Interview, Fireside Chat, Podcast
The core of effective altruism & how to argue for it | Benjamin Todd (2020)
- Ben Todd and Arden Kaler argue that effective altruism (EA) is widely misunderstood by the public and adjacent communities, often being conflated with specific tactics rather than its core intellectual project.
- A common misconception, exemplified by a definition offered by Sam Harris, restricts EA to immediate donations for the poorest living people, erroneously excluding future generations and existential risk.
- Current data indicates that while global health receives approximately 30% of EA-focused funding, the majority of resources are directed toward existential risks, global priorities research, community building, factory farming reduction, and criminal justice reform.
- The community is increasingly shifting from an "evidence-backed" approach (relying heavily on randomized control trials) to a "hits-based" approach (identifying high-risk, high-reward opportunities with strong qualitative arguments but limited empirical testing).
- Todd clarifies that EA does not define itself by a moral obligation to donate; rather, it posits that if an individual chooses to use resources (money, time, career) to help others, they should seek the most effective methods to do so.
- Todd defines the core claim of EA as seeking the best way to contribute to the "common good," which he characterizes as an impartial perspective on improving the welfare (happiness/fulfillment) of sentient beings.
- Todd proposes a three-premise argument for the importance of EA:
- Spread: There are massive differences in how much good different actions contribute to the common good.
- Identifiability: It is possible to identify these unusually high-impact actions with moderate effort.
- Novelty: The best actions often differ significantly from what common sense or tradition suggests.
- The argument posits that the moral "mistake" of not practicing EA is proportional to the magnitude of these three premises; if the best actions are only marginally better than average, or impossible to find, EA becomes less critical.
- Potential objections to EA involve rejecting these premises via:
- Epistemic Humility: Arguing that aggregating the views of many experts or deferring to common sense is superior to individual analysis, implying society is already prioritizing the right things.
- Regression to the Mean: Suggesting that extreme claims of high impact made by individuals are likely statistical errors that disappear upon deeper investigation.
- Career Constraints: Noting that highly neglected causes may be difficult to enter due to a lack of passion, skills, or community support, potentially offsetting the theoretical impact gains.
- Todd cites specific examples of impact magnitude:
- A typical US college graduate donating 10% of their income to effective global health charities could save approximately 200 lives over a 40-year career.
- Diet choices can yield a 1,000x difference in animal suffering reduction, with giving up chicken/fish/eggs being more impactful than giving up beef alone.
- Todd argues that existing institutions (markets, governments, nonprofits) contain significant "gaps" regarding long-termism, global public goods, and neglected cause areas, validating the need for EA to fill these voids.
- To improve communication, Todd suggests framing EA as an intellectual project centered on questions (e.g., "Which problems are most pressing?") rather than a community defined by specific behaviors like "giving 10% to malaria nets."
- Todd recommends de-emphasizing global health in introductory explanations to better reflect the community's actual diversity of focus, which now leans heavily toward existential risks and long-termism.
- Arden and Ben call for more rigorous empirical research into "global priorities" to quantify the actual spread of impact between different actions, noting a current lack of published data on this core premise.
- The episode references a recent article by Will McCaskill, "The Definition of Effective Altruism," and a 80,000 Hours blog post detailing misconceptions about the movement.
- Todd notes that objections to EA in practice (e.g., community biases) are distinct from objections to the core intellectual argument, which requires rejecting the premises of spread, identifiability, or novelty.