Venture capital's primary filter is whether a market is large enough to support billion-dollar outcomes. Nonprofits could achieve greater impact by similarly prioritizing the scale of a problem before evaluating a specific intervention, orienting them toward transformative solutions rather than incremental ones.

The Lead Exposure Elimination Project's story reveals a potential weakness in GiveWell's model. Its preference for proven, repeatable interventions can lead it to decline funding for more uncertain but potentially higher-impact "hits-based" approaches like policy reform, which Open Philanthropy, with its different risk tolerance, was able to support.

The outsized impact of a few charitable interventions mirrors the power law in venture capital. This occurs because success isn't additive; it's multiplicative. Many independent factors (e.g., timing, execution, scale) must all succeed, creating a distribution with fat tails where a few winners dominate.

Sir Ronald Cohen critiques the philanthropic model, arguing that relying on donations keeps charitable organizations small, underfunded, and perpetually begging for capital. This prevents them from achieving the scale needed to solve massive problems, a flaw that impact investing aims to correct by creating self-sustaining models.

Don't dismiss high-leverage but hard-to-measure interventions like government capacity building. Use "cost-effectiveness thinking": create back-of-the-envelope calculations and estimate success probabilities. This imposes quantitative discipline on qualitative decisions, avoiding the streetlight effect of only focusing on what's easily measured.

Reaching a 100x increase in charitable impact isn't from a single change but from combining principles that each act as a multiplier. For instance, shifting focus to a more neglected problem (10x) and choosing a leveraged policy solution (10x) can result in a 100x total improvement.

A critical flaw in philanthropy is the donor's need for control, which manifests as funding specific, personal projects instead of providing unrestricted capital to build lasting institutions. Lasting impact comes from empowering capable organizations, not from micromanaging project-based grants.

The 'effectiveness' in Effective Altruism creates a bias toward quantifiable problems like global health, while overlooking harder-to-measure but potentially higher-impact areas. For instance, preventing political dysfunction or misinformation among influencers could have a far greater downstream effect than many targeted donations, but it's not a typical EA cause because its impact is difficult to quantify in advance.

A charity like Make-A-Wish can demonstrably create value, even exceeding its costs in healthcare savings. However, the same donation could save multiple lives elsewhere, illustrating the stark opportunity costs in charitable giving. Effective philanthropy requires comparing good options, not just identifying them.