Opinion – Ezra Klein: Learn what effective altruism is and why the cryptocurrency market has transformed it

by

This is my annual column on philanthropy, so I’ll cut to the chase. I recommend donating to the four highest rated charities by GiveWell: Malaria Consortium, Against Malaria Foundation, Helen Keller International and New Incentives.

These organizations distribute medicine and mosquito nets to prevent malaria, vitamin A supplements to prevent blindness and infant death, and money to vaccinate underprivileged children against a whole range of diseases.

What distinguishes these entities is the confidence we feel in the good they do. There is no shortage of organizations seen by donors as excellent but whose programs are never scrutinized—and when they are, the benefits often disappoint. The organizations cited are different: their work has been evaluated by studies that have shown that they save lives and prevent disease at low cost.

That would be a banal set of recommendations, but this year it’s a little different.

Effective altruism, the philanthropic movement of which GiveWell is a part, is going through a reckoning after the fall of Sam Bankman-Fried, its financier and most famous participant. His case was unusual: he started trading cryptocurrencies after having lunch with Will MacAskill. The Oxford philosopher and co-founder of the movement told Bankman-Fried that he could probably do good on a larger scale if he made a lot of money and gave it away, rather than working for some non-profit organization.

Two years later, Bankman-Fried put this idea into practice. “Earning to donate” is advice commonly given: don’t join an entity and build schools abroad, but go work for a fund and in a short time donate US$ 500,000 a year for the purchase of mosquito nets.

Bankman-Fried distinguished himself by the scale of his apparent earnings and proposed donations. He founded the FTX cryptocurrency exchange and amassed a fortune that theoretically reached tens of billions of dollars, which he promised to give away almost all of it. He was a constant presence in the media, talking about how he had linked his assets to his ideals.

But the fortune was never real. FTX was linked to Alameda, Bankman-Fried’s brokerage, and the companies transferred money to each other, backing their reserves with crypto assets that had no functional value. If the fortune was false, what about the ideals that supposedly moved it?

Did effective altruism help to justify or even motivate the risks taken and limits crossed that ended up vaporizing billions of dollars? How could a movement that prides itself on its long-term thinking so completely ignore the foibles of its wonder boy?

These are reasonable questions, but I don’t want to repeat too much the moral to be drawn from what is really an old story: a young, daring and reckless financier who operates in a basically deregulated market made a fortune in a short time manipulating deposits and lost everything.

I don’t think effective altruism is to blame for this, and I don’t want to see a growing backlash dominate a movement that has done so much good and can do more. So let me suggest some areas where effective altruism really needs to recalibrate — not just because of Bankman-Fried — and get back to where it started.

We should view ‘earn to give away’ with more skepticism

Earn to give away has always struck me as a thought experiment masquerading as life advice. The problem is not in your logic, but people are not automatons.

Highly paid professions transform those who exercise them. Examples abound to indicate that the Spartan tastes and immaculate ideals of college students rarely survive prolonged immersion in the values, pressures, and possibilities of growing wealth.

Earning to give creates a more disturbing possibility of rationalizing the use of unethical means in the service of virtuous goals. Giving people more motivation to accumulate as much wealth as they can must be weighed against the psychological and historical truism that money is a corrupting force—and that few who relentlessly pursue it do so without compromise.

And there is also the risk that large donors pose to the movements they support. Museums, for example, have had to weigh the pros and cons of donations from the Sackler family, which profited from opiates.

This is a particular weakness of effective altruism, which tends to idolize today’s tech billionaires. But one lesson to take from the Bankman-Fried fiasco is that the more donors support you, the more you are seen as supporting them. That’s a risk — one that a mature movement needs to take more seriously.

We should view thought experiments with more skepticism

I want to be careful when I say this, because effective altruism is rooted in a thought experiment that turned out to be positive. Philosopher Peter Singer asked if you should jump into a lake to save a drowning child, even if it ruins your new shoes. “Of course I do” is the natural response. Singer then asks: does it matter if the child is in a lake in front of him or in a country on the other side of the world?

If you believe in this idea — I do, to a great extent — then you have to face the difficult question of how far it goes. When the choice is between your comfort and someone else’s survival, even the most modest luxuries seem immoral. Only the most virtuous of us can follow this moral logic to the end, but a little more altruism is within reach of many of us.

But the culture of effective altruism is a little too fascinated with thought experiments and theoretical models of the future. This is a risk, as the work of those who save lives and those who imagine future lives can point in opposite directions. That doesn’t mean it’s not worth thinking about the future. But I have already noticed that in recent years the energy of effective altruism has been turned much more towards its speculative interests and obsessions. It’s high time to make a correction.

In the midst of the repercussions and consequences of the Bankman-Fried fiasco, an excerpt from a dissertation by Nick Beckstead, who until recently directed the FTX Foundation, began to circulate on the networks. He wrote: “It seems plausible to me that saving a life in a rich country is substantially more important than saving a life in a poor country, other factors apart.”

According to his reasoning, since inhabitants of rich countries produce and innovate more (at least according to the usual economic criteria), saving lives in poor countries “may have significantly less ripple effect than saving and improving lives in rich countries”. The dissertation is complex, and philosophers need to be able to present arguments that would sound outlandish outside the confines of their discipline. The question is which way of looking at the world will ultimately define effective altruism.

Because, as GiveWell found, the logic takes a precisely opposite turn when we limit ourselves to the outcomes we can actually measure. One of the main achievements of effective altruism has been to persuade rich country donors to invest more in effective entities operating in poor countries.

We should view false accuracy with skepticism

It’s worth asking how one movement can encompass two such disparate wings, and I think the answer is rhetorical. I think the more speculative wing of the movement has developed a culture that uses probabilistic thinking and invented values ​​in ways that create an unfounded perception of accuracy. So it might seem that people who focus on what we can know and those who focus on what we can’t know are doing the same job. But they are not.

Toby Ord released a book in 2020 that reflects on many of the paths humanity can take to total ruin. At its core is a table containing estimates of the likelihood that different existential risks could bring about our destruction in a hundred years: asteroids have a one-in-a-million chance of annihilating humanity; natural pandemics, one chance in 10,000. A nuclear war and the climate crisis each have a one-in-a-thousand chance. But the risk of being wiped out by artificial intelligence that gets out of hand is one in ten.

This table summarizes both the value and the danger of converting speculation into probabilities. The value is that Ord tries to present his views as accurately as possible. The risk is that this cold parade of numbers will endow these estimates with an authority they don’t deserve.

This is not to say that speculating is wrong, but I think that false precision can make the unreal seem real and hide the precariousness of everything that preceded quantification.

This is the clearest link between the Bankman-Fried downfall and certain elements of effective altruistic culture: crypto-wealth is based on assigning values ​​and probabilities to fictitious assets and currencies.

This problem exists in other areas of capitalism, but it is crystal clear in crypto markets. There is only code and quantification, and numbers lend unearned solidity to an abstraction. I think effective altruists have a tendency to delude themselves in the same way.

We can and should do good more often

I’m afraid this column will be interpreted as a reason to dismiss any queries that sound strange the first time you hear them. That’s not what I’m saying.

Effective altruists have gone to great lengths to persuade people to care more about artificial intelligence—and they are right to do so. We can’t look at the remarkable performance of the latest AI models and not think it’s important to reflect on the consequences of tremendously more powerful systems.

But I think too much of the energy and talent of effective altruism is moving away from the compassionate rigor that distinguished the movement in the first place and that the world still needs. It will be much less effective if it loses touch with its original goal of improving the lives of people alive today.

You May Also Like

Recommended for you