Korva is a scroll of the internet’s best ideas.
Effective altruism stopped measuring
The original mission
The movement set out to measure what philanthropy actually accomplished and to optimize for tangible results. Somewhere along the line the measuring stopped, and it became about pet causes and money.
Pet causes take over
One of the first pet causes was pandemic prevention, when Sam Bankman-Fried was trying to create a Pandemic Prevention Center in California. That cause fell out of favor when he went to jail, and the movement shifted to the next existential risk it could hang its hat on: P doom, the existential risk of AI.
The turn to AI
Anthropic styled itself as the one frontier lab that would solve P doom or behave in a responsible manner that cared the most about it, and so it attracted many people from the movement. Rationalist communities like Slate Star Codex also got into AI early, and the same circles now have a couple of trillion dollars in market cap to work with.
How did GiveWell measure charities?
GiveWell was founded in 2007 by Holden Karnofsky and Elie Hassenfeld, who had worked at a hedge fund. It estimates how much good a donation does, such as the cost to save one life, and ranks charities on that basis. Its early top charities included the Against Malaria Foundation, which distributes insecticide-treated bed nets.
How did the rationalist community form?
Eliezer Yudkowsky founded the website LessWrong in 2009 as a place to discuss clear thinking and AI risk. Scott Alexander's blog Slate Star Codex, started in 2013, drew a large overlapping audience. Many members of these communities later worked at AI labs or on AI safety research.
How did AI risk enter effective altruism?
Oxford philosopher Nick Bostrom's 2014 book Superintelligence argued that advanced AI could pose an existential threat, and it circulated widely among effective altruists. The organization Open Philanthropy, which grew out of GiveWell and is funded largely by Dustin Moskovitz and Cari Tuna, began making grants on AI risk in the mid-2010s. The career guide 80,000 Hours later listed AI safety among its top recommended paths.