This edited volume assembles thirty essays examining whether our decisions today should be guided (primarily) by their effects thousands or millions of years hence. Longtermism is a relatively new branch of effective altruism, the popularity of which is largely thanks to the recent publication of The Precipice (Bloomsbury 2020) and What We Owe the Future (Oneworld 2022). The editors’ approach to defining longtermism in the introduction reflects the nascent nature of the field. They present longtermism as a “cluster of ideas” holding that considerations of the distant future are highly significant for decision-making today. The editors identify four key elements: that the far future matters morally; that present actions can meaningfully affect the far future’s balance of happiness over suffering and the number of future welfare subjects; that we have sufficiently good epistemic access to our potential far-future effects; and that our relationship to these possibilities grounds strong reasons, perhaps even a moral imperative, to act in ways that positively steer the far future. However, the editors simultaneously acknowledge that contributors define “longtermism” and “longtermist” in different ways, all loosely related to the general idea that “positively influencing the long-term future is a key moral priority of our time.” Rather than seeking to establish canonical definitions, they explicitly embrace this flexibility, noting that, “in the present state of the discussion, this flexibility is helpful.”
The first section attempts to establish the theoretical foundations for taking the far future seriously in moral deliberation. The opening chapter presents the standard argument for “strong longtermism” – the claim that, in important decision situations, every near-best option primarily benefits the far future rather than the present. This argument relies heavily on cost-effectiveness calculations for existential risk mitigation, assuming that preventing human extinction generates enormous expected utility by preserving vast future populations. The subsequent chapters in this section largely revolve around population ethics, examining whether longtermist conclusions survive various theoretical modifications. One considers “neutrality” views that adding lives with positive welfare is neither good nor bad; another argues for “prudential longtermism” based on radical life extension possibilities; a third questions whether human continuation is actually valuable given environmental destruction and animal suffering. These contributions take the form of technical exercises in moral philosophy, examining different axiological frameworks to determine which combinations yield longtermist conclusions. The section’s final chapters introduce contractualist and deontological objections, arguing that non-consequentialist moral theories resist longtermist demands – though these arguments operate within different theoretical frameworks from the utilitarian approach that dominates the rest of the volume.
The second section confronts what should be longtermism’s central challenge: our apparent inability to reliably forecast or control long-term outcomes. One chapter examines empirical evidence about forecast accuracy, finding that even short-term predictions often fail and accuracy decays over time. Another draws on paleobiology to argue that evolutionary history shows no predictable patterns that would support confident claims about million-year timescales. These chapters raise important questions about whether longtermist interventions can be meaningfully guided by our predictive capabilities. The remaining contributions in this section attempt various responses: one proposes a “stewardship” model focused on preserving decision-making capacity for future generations; another develops a mathematical framework for classifying different types of trajectory changes; a third considers lessons from cultural evolution. However, these responses leave the basic epistemological problem largely unresolved. If we cannot predict the far future, and if our interventions’ effects decay or become chaotic over time, the longtermist project faces significant foundational challenges.
The third section examines which interventions longtermist thinking supposedly recommends. Much attention focuses on existential risk mitigation, particularly from artificial intelligence. One chapter defends “urgent longtermism” against “patient longtermism,” arguing we live at a uniquely important historical moment requiring immediate action on technological risks. Others present detailed scenarios about artificial intelligence systems seeking power and potentially disempowering humanity. These chapters read more like science fiction speculation than rigorous risk assessment, relying on chains of uncertain assumptions about future technological development, artificial behaviour, and human responses. Other contributions consider whether longtermist conclusions are as radical as they initially appear. One argues that existential risk mitigation can be justified through conventional cost-benefit analysis without appealing to vast future populations. Another contends that even agents who don’t discount future welfare should often behave “myopically” due to causal and epistemic limitations. The section includes chapters on population dynamics, life extension, and animal welfare. While these topics are relevant to long-term thinking, their treatment often feels disconnected from the broader longtermist framework, suggesting a field still searching for coherent priorities.
The volume’s fourth section shifts attention from individual moral obligations to questions of institutional design and social psychology. One chapter explores “institutional longtermism,” arguing that when applied to institutions rather than individual actions, longtermism may sidestep various standard objections and find support from political values like justice and democracy. Another proposes a “futures assembly” with retrospective accountability mechanisms to combat political short-termism, while a third argues that longtermist social planners should actually be more willing to accept certain risks when evaluated against expected future risks. One contribution diagnoses “hardness bias” in academic economics as preventing economists from engaging with longtermist research, calling for methodological pluralism. The final two chapters present empirical findings: one reports survey evidence suggesting that both legal experts and laypeople endorse principles of “legal longtermism” more than typically assumed, while another finds that temporal partiality and nationalist ingroup favouritism are psychologically correlated, with people showing less nationalism when considering distant future generations.
Several problems emerge from this survey of contemporary longtermist thinking. Most fundamentally, the volume reveals a striking disconnect between longtermism’s theoretical ambitions and its practical limitations. The theoretical chapters assume we can make meaningful comparisons between present and far-future welfare, aggregate expected utilities across vast timescales, and identify interventions with persistent positive effects. Yet the empirical chapters often undermine these assumptions, showing that long-term prediction is unreliable and that complex systems exhibit chaotic rather than controllable behaviour. This tension isn’t resolved in the book, leaving readers unclear what they’re supposed to take away from the book.
The contributors’ approach to population ethics reveals other methodological problems. Multiple chapters examine whether longtermist conclusions survive modifications to utilitarian theory, testing robustness across different axiological frameworks. Yet this exercise presupposes that moral philosophy proceeds by selecting the theory that best fits our intuitions, then deriving practical conclusions. Such an approach treats moral philosophy as a purely academic exercise divorced from the practical questions it purports to address. The persistent focus on thought experiments involving vast future populations suggests a field more interested in theoretical puzzles than in understanding actual moral obligations.
The collection’s treatment of existential risk exemplifies longtermism’s tendency toward speculative reasoning disguised as rigorous analysis. The chapters on artificial intelligence risk present elaborate scenarios about future intelligence systems developing goals, seeking power, and potentially eliminating humanity. While such possibilities cannot be dismissed entirely, the confidence with which these authors assign probabilities to complex chains of future events appears unjustified. The chapters read more like exercises in creative writing than serious risk assessment, yet they form the empirical backbone of many longtermist policy recommendations. Rather than grappling seriously with concrete, well-understood problems like global poverty, disease, malnutrition, and inequality – issues where we have both clear causal understanding and proven interventions – the contributors gravitate toward speculative scenarios involving artificial intelligence takeover and other futuristic threats. This represents a peculiar moral prioritisation that privileges dramatic, science-fiction-like scenarios over the mundane reality of present suffering. The techno-fanatical focus appears to reflect the intellectual preferences of a particular academic subculture rather than a serious attempt to identify the most pressing moral problems. This has been my criticism of the effective altruist movement for many years now, and might be one of the many underlying reasons for the recent closure of the Global Priorities Institute. When longtermism cannot adequately grapple with how present-day atrocities shape the future, it forfeits any serious claim to moral priority.
The book ultimately reads as an academic exercise in moral philosophy rather than a compelling case for reorienting human priorities around the far future. While individual chapters contain interesting arguments, the collection fails to make a persuasive case that longtermism represents a coherent or compelling approach to moral reasoning. Readers seeking guidance on humanity’s long-term future would be better served by more modest approaches that acknowledge the limits of our predictive abilities and focus on building robust institutions rather than optimising for speculative scenarios.
