There are really a lot of EA grants for community-building, but very few for the acts of community destroying, even though there are significant limits to how large a healthy community can grow, and also some people smell bad. 

In general we should not assume that one big community is better than several little communities. While there are many redundancies from different groups attempting to meet the same goal, there are also overheads to coordinating large groups, and the scaling rate of these are plausibly much worse. Studies show that several measures of the efficiency and productivity of workers can decrease with firm size. With research groups, we find a doubling of the number of researchers less than doubles all measures of impact and we should expect the same to be true of communities. Yes, a larger group can do more - but can it do more than two smaller groups? This research would suggest not, at least beyond some size limit.

In-group co-ordinating is difficult at the best of times, but is even more so when a community fetishises both intellectual novelty and pedantry. This applies both at the object level and the meta-level. Objectively, people are constantly creating the most unhinged of hypotheses and the effort of discriminating between the odd-but-plausible and raving insanity is occupying far more of the EA cultural hivemind than is optimal. At the meta-level, EAs now spend more time discussing the social structure of EA and the design of its forum than they do actually improving the world. Look, you’re reading an article about it right now. Why? Do you really think reading this article is worth three minutes of your salary divided by $5000 of a life? 

Most people who hate EAs spend their time making the exact same criticisms as EAs make of themselves except much more angrily, therefore getting more people to read it. This means that anti-EAs are co-ordinating with EAs better than EAs co-ordinate with themselves, and without the same overheads. It would greatly benefit the world if EAs as a community would stop pretending that everyone who likes cost-benefit analysis needs to hang out together and read the same stuff. 

In this vein, I am proud to announce that our initial experiments in community-destruction within EA via several Sort by Controversial statements by senior figures has shown tremendous results. We anticipate more money being available for future work as everyone realises just how terrible other people are.




Sorted by Click to highlight new comments since:

Here's a practical idea for community destruction - we can encourage EAs to all live together, work together, date each other, and socialize exclusively with other EAs. My theory of changes is that when interpersonal conflicts inevitably arise, they'll be amplified by this interconnected structure.

There's a lot of opportunity for leverage here. If you set up the structure correctly, the community will self-destruct on its own!

I think the sacrifices made by senior figures to make this possible have been underappreciated. They have done--and continue to do--tremendously counterintuitive and self-destructive actions in order to accomplish this level of fragmentation.

I agree. Musk showed us that you can cut head count at a tech company by 90 percent and everything turns out fine. If we truly care about cost effectiveness we should do the same for tech-focused EA orgs like LessWrong, EA Forum, Ought, MIRI, ARC and Redwood. Anthropic comes to mind as another example.

Ideally all EAs should fear for their jobs - this fear will encourage them to work harder. It's well known that EA jobs are competitive with many applicants. If they can't handle the pressure, we can replace them with others who can.

The era of free-spending EA is over, and it's time to get back to our roots. We need to once again maximize impact per dollar, not just throw money at promising 17-year-olds. To that end, we ought to start firing unproductive community members. As we know from the Pareto principle, 20% of people have 80% of the impact, so we can easily 5x our cost-effectiveness by laying off 4/5ths of our staff. While we're at it, we should sell off all that wasteful office real estate OpenPhil and Constellation have acquired, in order to focus on the most effective assets, like Wytham Abbey.

I can appreciate playful satire as much as anyone, but the tone and content of this post might be quite puzzling to anyone who's not a native English speaker (or who's a slightly literal-minded & systematizing). 

I think for the sake of EA Forum's cross-cultural inclusivity, and out of respect for people who use English as a second language, we should try to avoid satirical posts unless they're clearly labelled as such.

PS this applies especially to any 'April Fools' posts, insofar as April's Fools is celebrated in only about 11 countries out of 200

This article isn't an exclusive list of the countries that celebrate it, merely a list of how it's celebrated in 11 noteworthy nations. It's also celebrated in Iran, China, Germany...

I hereby request funding for more overwrought posts about the community's social life, as they are a cost-effective way to do this.

More from Robin
Curated and popular this week
Relevant opportunities