Hacker News Reader

Comment by kr0bat | Hacker News Reader

I think the QALY criticism grazes the real fundamental problem with EA. You can't really quantify "good" good enough to create a meaningful measure. Being alive is better than being dead (in almost all cases). Being sighted is better than being blind. But is losing an arm objectively worse than losing a leg? Who has it worse, a gay man in a seriously homophobic country or a woman in a severely misogynistic country? Sure, donating money to an animal you find cute may not be the best use of money, objectively. But maybe it's better to be openly subjective, lest we end up codifying our own subjective beliefs.

View on HN

This is missing the forest for the trees. When I donate to http://givewell.org , I don't care about the relative value they place on arms vs legs. What I care about is avoiding pseudo-charities. You know, the ones that pop up in exposé articles every now and then where most of the money goes to marketing and admin, a tiny sliver goes to the actual cause, and then a clown car pulls up and unloads 30 people who try to rationalize how this is actually a good thing? I want to avoid those. As long as givewell can somewhat reliably steer my money away from pseudo-charities, as far as I am concerned they are earning their 10%.

> earning their 10%

Note that GiveWell only gets 10% of your donation if you explicitly request that they do. https://www.givewell.org/donate/more-information#How_is_Give...

Right. I happily tick the box because I believe in the importance of what they do.

If all "effective altruism" ever meant was "avoid giving money to inefficient charities, and we'll help you figure out which ones are which", then it would be an unambiguously positive thing. But, from my perspective, it's like most of the people originally involved in the movement got bored with that, and went all in on all this other stuff, justified by philosophical shenanigans, resulting in not just different but often contradictory conclusions.

For instance, lots of wealthy techie types who bought into EA donated heavily (and I think still do) to charities targeting things like risks to far-future humans and life extension. To me, those are also "pseudo-charities". There is no way to evaluate their effectiveness. They could be 100% effective or 0% effective, we won't know for generations or millenia...

And - just like all the "pseudo-charities" researched by GiveWell - every dollar that has gone to these organizations with unknowable effectiveness could have instead gone to known-effective charities.

Rich people have been using charity to avoid taxes, rationalize power acquisition, solicit donations for dubious causes, promote blatantly self-serving policy, etc since forever. We agree on classifying those as pseudo-charities, we agree that these things happened under the EA banner, and we agree that this should trigger some self-reflection from anyone who enabled them and rebranding from anyone who didn't.

However, I'm still going to do my charity through GiveWell. GiveWell makes many people with dubious pet causes very angry and these people have taken the adjacent scandal as an opportunity to attack the very idea of vetting charity. If someone wants to propose an alternative vetting strategy, I'm all ears, but I don't see the sins of SBF & co as a reason to stop vetting. I just don't. I am of the opinion that we need vetting now more than ever, and whatever we want to call that going forward I plan to support it.

> I don't see the sins of SBF & co as a reason to stop vetting.

Who's arguing against vetting?

100% agree. I just think of GiveWell as a totally different thing than "effective altruism" now, even though it is where I first heard of the concept. (And that predates the SBF scandal by quite a while!)

To their credit, they never recommended these "longtermism" initiatives as effective places to direct charitable funds. So I think their credibility as an organization comes out of this intact, despite some of the people originally associated with them coming out looking much worse.

But I dunno, maybe they and/or people like us who support them should be doing more to explicitly distance them from the rest of the EA movement.

> You know, the ones that pop up in exposé articles every now and then where most of the money goes to marketing and admin

Yeah, there are even ones that buy manor houses [0].

[0] Effectivealtruism.org purchased a £15M estate for its headquarters in 2021. https://news.ycombinator.com/item?id=33903850

From https://www.astralcodexten.com/p/my-left-kidney - "conference venues kept ripping them off, having a nice venue of their own would be cheaper in the long run, and after looking at many options, the “castle” was the cheapest".

You're anchoring on 'castle'. If it was a boring office building, no-one would care.

That quote you use was from a comment on that post, and the full context of that comment is arguing the complete opposite of what you are saying here:

> You said:

> "Obviously this kind of thing is why everyone hates effective altruists. People got so mad at some British EAs who used donor money to “buy a castle”. I read the Brits’ arguments: they’d been running lots of conferences with policy-makers, researchers, etc; those conferences have gone really well and produced some of the systemic change everyone keeps wanting. But conference venues kept ripping them off, having a nice venue of their own would be cheaper in the long run, and after looking at many options, the “castle” was the cheapest. Their math checked out, and I believe them when they say this was the most effective use for that money. "

> I'm curious to see the math for Wytham Abbey. All I ever saw was Owen Cotton Barrett's response when this was brought up over on the forum. The actual cost benefit analysis for paying £15 million wasn't made clear. As one comment pointed out, the reasoning given for purchasing the castle "could have been made virtually verbatim had the cost been £1.5m or £150m."

> You might have seen more information than me though.

> Owen Cotton-Barretts comment -> https://forum.effectivealtruism.org/posts/xof7iFB3uh8Kc53bG/...

I think taking this statement ("we were getting ripped off") without any data goes exactly against the while idea of effective altruism.

(A nit, but the quote was from the body of the post, not a comment on the post.)

A decent subthread on this on the EA Forum: https://forum.effectivealtruism.org/posts/xof7iFB3uh8Kc53bG/...

The money came from a grant from OpenPhil for the purpose. EVF didn't publish a cost-benefit analysis my knowledge, but a back-of-the-envelope calculation [https://forum.effectivealtruism.org/posts/xof7iFB3uh8Kc53bG/...] suggests it would plausibly cost £1~2M a year to hire similar venues. That said, whether it's worth it to purchase a dedicated venue depends a lot on how much use the venue actually gets.

Ultimately, "we were getting ripped off" is a financial judgement made by CEA. I don't think it really has a great deal to do with the wider effective altruist movement tbh.

I think there is no shortage of objective ways to do good. If we skip over those, seeking the more subjective ways, it is probably in service to one's self, not to the goal of doing good.

They bought two effective altruist castles because they said the ambiance would help their thinkers and spruce up their conferences.

Citation needed. From what i know, https://www.astralcodexten.com/p/my-left-kidney says "conference venues kept ripping them off, having a nice venue of their own would be cheaper in the long run, and after looking at many options, the “castle” was the cheapest". Zero mention of ambiance and improving thinking.

As a side note, Astralcodexten blog by Scott Alexander is a fascinating read even if you are neither effective nor altruist.

TFA quotes an "EA forum", and some Googling leads to https://forum.effectivealtruism.org/posts/xof7iFB3uh8Kc53bG/...

This might be the explanation astralcodexten was discussing [1]. Geoffrey Miller, a commenter, (I'm not involved in EA enough to know if that's a household name) says this

    I've been to about 70 conferences in my academic career, and I'm noticed that the aesthetics, antiquity, and uniqueness of the venue can have a significant effect on the seriousness with which people take ideas and conversations, and the creativity of their thinking. And, of course, it's much easier to attract great talent and busy thinkers to attractive venues. Moreover, I suspect that meeting in a building constructed in 1480 might help promote long-termism and multi-century thinking.

Of course this man didn't buy the castle, so I don't think it's fair to hold a random commenter's thoughts against CEA. I'm not big on cognitive decoupling, but I also find buying a castle for a charity event venue shocking, and I don't think there were enough numbers in this post to soothe me. If renting event venues was too expensive, couldn't they have met virtually?

[1] https://forum.effectivealtruism.org/posts/xof7iFB3uh8Kc53bG/...

Thank you for finding the original source, that's legitimately very helpful! I think this message can stand for itself much more then the parent can. You can argue with it and disagree, but at least the disagreement is more reasonable then they did it only for the ambiance.

Although it doesn't seem to mention the issue with being unable to find other conference centers, so that reason may still be valid or might be a mistake on Scott's part.

I suspect the real explanation has way more to do with Yudkowski's Harry Potter fanfiction and seeing the rest of the world as "Muggles" than they would be comfortable explaining.

It's right here: "having a nice venue"

Why do they need a "nice venue"? Why wouldn't any airport hotel's conference room be sufficient? They wanted something nice, more than utilitarian. They wanted luxury, and how does an "effective altruist" defend that? Because luxury makes them "more effective" of course.

I follow a business rule of thumb that relates to this. The "chandelier rule": the bigger the chandelier in the room, the worse the deal being pitched to you is.

If someone is spending lavishly on pitching to me[1], I start to really question the deal. At best, they're showing me that they don't mind wasting money on things that don't matter. If you want to impress me, wine and dine me by taking me to McDonald's.

[1] I have been picked up in expensive sports cars, I have been taken out to eye-wateringly expensive restaurants, etc. This is the sort of thing that raises big red flags to me.

Regardless of why they bought the castle, or whether it would be more accurate to call it a manorial estate, it's a long way from a simple website that sends out mosquito nets and deworming tablets.

> Regardless of why they bought the castle

Two separate castles:

https://twitter.com/paulmainwood/status/1600433194691502081

https://www.chateau-hostacov.cz/en/contact

They really ran with the exact opposite of Peter Singer's sacrifice your fancy boots to save a child parable.

And Effective Altruist Sam Bankman-Fried, who drove a Prius for optics, lived in what has been cited as potentially the most expensive private residence in the Bahamas. He was grilled on it a bit before the collapse and explained that worrying about what he did with 1% of his money wasn't justifiable use of time in the big picture--still dubious given the Singer origins of the movement)--but a bigger problem was close to 100% of the money was fictional: illiquid self-issued token valuation, and double spending/self-loaning of customer funds.

Just like other charities that spend nearly more of the money on fundraising and growth that were the main target of criticism of the movement, they did pretty much the same, while pitching it as a more effective alternative to Oxfam, etc.

The second one seems to be a link to a hotel. Where's the source for the claim that it was bought by EAs?

https://forum.effectivealtruism.org/posts/pbe8x4AQDqftQoaT5/...

your attitude to citations seems strange. Scott Anderson simply says, as part of a rant against anti-EA people, that the EA people said that they bought a castle for the right reasons, and that he chooses to believe them.

The problem with EA isn't a focus on objective METHODS to do good, but with the belief that there are potential recipients of "good" that are objectively better targets than others, and the belief that finding objectively better targets is even achievable.

So it's just as good to give 1,000 to someone with a 6 digit salary as to an African peasant?

No. In utilitarian EA terms it’s 100% the opposite, you should give it to the peasant because they’ll get a lot more utility out of it.

that was entirely my point. it's extremely hard to beat giving to the global poor.

EA does not believe there are objectively better recipients, only objectively more cost effective. It's not about moralizing at all. You can choose whatever objective function you want, there are some ways to achieve that objective that are more cost effective.

Does EA not believe in objectively better recipients?

    MacAskill writes that QALYs can be used to decide which charitable causes to prioritise: faced with a choice between spending $10,000 to save a 20-year-old from blindness or the same amount on antiretroviral therapy for a 30-year-old with Aids – a treatment that will improve their life and extend it by ten years – MacAskill argues it would be better to perform the sight-saving surgery, as the 20-year-old can expect to live another 50 years. He acknowledges that QALYs are an “imperfect”, “contested” measure but sees them as mostly good enough.

MacAskill does, but he's doing applied ethics here. EA is a normative ethical framework. You are free to disagree with his objective function and still be EA as long as you are thinking about how to maximize your own.

Reply on news.ycombinator.com