05.04.2026 –, The Rabbit Hole
How much is a human life worth? Governments routinely spend millions of euros to save a statistical life. Yet some global health interventions can prevent a death for a few thousand euros. The difference is not small — it is often orders of magnitude.
If you would save a drowning child in front of you, does distance matter? Does time? How easy is it, in practice, for someone in one of the richest countries in the world to statistically save a life each year? If we care about improving the world, shouldn't we allocate our resources where they have the greatest measurable impact?
This talk explores Effective Altruism as a framework for answering these questions. We'll look at cost-effectiveness reasoning, heavy-tailed impact distributions (why some charities are 10–100x more effective than others), and the idea of treating doing good as an optimization problem under uncertainty.
Most people donate based on proximity, emotion, or tradition. Effective Altruism starts from a different observation: impact is not evenly distributed. In many domains — startups, security vulnerabilities, open-source projects — outcomes follow heavy-tailed distributions, where a small number of interventions account for a disproportionate share of results. Evidence suggests that charitable effectiveness may follow similar patterns.
If that is true, then doing good becomes a problem of allocation under uncertainty. Which interventions actually save or improve the most lives per euro? How robust are those estimates? And what follows if some causes appear dramatically more leveraged than others?
We’ll look at concrete examples from global health, briefly touch on why some in the community prioritize long-term risks such as advanced AI systems, and explore how a hacker mindset — measuring, iterating, questioning assumptions — maps naturally onto impact-oriented decision-making.
tesuji currently isn’t working for money, but out of curiosity. He likes taking things apart — devices, models, occasionally even assumptions — and manages to put them back together most of the time. He hosts dinners with political and philosophical discussions, designs scavenger hunts, and has a weakness for leverage and optimizable systems.