that guidelines the world belongs to a fascist authorities or a rapacious trillionaire?
MacAskill requires higher regulation of AI analysis to protect area for reflection, open-mindedness, and political experimentation. Most of us wouldn’t object. However, as is usually the case in discussions of AI—and regardless of the salience of contingency—MacAskill tends to deal with the progress of know-how as a given. We are able to hope to control the transition to AGI nicely, however the transition is actually coming. What we do “may have an effect on what values are predominant when AGI is first constructed,” MacAskill notes—however not whether or not it’s constructed in any respect. Just like the thinker Annette Zimmermann, I hope that isn’t true.
MacAskill could also be reconciled to AGI, himself, by the hope that it’ll tackle one other long-term drawback: the specter of financial and technological stagnation. Once more, his argument is each fascinating and alarming. “For the primary 290,000 years of humanity’s existence,” MacAskill writes, “international development was near 0 % per 12 months; within the agricultural period that elevated to round 0.1 %, and it accelerated from there after the Industrial Revolution. It’s solely within the final hundred years that the world financial system has grown at a fee above 2 % per 12 months.” However it could actually’t go on perpetually: “if present development charges continued for simply ten millennia extra, there must be ten million trillion instances as a lot output as our present world produces for each atom that we may, in precept, entry”—that’s, for each atom inside ten thousand mild years of Earth. “Although after all we will’t be sure,” MacAskill drily concludes, “this simply doesn’t appear doable.”
There may be proof that technological course of has already slowed down exterior the areas of computation and AI. The speed of development in “whole issue productiveness”—our capability to get extra financial output from the identical enter—is declining, and in accordance with a latest examine by economists at Stanford and the London Faculty of Economics, new ideas are increasingly scarce. MacAskill illustrates this neatly, imagining how your life would change in fifty years. While you go from 1870 to 1920, you get working water, electrical energy, a phone, and maybe a automotive. While you go 1970 to 2020, you get a microwave oven and a much bigger TV. The one dramatic shifts are in computing and communications. With out the magic bullet of AGI, via which we’d construct limitless AI staff in R&D, MacAskill fears that we’re doomed to stagnation, maybe for lots of or 1000’s of years. However from his perspective, if it’s not everlasting, and folks don’t want they’d by no means been born, it’s not so dangerous. The worst issues about stagnation, for MacAskill, are the risks of misguided worth lock-in and of extinction or everlasting collapse.
Right here we come, ultimately, to existential dangers: asteroid collisions, which we’d not detect; deadly pandemics, which may be bioengineered; World Battle III, which could flip nuclear; and local weather change, which could speed up via suggestions within the local weather system. MacAskill applauds NASA’s Spaceguard program, requires higher pandemic preparedness and biotech security (specialists, he notes, “put the chance of an extinction-level engineered pandemic this century at round 1 %”), and helps a speedy shift to inexperienced vitality.
For MacAskill, survival in the long run counts for excess of a lower of struggling and loss of life within the close to future.
However what’s most alarming in his strategy is how little he’s alarmed. As of 2022, the Bulletin of Atomic Scientists set the Doomsday Clock, which measures our proximity to doom, at 100 seconds to midnight, the closest it’s ever been. In response to a examine commissioned by MacAskill, nonetheless, even within the worst-case situation—a nuclear battle that kills 99 % of us—society would doubtless survive. The longer term trillions can be secure. The identical goes for local weather change. MacAskill is upbeat about our probabilities of surviving seven levels of warming or worse: “even with fifteen levels of warming,” he contends, “the warmth wouldn’t move deadly limits for crops in most areas.”
That is stunning in two methods. First, as a result of it conflicts with credible claims one reads elsewhere. The final time the temperature was six diploma larger than preindustrial ranges was 251 million years in the past, within the Permian-Triassic Extinction, essentially the most devastating of the 5 nice extinctions. Deserts reached nearly to the Arctic and greater than 90 % of species had been worn out. In response to environmental journalist Mark Lynas, who synthesized present analysis in Our Closing Warning: Six Levels of Local weather Emergency (2020), at six levels of warming the oceans will grow to be anoxic, killing most marine life, and so they’ll start to launch methane hydrate, which is flammable at concentrations of 5 %, making a threat of roving firestorms. It’s not clear how we may survive this hell, not to mention fifteen levels.
The second shock is how way more MacAskill values survival in the long run over a lower of struggling and loss of life within the close to future. That is the sharp finish of longtermism. Most of us agree that (1) world peace is best than (2) the loss of life of 99 % of the world’s inhabitants, which is best in flip than (3) human extinction. However how significantly better? The place many would see a larger hole between (1) and (2) than between (2) and (3), the longtermist disagrees. The hole between (1) and (2) is a brief lack of inhabitants from which we are going to (or not less than might) bounce again; the hole between (2) and (3) is “trillions upon trillions of people that would in any other case have been born.” That is the “perception” MacAskill credit to the long-lasting ethical thinker Derek Parfit. It’s the moral crux of essentially the most alarming claims in MacAskill’s guide. And there’s no approach to consider it with out dipping our toes into the deep, darkish waters of inhabitants ethics.
Inhabitants ethicists ask how good the world can be with a given inhabitants distribution, specified by the variety of folks present at numerous ranges of lifetime well-being all through area and time. Ought to we measure by whole mixture well-being? By common? Ought to we care about distribution, ranking inequitable outcomes worse? As MacAskill writes, inhabitants ethics is “one of the advanced areas of ethical philosophy . . . usually studied solely on the graduate degree. To my data, these concepts haven’t been offered to a normal viewers earlier than.” However he offers it his greatest, and with trepidation, I’ll observe go well with.
On the coronary heart of the controversy is what MacAskill calls “the instinct of neutrality,” elegantly expressed by ethical thinker Jan Narveson in a much-cited slogan: “We’re in favour of constructing folks glad, however impartial about making glad folks.” The attraction of the slogan is obvious at scales each giant and small. Suppose you might be advised that humanity will go extinct in a thousand years but additionally that everybody who lives can have a adequate life. Do you have to care if the common inhabitants every year is nearer to 1 billion or 2? Neutrality says no. What issues is high quality, not amount.
Now suppose you might be deciding whether or not to have a toddler, and also you count on that your baby would have a adequate life. Should you conclude that it could be higher to have a toddler than not, except you may level to some countervailing purpose? Once more, neutrality says no. In itself, including an additional life to the world isn’t any higher (or worse) than not doing so. It’s totally as much as you. It doesn’t observe that you simply shouldn’t care in regards to the well-being of your potential baby. As a substitute, there’s an asymmetry: though it’s not higher to have a contented baby than no baby in any respect, it’s worse to have a toddler whose life is just not value dwelling.
Longtermists deny neutrality: they argue that it’s all the time higher, different issues equal, if one other individual exists, offered their life is nice sufficient. That’s why human extinction looms so giant. A world by which we’ve trillions of descendants dwelling adequate lives is best than a world by which humanity goes extinct in a thousand years—higher by an enormous, enormous, mind-boggling margin. An opportunity to scale back the chance of human extinction by 0.01 %, say, is an opportunity to make the world an inconceivably higher place. It’s a larger contribution to the nice, by a number of orders of magnitude, than saving 1,000,000 lives immediately.
What’s most alarming in MacAskill’s strategy is how little he’s alarmed.
But when neutrality is correct, the longtermist’s arithmetic relaxation on a mistake: the additional lives don’t make the world a greater place, all by themselves. Our moral equations will not be swamped by small dangers of extinction. And whereas we could also be doing a lot lower than we should always to handle the chance of a deadly pandemic, worth lock-in, or nuclear battle, the reality is way nearer to widespread sense than MacAskill would have us imagine. We must always care about making the lives of those that will exist higher, or in regards to the destiny of those that can be worse off, not in regards to the variety of good lives there can be. In response to MacAskill, the “sensible upshot” of longtermism “is an ethical case for area settlement,” by which we may enhance the long run inhabitants by trillions. If we settle for neutrality, in contrast, we can be glad if we will make issues work on Earth.
An terrible lot activates the instinct of neutrality, then. MacAskill offers a number of arguments in opposition to it. One is in regards to the ethics of procreation. In case you are considering of getting a toddler, however you’ve a vitamin deficiency which means any baby you conceive now can have a well being situation—say, recurrent migraines—it is best to take nutritional vitamins to resolve the deficiency earlier than you attempt to get pregnant. However then, MacAskill argues, “having a toddler can’t be a impartial matter.” The steps of his argument, a reductio advert absurdum, bear spelling out. Evaluate having no baby with having a toddler who has migraines, however whose life remains to be value dwelling. “In response to the instinct of neutrality,” MacAskill writes, “the world is equally good both method.” The identical is true if we examine having no baby with ready to get pregnant with a view to have a toddler who’s migraine-free. From this it follows, MacAskill claims, that having a toddler with recurrent migraines is nearly as good an end result as having a toddler with out. That’s absurd. In an effort to keep away from this consequence, MacAskill concludes, we should reject neutrality.
However the argument is flawed. Neutrality says that having a toddler with a adequate life is on a par with staying childless, not that the result by which you’ve a toddler is equally good no matter their well-being. Think about a frivolous analogy: being a thinker is on a par with being a poet—neither is strictly higher or worse—nevertheless it doesn’t observe that being a thinker is equally good, whatever the pay.
A putting truth about circumstances just like the one MacAskill cites is that they’re topic to a retrospective shift. In case you are planning to have a toddler, it is best to wait till your vitamin deficiency is resolved. However should you don’t wait and also you give delivery to a toddler, migraines and all, it is best to love them and affirm their existence—not want you had waited, in order that they’d by no means been born. This shift explains what’s flawed with a second argument MacAskill makes in opposition to neutrality. Considering of his nephew and two nieces, MacAskill is inclined to say that the world is “not less than a little bit higher” for his or her existence. “In that case,” he argues, “the instinct of neutrality is flawed.” However once more, the argument is flawed. As soon as somebody is born, it is best to welcome their existence as an excellent factor. It doesn’t observe that it is best to have seen their coming to exist as an enchancment on the earth earlier than they got here into existence. Neutrality survives intact.
In rejecting neutrality, MacAskill leans towards the “whole view” on which one inhabitants distribution is best than one other if it has larger mixture well-being. That is, in impact, a utilitarian strategy to inhabitants ethics. The whole view says that it’s all the time higher so as to add an additional life, if the life is nice sufficient. It thus helps the longtermist view of existential dangers. But it surely additionally implies what is called the Repugnant Conclusion: which you can make the world a greater place by doubling the inhabitants whereas making folks’s lives a little bit worse, a sequence of “enhancements” that ends with an inconceivably huge inhabitants whose lives are solely simply value dwelling. Adequate numbers make up for decrease common well-being, as long as the extent of well-being stays optimistic.
Many regard the Repugnant Conclusion as a refutation of the full view. MacAskill doesn’t. “In what was an uncommon transfer in philosophy,” he experiences, “a public assertion was not too long ago revealed, cosigned by twenty-nine philosophers, stating that the truth that a idea of inhabitants ethics entails the Repugnant Conclusion shouldn’t be a decisive purpose to reject that idea. I used to be one of many cosignatories.” However you may’t outvote an objection. Think about the worst life one may stay with out wishing one had by no means been born. Now think about the form of life you dream of dwelling. For individuals who embrace the Repugnant Conclusion, a future by which trillions of us colonize planets in order to stay the primary type of life is best than a future by which we survive on Earth in modest numbers, attaining the second.
Many regard the Repugnant Conclusion as a refutation of his view of inhabitants ethics. MacAskill doesn’t.
MacAskill has a closing argument, drawing on work by Parfit and by economist-philosopher John Broome. “Although the Repugnant Conclusion is unintuitive,” he concedes, “it seems that it follows from three different premises that I’d regard as near indeniable.” The main points are technical, however the upshot is a paradox: the premises of the argument appear true, however the conclusion doesn’t. Because it occurs, I’m not satisfied that the premises are compelling as soon as we distinguish those that exist already from those that might or might not come into existence, as we did with MacAskill’s nephew and nieces. However the principle factor to say is that basing one’s moral outlook on the conclusion of a paradox is dangerous type. It’s a bit like concluding from the paradox of the heap—including only one grain of sand is just not sufficient to show a non-heap right into a heap; so, irrespective of what number of grains we add, we will by no means make a heap of sand—that there aren’t any heaps of sand. This can be a far cry from MacAskill’s “easy” start line.
Nor does MacAskill cease right here; he goes nicely past the Repugnant Conclusion. Because it’s not simply human well-being that counts, for him, he’s open to the view that human extinction wouldn’t be so dangerous if we had been changed by one other clever species, or a civilization of acutely aware AIs. What issues to the longtermist is mixture well-being, not the survival of humanity.
Nonhuman animals depend, too. Although their capability for well-being varies broadly, “we may, as a really tough heuristic, weight animals’ pursuits by the variety of neurons they’ve.” Once we do that, “we get the conclusion that our total views needs to be nearly totally pushed by our views on fish.” By MacAskill’s estimate, we people have fewer than a billion trillion neurons altogether, whereas wild fish have three billion trillion. Within the whole view, they matter thrice as a lot as we do.
Don’t fear, although. We shouldn’t put their lives earlier than our personal, since there may be purpose to imagine their lives are horrible. “If we assess the lives of untamed animals as being worse than nothing on common, which I feel is believable (although unsure),” MacAskill writes, “we arrive on the dizzying conclusion that from the attitude of the wild animals themselves, the big development and growth of Homo sapiens has been an excellent factor.” That’s as a result of human development and growth are sparing them from all that distress. From this attitude, the anoxic oceans of six-degree warming come as a merciful launch.
In Plato’s Republic, potential philosopher-kings start their training in dialectic or summary reasoning at age thirty, after years of gymnastics, music, and math. At thirty-five, they’re assigned jobs within the administration of town, like minor civil servants. Solely on the age of fifty do they flip to the Good itself, leaving the cave of political life for the daylight of philosophy, a present they repay by deigning to rule.
MacAskill is simply thirty-five. However like a philosopher-king, he follows the trail of dialectic from the shadows of conference to the blazing glare of a brand new ethical imaginative and prescient, returning to the cave to inform us a few of what he’s discovered. MacAskill calls the early Quaker abolitionist Benjamin Lay “an ethical entrepreneur: somebody who thought deeply about morality, took it very significantly, was completely prepared to behave in accordance together with his convictions, and was thought to be an eccentric, a weirdo, for that purpose.” In response to MacAskill: “We must always aspire to be weirdos like him.”
MacAskill kinds himself as an ethical entrepreneur too. His purpose is to construct a social motion, to win converts to longtermism. In any case, if you wish to make the long-term future higher, and our values are among the many most important, persistent, contingent determinants of the way it will go, there’s a longtermist case for working laborious to make lots of us longtermists. For all I do know, MacAskill might succeed on this. However as I’ve argued, the reality of his ethical outlook—which rejects neutrality and offers no particular weight to human beings—is lots much less clear than the injustice of slavery.
To his credit score, MacAskill admits room for doubt, conceding that he could also be flawed in regards to the whole view in inhabitants ethics. However he additionally has a view about what to do whenever you’re undecided of the ethical reality: assign a chance to the reality of every ethical view, “then take the motion that’s the greatest compromise between these views—the motion with the very best anticipated worth.” This raises issues of each idea and apply.
In apply, there’s a risk that longtermist considering will dominate anticipated worth calculations in the identical method as tiny dangers of human extinction. If there may be even a 1 % probability of longtermism being true, and it tells us that lowering existential dangers is many orders of magnitude extra necessary than saving lives now, these numbers might swamp the prescriptions of extra modest ethical visions.
Longtermism’s ethical arithmetic is simply nearly as good as its axioms.
The theoretical drawback is that we must be unsure about this manner of dealing with ethical uncertainty. What ought to we do when uncertainty goes all the best way down? Sooner or later, we fall again on ethical judgment and face what philosophers have referred to as the issue of “moral luck.” What we must do, no matter our beliefs, is to behave in accordance with the ethical reality of find out how to act with these beliefs. There’s no approach to insure ourselves in opposition to ethical error—to ensure that, whereas we might have made errors, not less than we acted as we should always, given what we believed. For we could also be flawed about that, too.
There are profound divisions right here, not simply in regards to the content material our ethical obligations however in regards to the nature of morality itself. For MacAskill, morality is a matter of indifferent, impersonal theorizing in regards to the good. For others, it’s a matter of rules by which we may moderately agree to control ourselves. For nonetheless others, it’s an expression of human nature. On the finish of his guide, MacAskill features a postscript, titled “Afterwards.” It’s a fictionalized model of how the long run may go nicely, from the attitude of longtermism. After planning to colonize area, MacAskill’s utopians pause to consider how.
There adopted a interval of in depth dialogue, debate, and commerce that grew to become identified, merely, because the Reflection. Everybody tried to determine for themselves what was really invaluable and what a really perfect society would seem like. Progress was sooner than anticipated. It turned out that ethical philosophy was not intrinsically laborious; it’s simply that human brains are ill-suited to sort out it. For specifically educated AIs, it was baby’s play.
I feel this imaginative and prescient is misguided, and never simply because there are serious arguments in opposition to area colonization. Ethical judgment is one factor; machine studying is one other. No matter is flawed with utilitarians who advocate the homicide of 1,000,000 for a 0.0001 % discount within the threat of human extinction, it isn’t an absence of computational energy. Morality isn’t made by us—we will’t simply resolve on the ethical reality—nevertheless it’s made for us: it rests on our widespread humanity, which AI can not share.
What We Owe the Future is an instructive, clever guide. It has lots to show us about historical past and the long run, about uncared for dangers and ethical myopia. However an ethical arithmetic is simply nearly as good as its axioms. I hope readers strategy longtermism with the open-mindedness and ethical judgment MacAskill needs us to protect, and that its values are explored with out ever being locked in.
