A place for new ideas to settle.

08 September 2015

REVIEW: "Doing Good Better" and "Living on One Dollar" -- The varieties of altruistic experience

I.

I don't know if another book has sparked more cognitive dissonance in me than Doing Good Better.

Written just this year by William McAskill, a philosopher and co-founder of both 80,000 Hours and Giving What We Can, the book represents a sort of manifesto for the Effective Altruism (EA) movement. I did want to read the book for straightforward reasons—my parents pretty well instilled in me the ethic of "live within your means, and give some of the extra to those less fortunate," so any way to do that better is welcome—but also for more circumspect ones. Namely; EA has grown pretty fast since its generally-agreed-upon inception in 2007 (the year GiveWell was founded; but the EA brand-name wasn't invented until 2011), and has just as quickly attracted some fairly harsh criticism for something so... altruistic. I'll get into that later in the post, but for now it was the other reason to read the book: could I see in it the misguided or even misanthropic tendencies alleged by EA's critics?

So, the book: Doing Good Better is persuasive and also frustrating. While the first half's argument for why effective altruism is worthwhile is pretty great, the second half's implied philosophy of how to be an effective altruist struck all sorts of nerves with me, and I'm still trying to parse why that was. Part of it is that the how-to strategies don't seem able to be generalized; for example, at what point does "an ER doctor" become the best option for someone considering their future career? Career comparison, moreover, is rather hard for paths that require a lot of invested effort—software engineering, brought up more than a few times in the book, is unusual in this regard—and the ability to gather a decent sample of possible careers seems like it would be rare in a community, even the industrialized West.

But let's take it a bit slower. Why "effective altruism"? McAskill's argument is that we need to be very careful about which aid organizations get our donations, because the very best organizations are responsible for the vast majority of improvement in their respective causes, even compared to the next lowest quantile of merely very good organizations. So basically, if you just randomly or thoughtlessly donate your money, there's a very good chance it's as if you just burned the check.

He backs it up with quite a few examples, some of them very high profile. The worst realization for me came right in the introduction, where McAskill recounts the story of the PlayPump. The basic theme is the same as pretty much every high-profile failure: celebrities and corporations joyously endorse a solution that's completely outside-context and an utterly poor fit for the local problem. The PlayPump, for example, did a worse job than the mediocre hand pumps the villagers had been using before! Not to mention the part where some PlayPumps were installed without consent.

And then, in the face of overwhelming counter-evidence, the PlayPump organization... kept right on making them. That was my cynical realization: bad charity is pretty much like pseudoscience or alternative medicine. Debunk it all you want, it won't die. As long as it helps someone, just once, it's "worthwhile" and keeps pulling in funding.

After raising the reader's hackles against the status quo in the first few chapters, McAskill lays out a framework for deciding how effective a charity is. This is really good stuff, and should be considered by everyone: there's really no excuse, if your motives are truly altruistic, for not doing at least a little homework. Charities, too, ought to be more diligent about transparency and open data.

Now to the second part: how to be an effective altruist. As I said above, my big problem with this section was its seeming pretension to generality without that being obvious, or else its lack of stipulation that the rules can't apply to everyone. The whole concept of "earning to give" as a terminal goal hinges on the idea that the rest of the workforce is pretty well in hand, so you as an effective altruist are better served going for the highest-paying version of whatever career path suits you best. Doctor? How about the kind of doctor that operates really expensive equipment, or only on rich people who are insecure about their appearance? That sort of thing. And the ever-popular options of "software engineer" and "quantitative financial analyst" came up, too.

Except that line of argument seems really facile. Taken as a general strategy, it totally ignores the uneven distribution of jobs by specialization and geography. Yes, you might be marginally non-useful as the 800,001st doctor, but what about the thirteenth doctor in a rural county? Or the first abortion-providing doctor in the state? Quite a different marginal impact. This raised a question in my mind about the ethos of effective altruism, which I'll get into below.

There's also the issue of calculation: the heuristic used in the book is Quality-Adjusted Life-Years, or QALY (rhymes with "Wally"), which McAskill generalizes from medical charity research. It's not clear how well QALYs work with other kinds of causes and interventions; however, I recognize the lack of a robust method since EA is a relatively new prospect. Something is better than nothing, but I worry (see below) that arguments based on QALY estimates are not being properly discounted.

Overall it's certainly a conversation-starter, and a good introduction to why we should be thinking rigorously about our charity. And as a framework for evaluating charities and causes, I highly recommend it. (Along with GiveWell and Giving What We Can.)

II.

I also watched Living On One Dollar, a 2013 documentary produced by some of my high-school classmates. As a film, it's pretty great, especially given that it was filmed on a shoestring budget by college sophomores. They bring an earnestness and positivity to a genre that, in their own opinion (and mine, to some extent), relies more often on guilt-tripping the audience—and yet they don't veer too far into the empty optimism or dangerous naivite of something like the #KONY2012 thing.

In the film, college friends Zach (Bainbridge High School represent!) and Chris decide to study rural extreme poverty by traveling to Guatemala and living on just $1 per person per day. They are joined by friends and filmmakers Sean (BHS!) and Ryan (er, actually I don't know either way), who are good at photography but ignorant of Spanish, much less the native Mayan dialect. Zach and Chris, on the other hand, do speak Spanish, and thus acquire a very different understanding of the hardships that come with extreme poverty—a difference that comes to a head in the film's climax.

They also further simulate the experience of poverty by randomizing their daily income, anywhere from $9 to $0 drawn out of a hat each day. Between the villager interviews and a series of extremely-low-draw days, you approach a sense of the chaos inherent in that situation.

The film does a pretty good job of showcasing just a few of the barriers that stand between the villagers of Peña Blanca and a stable (if lower-working-class) life:
  • Language—All of the villagers speak the Kaqchikel a Mayan language, and not many can afford to learn Spanish. Wiki says Guatemala isn't terrible about this and there are bilingual classes in school, but lack of institutional support for indigenous and minority languages is a huge barrier in many places around the world.
  • Labor—In the film, only one villager, Antonio, has a stable job. The rest work informally as day laborers working on farmland.
  • Nutrition—Food is not that expensive in Guatemala, but uncertain, poverty-level income makes it so nobody's really eating well. Even on rice, beans, lard, and bananas, the guys are barely breaking 1000 calories a day and one of them loses something like 20 pounds over the two months they stay in Pi a Blanca
  • Healthcare—When Chris comes down with Giardia parasites they get him checked out at a doctor in the nearby town. Medicine is stupidly expensive, essentially beyond the financial reach of any villager without some serious financial trauma.
  • Credit—Traditional banks in Guatemala, consistent with what I've read about in general, pretty much refuse to extend lines of credit to poor people. You have to meet an absurd number of requirements that are simply beyond the means of subsistence farmers and informal day-laborers.
  • Education—Schools in Guatemala require tuition on the order of $25 per student per year. Basically, if anything bad happens (disease, injury, hurricane...) there goes the money for school.
The film doesn't get into the issue of land rights, which I understand are a big issue in some countries—title and use rights are often legally fuzzy in these places, so poor farmers are at risk of being exploited and deprived of potentially a lot of wealth. Or they simply have no real access to the equity that's literally beneath their feet. This idea formed the basis of Hernando de Soto's (not the conquistador) work on property rights and the informal economy. But then again, given the film's scope, I can see why they never went there.

My biggest  nitpick is that the film mentions that microfinance (institutions like the Grameen Bank, which actually does make small loans to extremely poor people) has a mixed reputation, but they never follow through on it. On the one hand, it's great that the villagers who took out loans seemed to be doing really well (starting businesses, which would probably help pay off the loan rather quickly)... but what happens if someone has to default? The guys take out a loan themselves—to represent buying the farm-hut they're living in—and risk not being able to make a payment ($6.25 every three weeks, I think!) but they never go into detail about what would happen. Given the volatility of a typical villager's daily income, it seems like a problematic omission.

Oh yeah, and the film gets you right in the feels any time they talk to kids. Or Antonio. Or Rosa. Or...

Overall it's a pretty good effort by some people I actually knew in real life. Well done!

III.

Taken together, Doing Good Better and Living On One Dollar present an interesting juxtaposition of altruistic experiences. The book emphasizes high impact, global efforts. The film zooms in on a highly local struggle, though one reproduced all over the developing world. Both focus on experiences pretty much exclusive to the developing world—there's severe poverty in the United States, for example, but its character seems very unlike that of the rural Guatemalans in Living On One Dollar. And both aim, in their own ways, to break the mold of how we think about charity and foreign aid.


They also highlight the extremely privileged situation the generic reader or viewer will be in: literally anyone with a job in the United States is better off economically than billions and billions of other human beings. (Even with all the stories of economic hardship that working-class and middle-class Americans face.) That's the argument Doing Good Better makes, and uses a lot of graphs and statistics to back it up. The film goes one better, and shows us two guys from very comfortable backgrounds—New York for Chris, Bainbridge Island ("Seattle") for Zach—experiencing some fraction of the daily struggle of rural Guatemalans firsthand. I've written about this sort of educational tourism before, and I think the filmmakers pretty much avoided the pitfalls of WEIRDness as much as could be expected.

In the end the message is the same: You, reader, occupy such a privileged position in the world that you have more or less a moral duty to contribute to these causes. If you don't, who will?

Fair enough. But now we come to the crux of EA's oddness: what cause(s) should you donate to? Well, one EA principle is that of "impartiality"—we shouldn't discount the needs of any moral agent. This includes people at the furthest reaches of human habitation, other animals, and any of these that may come to exist at some future time. This generates, in order, EA's championing of global poverty and disease eradication; animal advocacy and veganism; and existential risk mitigation.

Ah, x-risk. This is what makes EA look like a movement of self-obsessed tech nerds. Not that, say, making sure we can see oncoming asteroids isn't important. Nor is research into neutralizing rogue technology, whether nuclear, bio, or nano. No, it comes down to two factors.

First, the "every life is sacred" principle massively weights the argument in favor of x-risk as the exclusive cause to donate to. Nick Bostrom, a philosopher researching x-risks, calls this the argument from "astronomical waste":
As I write these words, suns are illuminating and heating empty rooms, unused energy is being flushed down black holes, and our great common endowment of negentropy is being irreversibly degraded into entropy on a cosmic scale. These are resources that an advanced civilization could have used to create value-structures, such as sentient beings living worthwhile lives.

The rate of this loss boggles the mind. One recent paper speculates, using loose theoretical considerations based on the rate of increase of entropy, that the loss of potential human lives in our own galactic supercluster is at least ~10^46 per century of delayed colonization.[1]
He goes on to refine this estimate but it's still mind-bogglingly huge. And that's how you end up with EA pivoting to artificial-intelligence-safety-research for the keynote speech at the EA Global 2015 conference. Not "how many diseases can we eradicate?", or "how can we end factory farming?", but pretty much "how can we appease an alien god?"

And that's the second problem: If you posit superintelligence, you can get away with a lot of whizz-bang "predictions" that are way to close to "what the Kingdom of God will be like once Jesus returns" for my liking. And for all the admonishments to not take futurologists seriously if they start throwing out too many specific claims:
Which is to say:  Adding detail can make a scenario SOUND MORE PLAUSIBLE, even though the event necessarily BECOMES LESS PROBABLE.

If so, then, hypothetically speaking, we might find futurists spinning unconscionably plausible and detailed future histories, or find people swallowing huge packages of unsupported claims bundled with a few strong-sounding assertions at the center.
Well, there are a lot of specific claims floating around by those same admonishers:
Amazon screenshot showing related purchases for "Smarter Than Us" by Stuart Armstrong
What happens when machines become smarter than humans? Forget lumbering Terminators. The power of an artificial intelligence (AI) comes from its intelligence, not physical strength and laser guns. Humans steer the future not because we're the strongest or the fastest but because we're the smartest. When machines become smarter than humans, we'll be handing them the steering wheel. What promises—and perils—will these powerful machines present? Stuart Armstrong’s new book navigates these questions with clarity and wit. Can we instruct AIs to steer the future as we desire? What goals should we program into them? It turns out this question is difficult to answer!

[...]


Are we up to the challenge? A mathematician by training, Armstrong is a Research Fellow at the Future of Humanity Institute (FHI) at Oxford University. His research focuses on formal decision theory, the risks and possibilities of AI, the long term potential for intelligent life (and the difficulties of predicting this), and anthropic (self-locating) probability. Armstrong wrote Smarter Than Us at the request of the Machine Intelligence Research Institute, a non-profit organization studying the theoretical underpinnings of artificial superintelligence.
Armstrong begins that book, Smarter Than Us: The Rise of Machine Intelligence, with a parable about the Terminator getting its digital shit wrecked by a superhuman A.I.:
"I thought I should keep you up to date as to what I've been doing," [the superintelligence] said. "Well, I started by locating the project that would become Skynet and leaked its budget to various Senate subcommittees. The project will become a political football between budget hawks and military hawks before finally being cut in a display of bitpartisanship in about three months' time. I also figured out how to seduce a photogenic fireman, who'll be leader of the new political party I'm setting up—funded by my investments. (Do you have any idea how easy it is for me to predict the stock market?) I have already written a few speeches that will bring tears to the eyes of every human who hears them. It'll ensure no alternative version of Skynet is ever built, anywhere or anywhen."
At least when Iain M. Banks posited godlike machine Minds, he had the decency to give them hyperspatial computing substrates. (This is the genesis of a whole 'nother post so I'll have to leave that there for now.)

IV.

The ancient Stoics had a meditation exercise. Imagine progressively generating these circles:

Source: Massimo Pigliucci, How To Be a Stoic (https://howtobeastoic.wordpress.com/stoicism-101/)
The idea is that compassion begins with the self, then extends with less and less natural ease to family, neighbors, countrymen, and humanity as a whole. The Stoic meditation is an exercise in fellow-feeling towards these wider circles, in progressive waves, so as to expand one's compassion towards humanity. (You can, of course, continue from "humanity" to "sentient creatures" or even "conscious creatures.")

In some sense the Effective Altruism movement is trying to radically achieve that last circle, or even extend it beyond to "conscious creatures" (in the form of animal-rights activism), but by "hacking" around natural fellow-feeling. I think it would be well-served to seek out more human stories to generate that compassion, rather than just the warm glow of having "optimized" something.

To that end I think Living on One Dollar offers a more humanistic approach. By all means, be effective in your charity, crunch all the data you need to—but seek out the human stories that come of it, because it's not really the data that needs your aid.

No comments:

Post a Comment