Longtermism Considered

Quo vadis

Longtermism is the view that positively influencing the long-term future of humanity (and potentially other sentient beings) is a key moral priority—sometimes argued to be the key priority. It is closely tied to existential risk (x-risk) reduction: efforts to prevent events that could cause humanity’s premature extinction or permanently curtail its potential.

This framework emerged prominently within the Effective Altruism movement but draws on broader philosophical traditions. It represents one extension of the impartial, evidence-based, aggregate-welfare approach we discussed in the context of radical utilitarianism.

Core Ideas and Definitions

Philosopher William MacAskill defines longtermism as “the view that positively influencing the long-term future is a key moral priority of our time.” He distinguishes strong longtermism as the stronger claim that it is the key moral priority.

Toby Ord (in The Precipice, 2020) frames it as a “moral re-orientation toward the vast future that existential risks threaten to foreclose.” The central insight is that:

•  Future people matter morally as much as present ones (impartiality across time).

•  The future could be enormous: if humanity survives as long as a typical mammalian species (~1 million years) at current population levels, future people could outnumber all who have ever lived by thousands to one.

•  We live in a uniquely pivotal era (“the precipice”) where technological power has surged while wisdom and governance have not kept pace—making our actions today capable of locking in trajectories for millennia or longer.

Longtermism emphasizes tractability (some future-shaping actions are feasible now) and importance (the scale of potential impact dwarfs many near-term interventions when future lives are counted).

Existential Risks: Categories and Estimates

Existential risks are those that threaten the extinction of humanity or the permanent loss of its potential (e.g., through collapse into a state from which recovery is impossible or vastly diminished).

Toby Ord’s estimates (from The Precipice, with later reflections) for the chance of existential catastrophe in the next 100 years (his “best guesses,” acknowledging wide uncertainty):

•  Unaligned artificial general intelligence (AGI): ~1 in 10 (highest single risk in his assessment).

•  Engineered pandemics: ~1 in 30.

•  Nuclear war: ~1 in 1,000.

•  Climate change (extreme scenarios): ~1 in 1,000.

•  Natural risks (asteroids, supervolcanoes, stellar explosions, etc.): Much lower, on the order of 1 in 10,000 or less combined.

•  Overall: Roughly 1 in 6 (~16–17%) chance of existential catastrophe this century.

Natural risks have been low throughout human history (we’ve survived ~2,000 centuries). Anthropogenic risks dominate today because our technological capabilities (nuclear weapons, synthetic biology, advanced AI) now exceed our collective ability to manage them safely. Ord updated some numbers in 2024 reflections (climate risk somewhat lower, nuclear somewhat higher, AI/pandemics mixed), but the overall picture and order of magnitude remain similar.

Other categories sometimes discussed include unrecoverable dystopias, permanent value lock-in (e.g., via advanced AI or totalitarian systems), and risks from space expansion or other emerging technologies.

Arguments in Favor

Proponents argue:

•  Scale and impartiality: If future lives have moral weight comparable to present ones, even small reductions in x-risk can have astronomical expected value.

•  Leverage: We are early in history and can influence foundational technologies and institutions (e.g., AI alignment research, biosecurity governance, nuclear policy) at relatively low cost compared to the stakes.

•  Empirical grounding: Historical near-misses (e.g., Cuban Missile Crisis) and current technological trends show real vulnerabilities.

•  Practical examples: Work on AI safety, pandemic preparedness, and robust governance can have both near-term and long-term benefits.

Books like Ord’s The Precipice and MacAskill’s What We Owe the Future (2022) lay out these cases with data and thought experiments.

Criticisms and Counterarguments

Longtermism is contested, even within Effective Altruism circles, and has drawn sharp external critique:

•  Philosophical concerns: It often relies on total utilitarianism or population ethics, which can lead to counterintuitive implications (e.g., the “repugnant conclusion” where vast numbers of lives barely worth living are preferred over fewer high-quality ones). Critics argue it risks “fanaticism” by prioritizing tiny probabilities of enormous future payoffs over certain present suffering.

•  Uncertainty and predictability: Effects on the far future are extremely hard to forecast reliably. Overconfidence in speculative scenarios (especially around AGI) is a common charge.

•  Neglect of the present: Prioritizing hypothetical future trillions can deflect resources and attention from urgent current issues like global poverty, animal suffering, or injustice. Some see this as morally myopic or even “toxic” when it justifies deprioritizing today’s harms.

•  Social and political risks: Associations with elitism, transhumanism, or powerful tech figures have led to accusations that it can rationalize power concentration or questionable means. The FTX scandal (involving major EA/longtermist funder Sam Bankman-Fried) amplified concerns about ends-justifying-means thinking.

•  Empirical pushback: Some argue the 1-in-6 figure or specific AI risk estimates are overstated or under-supported by evidence.

Defenders respond that these critiques often mischaracterize the view (longtermism is compatible with strong near-term action) and that ignoring tail risks with civilization-scale consequences would itself be irresponsible.

Current Landscape (as of 2026)

Longtermism remains influential in parts of Effective Altruism, AI safety research, and certain philanthropic circles, with organizations working on biosecurity, AI governance, and related areas. However, post-FTX (2022 onward), the broader EA ecosystem has diversified, with greater emphasis on near-term causes, transparency, and avoiding over-reliance on any single framework.

Research continues at institutions connected to Oxford and elsewhere, alongside ongoing philosophical debate. The ideas have entered wider public discourse (e.g., via books, articles, and policy discussions on AI and pandemics), but they face sustained scrutiny regarding both their foundations and practical implications.

Why It Matters

Longtermism and existential risk work highlight a profound asymmetry: humanity’s technological power has grown explosively, while our collective wisdom and institutions have not kept pace. Whether one fully endorses strong longtermism or not, the underlying questions—How should we weigh the interests of future generations? What are the biggest threats to civilization’s continuation? How can we steer toward better long-term outcomes?—are among the most consequential we face.

These ideas connect directly to the utilitarian and impartial frameworks explored earlier: extending concern across time and (in broader versions) across all sentient life. They are not settled doctrine but an active area of inquiry with significant real-world influence and vigorous internal and external debate.

For deeper engagement, key starting points include Toby Ord’s The Precipice, William MacAskill’s What We Owe the Future, and critical responses in philosophy journals and forums. The field continues to evolve with new evidence on risks and governance.