Contra MacAskill on saving money for the intelligence explosion

LessWrong is hosting a debate about the timing of large charitable donations aimed at AI risk. Effective altruism leader Will MacAskill recently argued that wealthy donors should invest their money now and wait until after a possible "intelligence explosion"—rapid AI advancement expected in the early twenty-thirties—to give it away. A contributor disagrees, saying donors should give away thirty to fifty percent of their philanthropic budget within the next few years instead. The core argument is economic. If you wait for your investments to multiply tenfold, so will the wealth of other major donors already invested in AI companies. You turn one billion into ten billion, but the hundred billion in AI-founder wealth held in Anthropic and OpenAI will also multiply tenfold, creating a vastly larger philanthropic pool. Your ten billion becomes a tiny fraction. Your impact depends on your share of total philanthropic funding—so giving now, when your percentage is bigger, means your money influences more of what gets funded. There's also a strategic risk. Waiting until AI can solve alignment problems itself and then handing it hundreds of billions of dollars sounds like surrender. Donating now, while humans still hold real influence, lets donors direct AI toward good. Early giving also seeds organizations with years to grow and build legitimacy. The LessWrong piece flags specific opportunities: the twenty billion dollars in the twenty twenty-six US election, or backing early-stage projects with time to compound their impact.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton