"I want my life to flash before your eyes."
July 8, 2022 6:56 AM Subscribe
"And every minute you spend with me is a minute that they too get to look for beauty." "The Unweaving of a Beautiful Thing" by atb depicts a battle between a witch and Death. It was posted to the Effective Altruism forum but is much more about character than calculations. 'There were two words that Superman lived by, and they were “pay me”.' Over on Archive of Our Own, "A Common Sense Guide to Doing the Most Good" by cthulhuraejepsen is an unfinished narrative of "Clark Kent, effective altruist" that addresses "the Crank Problem".
Most of the things about the effective altruism approach that I hear about are kind of ugly and inhumane, or are connected to proposals to take extreme steps regarding stuff like artificial general intelligence (AGI). That's why I thought it was interesting, and worth sharing, when I read these two pieces, which are enjoyable to read and which engage with effective altruism ideas in ways I found more sympathetic to my values.
posted by brainwane at 7:35 AM on July 8, 2022 [5 favorites]
posted by brainwane at 7:35 AM on July 8, 2022 [5 favorites]
I really enjoyed and was moved by the first story you linked, brainwane. Thank you.
posted by minervous at 8:49 AM on July 8, 2022 [5 favorites]
posted by minervous at 8:49 AM on July 8, 2022 [5 favorites]
ea/rationlism represent the most underexamined pipeline to far right politics on the internet today. that's really all i have to say about them and i feel obligated to point it out every time they are mentioned here which is surprisingly often.
posted by AlbertCalavicci at 8:52 AM on July 8, 2022 [10 favorites]
posted by AlbertCalavicci at 8:52 AM on July 8, 2022 [10 favorites]
I think the Superman one is fascinating but I see why it couldn't be finished: because by virtue of the premise, by the end of the piece it has already come perilously close to stumbling over the intrinsic problem at the extreme end of "effective altruism" - that the logical dictate of maximally effective altruism would be to identify the Most Effective Altruist(s) and then eliminate choice/free will from everyone else. (This underlying conceit, I suspect, also has something to do with why it's such an effective proto-fascist pipeline.) Unless you can quantify the value of free will, any "effective altruism" calculations become inherently suspect the moment you start influencing other peoples' behavior.
posted by mstokes650 at 10:14 AM on July 8, 2022 [7 favorites]
posted by mstokes650 at 10:14 AM on July 8, 2022 [7 favorites]
Thanks brainwane. Liked the first story quite a bit.
posted by aleph at 10:20 AM on July 8, 2022 [2 favorites]
posted by aleph at 10:20 AM on July 8, 2022 [2 favorites]
Like the money shot from Superman: Red Son: “why don’t you put the whole world in a bottle, Superman?”
posted by notoriety public at 10:21 AM on July 8, 2022 [2 favorites]
posted by notoriety public at 10:21 AM on July 8, 2022 [2 favorites]
Just learned what a "hedgehog slice" is.
Reading more now.
posted by Well I never at 10:57 AM on July 8, 2022 [2 favorites]
Reading more now.
posted by Well I never at 10:57 AM on July 8, 2022 [2 favorites]
I wanted to understand more about this observation - this sounds like one of those "innocuous and enticing when summarized but look at the loudest proponents" challenges.
The Slate Star Codex crew opines EA is fundamentally conservative (this reads more like appropriation of the seductive marketing term though). You can Google it I'm not linking it.
A voice I'd trust, Timnit Gebru weighs in, her experience is with the most techbro moguls involved and is unimpressed.
One doesn't necessarily need a Movement to do good as best they can.
Allllll that said I'll go read the links now - is this a concept resilient enough to be rehabilitated? It sounds more like a marketing term like Objectivist or Sapiosexual - maybe starting out with an uncharged meaning but quickly picking up layers of untested hypocrisy and panglossian lethargy.
posted by abulafa at 11:32 AM on July 8, 2022 [1 favorite]
The Slate Star Codex crew opines EA is fundamentally conservative (this reads more like appropriation of the seductive marketing term though). You can Google it I'm not linking it.
A voice I'd trust, Timnit Gebru weighs in, her experience is with the most techbro moguls involved and is unimpressed.
One doesn't necessarily need a Movement to do good as best they can.
Allllll that said I'll go read the links now - is this a concept resilient enough to be rehabilitated? It sounds more like a marketing term like Objectivist or Sapiosexual - maybe starting out with an uncharged meaning but quickly picking up layers of untested hypocrisy and panglossian lethargy.
posted by abulafa at 11:32 AM on July 8, 2022 [1 favorite]
I listened to the Ezra Klein interview with one of the effective altruism leaders a while ago; it's worth the time to listen.
EA is weird because it's a big tent. It includes both people who think that a large portion of every dollar not spent on bed nets is wasted... And a contingent of libertarian Reddit folks who are way too excited to be the first ones eaten by AGI.
The fundamental problem is that taking EA seriously requires thinking about the discount rate for future lives (a hard philosophical choice) multiplied by the probability of any particular intervention having an effect on a future historical branch you've chosen as your pet project (which is effectively impossible to guess well at). So anyone in EA who is overly concerned with distant future problems is going to be mostly a bullshit artist, since these things are impossible to estimate. This suggests to me that the right approach really is a large discount rate for preventing most future calamities, since we don't really have any idea whether they will really come to pass, or whether anything we can do today will be effective at stopping or altering the outcome of the calamity. The devil comes when you start actually arguing about the probability of Killer AGI (hard to say) vs a Killer Asteroid (probability 1 on a long enough time horizon)...
Anyway, my takeaway was that there really are two EA movements, one which is pretty serious in what it's trying to do, and another made up of jackasses on Reddit with boners for Slate Star Codex. YMMV.
posted by kaibutsu at 11:46 AM on July 8, 2022 [5 favorites]
EA is weird because it's a big tent. It includes both people who think that a large portion of every dollar not spent on bed nets is wasted... And a contingent of libertarian Reddit folks who are way too excited to be the first ones eaten by AGI.
The fundamental problem is that taking EA seriously requires thinking about the discount rate for future lives (a hard philosophical choice) multiplied by the probability of any particular intervention having an effect on a future historical branch you've chosen as your pet project (which is effectively impossible to guess well at). So anyone in EA who is overly concerned with distant future problems is going to be mostly a bullshit artist, since these things are impossible to estimate. This suggests to me that the right approach really is a large discount rate for preventing most future calamities, since we don't really have any idea whether they will really come to pass, or whether anything we can do today will be effective at stopping or altering the outcome of the calamity. The devil comes when you start actually arguing about the probability of Killer AGI (hard to say) vs a Killer Asteroid (probability 1 on a long enough time horizon)...
Anyway, my takeaway was that there really are two EA movements, one which is pretty serious in what it's trying to do, and another made up of jackasses on Reddit with boners for Slate Star Codex. YMMV.
posted by kaibutsu at 11:46 AM on July 8, 2022 [5 favorites]
Thanks kaibatsu - that was a more succinct way to phrase it than my meandering.
Aside from TFA are there any recommendations you'd suggest for non-reddit-rationalist ummm ... Forums? I guess?
posted by abulafa at 12:06 PM on July 8, 2022
Aside from TFA are there any recommendations you'd suggest for non-reddit-rationalist ummm ... Forums? I guess?
posted by abulafa at 12:06 PM on July 8, 2022
If these billionaires were really serious about effective altruism, they wouldn’t be billionaires.
posted by parm at 1:41 PM on July 8, 2022 [3 favorites]
posted by parm at 1:41 PM on July 8, 2022 [3 favorites]
I agree with you, parm, but I think the reason "effective altruism" is appealing to these types is because it allows them to construct a narrative in which resources are most effectively allocated when they're put in the hands of the smartest, best people (themselves).
posted by crosley at 2:00 PM on July 8, 2022 [6 favorites]
posted by crosley at 2:00 PM on July 8, 2022 [6 favorites]
I'm reminded of the series finale of The West Wing, where CJ's considering her career options after leaving the White House, and one of her choices is to help do PR for a charity whose work in Africa is simply building roads. As it was explained to her, a non-profit wanted to pave and maintain roads for safer transportation and shipping in regions that had never had them before, but since there's nothing trendy or photogenic about the problem, they had a tough time getting traction (so to speak).
posted by The Pluto Gangsta at 2:15 PM on July 8, 2022 [1 favorite]
posted by The Pluto Gangsta at 2:15 PM on July 8, 2022 [1 favorite]
For reasonable effective altruism, it's also worth talking a look at Give Directly, which advocates for just giving direct cash payments to the poor and letting them use it as they see fit. This is a very effective strategy, and completely circumvents the perform of funders needing to guess at people's problems.
Give Directly also pioneered randomized controlled trials (RCTs) for philanthropic work, which is vitality important to understanding where to invest for impact. (And, let us also note in passing, that you can't do RCTs for hypothetical future catastrophes, putting this completely outside the Jackasses on Reddit conversation.)
I get the desire to bag on billionaires, but please take a moment to consider that the problem of how to do the most good with a given amount of money is a real problem worth solving. And it's a real question at pretty much every level of wealth/income. For people making a decent amount of money in professional work, it's really something like an extra job to figure out how to give money well. Collecting data on what actually works is this important work.
And if you're worried about the ego trips of the rich, let's also consider that government and ngo initiatives often have a lot more to do with appeal to voters and funders than actual effectiveness. (For example, the story about road building up above.) Unsexy but vital work can have a hard time getting funded exactly because the incentives are all screwed up, wherever you look in the philanthropic space. Collecting data is a way to get around some of the misaligned incentive problems.
Now, there's a real critique available here about whether you can actually solve everything with RCTs and data collection. The world is never going to be 'legible' enough for this strategy to work in every case. You can also argue that this live of thinking will tend towards problems that actually can be solved with money (bed nets) and away from those with more complicated social dimensions (racism). (And there's the rub... should we spend money to since problems that can't be solved with money? And how much?) But these are the critiques that are actually worth engaging with and thinking about - they're a hell of a lot more interesting than Roko's Basilisk.
posted by kaibutsu at 4:27 PM on July 8, 2022 [8 favorites]
Give Directly also pioneered randomized controlled trials (RCTs) for philanthropic work, which is vitality important to understanding where to invest for impact. (And, let us also note in passing, that you can't do RCTs for hypothetical future catastrophes, putting this completely outside the Jackasses on Reddit conversation.)
I get the desire to bag on billionaires, but please take a moment to consider that the problem of how to do the most good with a given amount of money is a real problem worth solving. And it's a real question at pretty much every level of wealth/income. For people making a decent amount of money in professional work, it's really something like an extra job to figure out how to give money well. Collecting data on what actually works is this important work.
And if you're worried about the ego trips of the rich, let's also consider that government and ngo initiatives often have a lot more to do with appeal to voters and funders than actual effectiveness. (For example, the story about road building up above.) Unsexy but vital work can have a hard time getting funded exactly because the incentives are all screwed up, wherever you look in the philanthropic space. Collecting data is a way to get around some of the misaligned incentive problems.
Now, there's a real critique available here about whether you can actually solve everything with RCTs and data collection. The world is never going to be 'legible' enough for this strategy to work in every case. You can also argue that this live of thinking will tend towards problems that actually can be solved with money (bed nets) and away from those with more complicated social dimensions (racism). (And there's the rub... should we spend money to since problems that can't be solved with money? And how much?) But these are the critiques that are actually worth engaging with and thinking about - they're a hell of a lot more interesting than Roko's Basilisk.
posted by kaibutsu at 4:27 PM on July 8, 2022 [8 favorites]
Did Lex never have Clark read Marx? Or Aaron Bastani's "Fully-Automated Luxury Space Communism"?
posted by k3ninho at 12:51 AM on July 10, 2022
posted by k3ninho at 12:51 AM on July 10, 2022
Love the Superman story, branewane. Thanks! (Haven't read the other one yet, but I will.)
posted by straight at 12:50 PM on July 10, 2022 [1 favorite]
posted by straight at 12:50 PM on July 10, 2022 [1 favorite]
It has already come perilously close to stumbling over the intrinsic problem at the extreme end of "effective altruism" - that the logical dictate of maximally effective altruism would be to identify the Most Effective Altruist(s) and then eliminate choice/free will from everyone else.
Unless you can quantify the value of free will, any "effective altruism" calculations become inherently suspect the moment you start influencing other peoples' behavior.
Some people would say that democracy isn't the best form of government, it's just the least bad form of government.
The problem with people who point that out is that they believe at some level that they themselves (or maybe somebody better than them like God or Benevolent AI or Dolly Parton or Elon Musk) could rule better than democracy.
You're arguing that free will might be a good in itself rather than just a means to prevent tyranny and I wish to subscribe to your newsletter.
posted by straight at 12:58 PM on July 10, 2022 [1 favorite]
Unless you can quantify the value of free will, any "effective altruism" calculations become inherently suspect the moment you start influencing other peoples' behavior.
Some people would say that democracy isn't the best form of government, it's just the least bad form of government.
The problem with people who point that out is that they believe at some level that they themselves (or maybe somebody better than them like God or Benevolent AI or Dolly Parton or Elon Musk) could rule better than democracy.
You're arguing that free will might be a good in itself rather than just a means to prevent tyranny and I wish to subscribe to your newsletter.
posted by straight at 12:58 PM on July 10, 2022 [1 favorite]
A voice I'd trust, Timnit Gebru weighs in, her experience is with the most techbro moguls involved and is unimpressed.
Gebru links to this article in abulafa's link, which seems like good critical EA overview:
The Dangerous Ideas of “Longtermism” and “Existential Risk”
posted by straight at 1:15 PM on July 10, 2022 [1 favorite]
Gebru links to this article in abulafa's link, which seems like good critical EA overview:
The Dangerous Ideas of “Longtermism” and “Existential Risk”
posted by straight at 1:15 PM on July 10, 2022 [1 favorite]
For reasonable effective altruism, it's also worth talking a look at Give Directly, which advocates for just giving direct cash payments to the poor and letting them use it as they see fit. This is a very effective strategy, and completely circumvents the perform of funders needing to guess at people's problems.
Most proponents of EA would normally be all over the idea that the decisions of individual consumers will always be a more efficient way of allocating resources than some central planner trying to omnisciently select the "best" way to do it.
And surely the actual truth is somewhere in the middle: There's limits to how well you can predict the consequences of your actions--especially in the longer term--so individual decisions are often better than centrally planned ones. But we should pay attention, study, and act because many times we do have enough evidence to say that collective action is better than anything individuals can do separately.
posted by straight at 1:27 PM on July 10, 2022
Most proponents of EA would normally be all over the idea that the decisions of individual consumers will always be a more efficient way of allocating resources than some central planner trying to omnisciently select the "best" way to do it.
And surely the actual truth is somewhere in the middle: There's limits to how well you can predict the consequences of your actions--especially in the longer term--so individual decisions are often better than centrally planned ones. But we should pay attention, study, and act because many times we do have enough evidence to say that collective action is better than anything individuals can do separately.
posted by straight at 1:27 PM on July 10, 2022
I almost didn't notice the most important part of the Superman story. The part at least as fantastical--and which does even more of the heavy lifting than Clark--is that army of accountants, scientists, ethicists and other smart people whose superpower is collecting accurate data, analyzing it properly, and making reliable analyses and predictions about the consequences or potential consequences of Superman's actions. (Which I think is kind of what kaibutsu was saying.)
That's the part the EA people think is potentially non-fiction.
posted by straight at 2:41 PM on July 11, 2022 [1 favorite]
That's the part the EA people think is potentially non-fiction.
posted by straight at 2:41 PM on July 11, 2022 [1 favorite]
it's also worth talking a look at Give Directly
I really want to like GiveDirectly. I even DO like a lot about GiveDirectly, and donate to them occasionally. But there's something about it that just feels so very... male to me. Like, there's nothing inherently wrong with being male or anything. I'm not even sure what exactly I mean by that. It's bothered me for years, but it's also been a few years since I've looked at them, have they gotten better recently?
I just checked, and their overall staff looks like it is almost exactly gender-balanced, but the leadership looks like it's like 85% male.
Anyway, so I've been wanting something that's LIKE Give Directly, but NOT. Is it out there?
posted by aniola at 10:01 AM on July 14, 2022
I really want to like GiveDirectly. I even DO like a lot about GiveDirectly, and donate to them occasionally. But there's something about it that just feels so very... male to me. Like, there's nothing inherently wrong with being male or anything. I'm not even sure what exactly I mean by that. It's bothered me for years, but it's also been a few years since I've looked at them, have they gotten better recently?
I just checked, and their overall staff looks like it is almost exactly gender-balanced, but the leadership looks like it's like 85% male.
Anyway, so I've been wanting something that's LIKE Give Directly, but NOT. Is it out there?
posted by aniola at 10:01 AM on July 14, 2022
« Older 300 days, one house | We’re battling a sleep loss epidemic. California... Newer »
This thread has been archived and is closed to new comments
posted by lalochezia at 7:16 AM on July 8, 2022 [4 favorites]