Strong Longtermism: Future generations matter 10^58 times more than present

Bentham's Bulldog on LessWrong presents a compelling case for Strong Longtermism: the idea that the most important consequences of our actions are those affecting the long-run future. The core argument is simple: the future could contain an astronomically large number of people—by some estimates 10^58 or more. While humanity might collapse soon, we might also survive for billions of years, and our actions today have some chance of affecting cosmic timescales. Since future people vastly outnumber present ones, nearly all expected moral impact is on them. The author illustrates with a cave-dweller analogy: if 99.999999999999999999999999999999% of all people lived in caves and our actions affected them, would we not prioritize them? Future people are analogous.

The expected value calculation is staggering. Assuming an existential catastrophe would reduce future population by 10^40, and that $10 billion can reduce existential risk by 0.0001%, each dollar spent adds 10^24 expected future people in expectation. Even if you are skeptical about the numbers, the argument holds as long as there is any non-trivial chance of a huge future. The median size doesn't matter; only the mean expected value. This has profound implications for philanthropy, policy, and AI safety: the most effective interventions are those that reduce existential risks or otherwise improve the far future. The article concludes that Strong Longtermism is not bizarre but follows from standard ethical premises, making it a critical framework for anyone thinking about long-term impact.