|
Abstract
|
Shouldn't charity serve the needs of recipients, not givers?
Isn't it better to do more good than less?
Shouldn't there be some way to measure that?
Effective altruism is the philosophy that answers "yes" to all these ques- tions. Put this way, it sounds entirely innocuous. So why was it one of the hottest ideas in tech circles in the 2010s? And why is it playing a central role in so many Silicon Valley controversies of the 2020s?
If we use headlines as our guide, effective altruism has fallen from grace. One of its leaders, Eliezer Yudkowsky, also a founder of AI safety research, notoriously called in Time last year for global limits on AI development that are enforceable by airstrikes on rogue data centers. Sam Bankman-Fried claimed it as a motive for what turned out to be his multi-billion-dollar fraud. It reportedly drove board members of OpenAI to fire Sam Altman over concerns that he wasn't taking AI safety seriously enough, a few days before some of them were pushed out in turn. And it has spurred the pro-AI backlash movement of "effective accelerationism," which regards effective altruism as the second coming of Ted Kaczynski.
In public view, effective altruism shows up as a force of palace intrigue in the halls of Silicon Valley. And it is losing the favor of the court. How did a movement based on an idea so obvious it seems trivial -- "doing good better" -- become so strange, and fall so dramatically?... The Drowning Child
For some effective altruists, simply doing more to help, for people we're not helping now, really is all that it's about. Julia Wise, a former president of Giving What We Can and one half of the Boston couple mentioned above, counsels that it's okay for people to pick and choose which parts of the philosophy work well for them. And Scott Alexander, a major EA figure and its most philosophically serious, argues against "extreme rationality" and says that the movement is about significant directional improvement rather than perfection. To these figures, the weird parts are just a sideshow. Should we judge every movement by its extremists?
The problem here is not just an attitudinal one. Consider the famous "drowning child" thought experiment proposed by Princeton philosopher Peter Singer. Though later figures were more directly responsible for founding EA, Scott Alexander wrote in 2022 that "the core of effective altruism is the Drowning Child scenario." Alexander is right -- this spe- cific scenario is a fixation of EA writing. And we can see in it that the movement's problems grew right out of its intellectual roots. Singer's experiment, beginning in a 1972 paper in Philosophy and Public Affairs and expanded in later writings, asks us to imagine stumbling upon a child drowning in a pond, and then imagine another child drowning or starving or succumbing to disease on the other side of the world.
OBJECTIVEly, we must know that the distant child is no less urgent a problem than one I just so happen across: "If we accept any principle of impartiality, univer- salizability, equality, or whatever, we cannot discriminate against someone merely because he is far away from us (or we are far away from him)." Thus we must use reason to extend to distant causes an equal level of concern to those near. Singer, borrowing from the Irish historian William H. Lecky, calls this outward motion the "expanding circle" of moral concern. ... |