Will MacAskill’s new book, What We Owe The Future has just been released in the US and will be available in the UK from September 1. You might already be turning to book reviews or podcasts to inform whether you should buy a copy. To help, I’m writing a quick summary of the book, the three new insights I gained, and the three questions it left me asking.
But it’s worth being upfront to the reader about where I sit. MacAskill entwines rigorous arguments with compelling metaphors to promote a profoundly important idea: we can make the future go better, and we should. It’s filled with rich, relevant and persuasive historical examples, grounding his philosophical arguments in the real world. It’s a book for people who are curious to learn, but also motivated to act — I strongly recommend it.
The book makes a case for longtermism, the view that positively influencing the long-term future is a key moral priority of our time. The overarching argument is simple:
The book argues for the first claim at the outset in straightforward and intuitive terms, but MacAskill also takes the reader through rigorous arguments grappling with population ethics, the area of philosophy that focuses on these sorts of questions.
MacAskill’s argument for the second claim is that there are far more people who could potentially live in the future than have ever lived in the past. On certain assumptions about the average future population and expected lifespan of the human species, the number of people who could live in the future dramatically outweighs the number of people who have ever lived. This kind of analysis may have inspired Our World In Data’s visualisation of how vast the long-term future could be.
I recommend Kurzegast’s “The Last Human — A Glimpse Into The Far Future” which evocatively draws out the potential magnitude of our long-term future. A substantial amount is at stake: if the future goes well, it could be enormously valuable, but if it doesn’t, it could be
The third claim is the central focus of the book. MacAskill aims not just to argue that we can in principle influence the long-term future (which is the standard of most philosophical arguments) but that we can and here’s how (the standard for those who want to take action). MacAskill argues that one of the best ways to focus on the long-term future is to reduce our risk of extinction. Though he also argues that it’s not just about whether we survive; it’s also about how we survive. The case for focusing on the ways we can improve the quality of the long-term future is one of the key lessons I took from the book.
Much of what I read was new to me, even as someone who’s been highly engaged with these ideas. If I were to list all the historical examples that were new to me, I’d essentially be rewriting the book. Instead, here are the top three lessons I learned.
One of the book’s key ideas is that if we could re-run history again, it’s unlikely we’d end up with the same values — instead, they’re contingent. This is not something I believed before reading the book.
If I can make a personal confession: I’m a (perhaps naive) supporter of a philosophical view called hedonistic-utilitarianism, which claims the best actions are those that increase the total amount of experienced Before reading this book, I hadn’t thought much about whether the values prevalent today were historically contingent. Instead, somewhat smugly, I assumed that there’d been a more-or-less consistent trend over time toward my favourite view (I say smugly because I thought the explanation of this trend was simple: my favourite view is I now think I was wrong.
MacAskill develops a new conceptual framework to help address the question of whether our values are historically contingent, and he applies this framework to several historical case studies. The most fascinating — and horrifying — is his argument that the abolition of slavery was highly contingent. Before I read the book, I thought the abolition of slavery was inevitable. I reasoned that slavery was so obviously appalling that, eventually, humanity would have abolished it. I now think we’re extremely fortunate and indebted to all those people who managed to (by and large) end it. Without them, I’m now convinced that there’s a disturbingly high chance it would have continued to this day.
The contingency of our values has important implications for people motivated by longtermism.
One of the most straightforward ways we could ruin the long-term future of humanity is if we end it. That’s why longtermists generally focus on threats to our continued survival, such as those coming from misaligned artificial intelligence, pandemics, and nuclear war. Toby Ord’s extensively analysed these risks in The Precipice, and What We Owe The Future compliments this work with its coverage of these risks. I was generally quite familiar with these ideas, but what I was less familiar with — and what I see as MacAskill’s novel contribution — is how to think about other ways we can improve the long-term future.
MacAskill develops his idea that values have been historically contingent into the idea that civilisation’s future values may be determined by how we act today. And given how many people who might be guided by these values, getting them right matters. MacAskill’s conclusion is that we ought to build “a morally exploratory world” — I’ll walk through some of the steps he takes to arrive there, and say more about what that means.
Key to this argument is one of MacAskill’s sustained metaphors: the moral views that shape society are like molten glass. While hot, glass is highly malleable and can easily be blown into a variety of shapes. But it doesn’t stay hot for long, and once it cools, it sets. Just like glass, MacAskill suggests that societies often have periods of great plasticity, where the dominant moral view could easily change, but subsequently, a lock-in event occurs, causing whatever the dominant value system is to persist for an extremely long time — just like the glass sets once cooled. The Hundred Schools of Thought in China, and the tumultuous beginnings (but long-lasting prevalence) of many religious beliefs and practices, as examples of this trend.
MacAskill argues that we are in one of these unusually plastic times where the glass is molten. Yet, even if you think society is far more moral than it has ever been, we should be fearful of the glass setting:
It is extraordinarily unlikely that, of all generations across time, we are the first ones to have gotten it completely correct. The values you or I endorse are probably far from the best ones.
Given this, MacAskill suggests we build “a morally exploratory world” allowing us to make moral progress before any lock-in event occurs. MacAskill accompanies this with several suggestions, such as promoting conservation efforts, experimenting with different political structures like charter cities, and upholding certain forms of free speech that facilitate good-faith debate.
I found reading through these arguments exhilarating. How we can positively influence the long-term future is something I’ve been interested in for many years, but this part of the book provided me with conceptual frameworks and historical context that were completely new to me.
In 2018, when I read Nick Bostrom’s Astronomical Waste, I became extremely motivated to improve the long-term future. In this article, Bostrom argues that, even though speeding up how long it takes us to colonise the galaxy by a second could result in a further 10^29 lives, it’s dwarfed by the value of even tiny reductions in existential risk. The conclusion is simple: we should reduce existential risk.
I think this is an important argument, and sits among other arguments that lay the foundation for longtermism. But though Bostrom’s argument was compelling to me, it is not compelling to everyone. In the early days, this made me think that longtermism was an inherently niche idea, unlikely to be widely adopted — or even discussed — by the broader community. Though over the last few years I’ve been changing my mind on this issue, after reading What We Owe The Future, I am confident I was wrong.
Though the idea of astronomical waste supports longtermism as an idea, longtermism does not require it. MacAskill’s case for longtermism does not require that we believe we might colonise the galaxy, only that the future is big. It does not focus on reducing existential risk in the abstract, but instead points to specific work being done, and tangible actions that readers can take. This is not a slight to the kinds of arguments that initially convinced me of longtermism: MacAskill was able to write this book in this way by using the foundation early longtermists built, and pointing to the work they inspired.
Though I learned important lessons from the book, I also have some lingering questions. Here are three.
In one of the book’s later chapters, MacAskill claims that the recent trend of viewing the choice to have children as unethical (due to the carbon emissions the child will generate) is mistaken. He is very clear that this is a deeply personal choice, and he does not aim to pass judgement on others for their choice, nor suggest governments restrict people’s reproductive rights to decide whether and when to have a child. His aim is to present the case that having children is a way of positively contributing to the world.
Here’s a rough summary:
To state my bias, I want to have kids (and am likely going to even if I decide it’s unethical), so I’m tempted to listen to these arguments as music to my ears. However, though I find the final point about the value of the child’s own life convincing, I’m sceptical of the argument that in the past, population growth has been good (for humans), so an additional child now is likely good too.
Here are a few reasons why:
For what it’s worth, I do think (overall) that MacAskill’s conclusion is right, even though I’m not persuaded by some of his reasoning.
In an 80,000 Hours Podcast from 2020, MacAskill discussed the feedback he received after giving a series of lectures on longtermism. He reported the following as one of the most common objections:
many more people are willing to say future people just don’t matter than I would expect.
I’ve encountered the same objection in my own life when I’ve discussed these ideas with friends and family. In this book, MacAskill makes a case for caring about these future people in ways I found moving, and also — when he dives into the arguments — extremely convincing. But then, I was already convinced, so I am left to wonder if others will be too.
Though one of the key lessons I learned was that extending our efforts beyond just reducing our risk of extinction could be hugely impactful, I now wonder: if you aim to make the future better, what should you do?
I was a co-author of Giving What We Can’s strategic update in 2022, which highlighted promoting positive values as a key motivation behind the organisation’s mission. At the time, we acknowledged this was difficult to evaluate — how could we tell if we were doing a good job? I similarly wonder how else we might improve the long-term values of humanity. I mentioned some of MacAskill’s suggestions earlier (conservation, charter cities, free speech) but I find it difficult to evaluate just how important each is. I look forward to new work in this area, which I imagine is likely to come.
I found What We Owe The Future to be a fascinating read, and I enjoyed grappling with its new ideas and examples. I recommend it to those who are already quite familiar with longtermism, and I’ve already purchased several copies to give away to some of my friends and family who aren’t.
On a personal note, I’m someone who can often feel a little embarrassed if I feel I’m being perceived as overly ambitious. The project of longtermism has often felt awkward to me in this way; at times, it seemed too grandiose. The book’s historical grounding and emphasis of longtermism as both an idea and a movement, makes me feel much more comfortable with my commitment to positively influence the long-term future.
For those of you who have read the book, I’m interested in hearing about other people’s takeaways. For those that haven’t, and would like to know more about its contents before deciding to buy it, I’m happy to share my thoughts. You can comment on the Effective Altruism Forum version of this review, or send me an email at michael.townsend[at]givingwhatwecan.org.
Thank you to Luke Freeman and Vanessa Li for your helpful comments.