1305
Meals
(slrpnk.net)
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 13 Jul 2025
1305 points (98.3% liked)
Microblog Memes
8527 readers
3032 users here now
A place to share screenshots of Microblog posts, whether from Mastodon, tumblr, ~~Twitter~~ X, KBin, Threads or elsewhere.
Created as an evolution of White People Twitter and other tweet-capture subreddits.
Rules:
- Please put at least one word relevant to the post in the post title.
- Be nice.
- No advertising, brand promotion or guerilla marketing.
- Posters are encouraged to link to the toot or tweet etc in the description of posts.
Related communities:
founded 2 years ago
MODERATORS
Bayesianism is about reconciling your squishy priors with hard math. If there's a round peg and a square hole, the square hole is Frequentism.
I don't understand your point about Bill Gates. You're saying he had one plan, but then found another plan worked better. What does this have to do with EA? Givewell isn't an armchair-thinktank, it does pretty solid research and analysis comparing the effectiveness of real-world charities that already exist.
The loss of USAID was really bad. Here's EA Scott Alexander talking about just how bad the scaling back of USAID is. If there were self-proclaimed EA's involved with villifying USAID, that is ironic indeed.
Well I agree. I don't have executive fiat. I'd like to increase the amount of popular buy-in. This is one of the main reasons I post on Lemmy. However, that socialism requires concensus whereas charity does not -- this is exactly Ozy Brennan's point. So I think that we don't disagree at all. Ozy's observation is that EA charity organizations generally focus on the opposite of buy-in; they look for areas of neglect -- places where big strides can be made because other people aren't working hard on those problems yet. Perhaps because they sound strange. Like electrocuting shrimp so they don't feel pain when they die in factory farms (yes this is a real charity).
Frequentism won't work with a contained set of inputs. But now we're getting into Abstract Algebra rather than probability.
I'm saying he kept coming at the problem dead on without exploring the second and third order consequences of did policies.
Lots of maths up front but the models were shit. The end result was a reactionary mess precisely because Gates and his lackeys didn't care about the popular politics of their policies.
The observation that mosquitoe nets and medical interventions have a long term benefit isn't a problem on its face. But, again, Ozy is attacking a complex problem of supply chains and sustainable development from a very boiled down "do things that look good on my spreadsheet" as the "Effective" solution.
When these plans fall apart, because the proponents fail to account for second order problems, they denounce everyone else as another problem they need to strike head on, rather than considering where they went wrong.
Case in point
Because the focus was on disease and food security was discounted as a less pressing problem, the primary tool for mitigating disease spread became an environmental catastrophe.
I'm legitimately curious how Abstract Algebra relates here. I thought that was all about group theory and such.
Mkay, but, this doesn't mean math is wrong. It means actual research is needed. Trials and case studies and comparative analysis and so on. Fortunately, that's exactly what givewell does. You can criticize Gates for not predicting second and third order consequences, but I'd argue the only thing we can do in the world where the higher-order consequences are somewhat predictable in advance is preserving the status quo.
The misuse of mosquito nets for fishing is bad, yes -- and depressingly ironic -- but you should check out the Against Malaria Foundation's response, where they say basically the misuse of malaria nets is not very widespread.
https://en.wikipedia.org/wiki/Group_theory
I got a degree in it, so I know a few things.
The application of a model to a set of data which fails to predict outcomes reliably is "Wrong Math".
The big problem with EAs is empirical. They don't deliver on their promises.
The distribution of nets had failed to yield the promised benefits. I site the misuse as a very prominent example of how EAs misjudge externalities, but its one data point in a much broader picture.
If you really want to drop the hammer on EAs - particularly chronic fraudsters like SBF and the Zuckerberg CZI - it is that they're fair more interested in self-enrichment than altruism in the basic sense.
I cite the mosquito netting distribution effects as a very straightforward calculation error, because it is at least superficially a sincere effort with lackluster results. But once you get under the tip of the iceberg, EAs are as riddled with con-artists and bullshitters as any Clinton Foundation or UN Food for Oil initiative.
That's the wages of unilateralism in a nutshell.
Please explain. I took a few courses in group theory, ring theory, etc., though I was never particularly good at it. How does it relate to probability?
Zuckerberg/CZI are not EA, and SBF was disowned by EA. It's not obvious to me that SBF was not interested in altruism, I think he was just catastrophically bad at it.
AMF stopped 20 million cases of malaria in 2023.
Sure, let me just dust off my notes from a decade ago.
It has to do with the available range of outputs given all available inputs. And the degree to which iterative actions can have a feedback effect.
But the math on this kind of thing gets hairy fast.
Zuckerberg hires from the community and its affiliates. Sarah Wynn-Williams being an excellent example.
SBF being disowned after he went broke is hardly a point in the movement's favor.
FFS, they printed this in November of 2023. Really getting out ahead of your skis, when you're a data driven organization that's making claims on total reduction in cases before the period is even closed.
To date, I can find no evidence of a 10% drop in malaria cases in any of the targeted countries between '22 and '23.
On the contrary, the WHO reports an 11M case rise from the prior year. Neither have we seen a plunge in cases in '24 or the Q1 & Q2 of '25.