(This was a quick post, written in around 30 min. It was originally posted on Facebook, where it generated some good discussion.)
I really wish EA had better internal communications.
If I wanted to make a blog post / message / recording accessible to a "large subset of effective altruist professionals", I'm not sure how I'd do that.
I don't think we yet have:
- One accepted chat system
- An internal blogging system
- Any internal email lists (for a very wide net of EA professionals)
It's nice to encourage people to communicate publicly, but there's a lot of communication that's really not meant for that.
Generally, the existing options are:
- Post to your internal org slack/emails (note: many EA orgs are tiny)
- Share with people in your office
- Post to one of a few domain-specific and idiosyncratic Slacks/Discords
- Post publicly, for everyone to see
I think the SBF situation might have shown some substantial vulnerability here. It was a crisis where public statements were taken as serious legal statements. This meant that EA leadership essentially didn’t have a real method of communicating with most EAs.
I feel like much of EA is a lot like one big org that tries really hard not to be one big org. This gives us some advantages of being decentralized, but we are missing a lot of the advantages of centralization. If "Professional EAs" were looked at as one large org, I'd expect that we'd look fairly amateur, compared to other sizeable organizations.
A very simple way to make progress on internal communications is to separate the issue into a few clusters, and then attack each one separately.
- Access/Onboarding/Offboarding
Make official lists that cover "professional/trusted members". You could start with simple criteria like "works at an org funded by an EA funder" or "went to 2+ EAGs". - Negotiation and Moderation
"EA Professionals" might basically be an "enterprise", and need "enterprise tools". These often are expensive and require negotiation. - A Responsible Individual
My preference would be that we find someone who did a good job at this sort of thing in other sizeable companies and try to get them to do it here.
I bet with $200k/year for the talent, plus maybe $200k-$1k/year, we could have a decent setup, assuming we could find good enough talent. That said, this would definitely be work to establish, so I wouldn't expect anything soon.
A modest step might be to allow a forum post author to restrict visibility of their post to those who enter a forum username/password (i.e., the post could not be indexed by search engines). You could go a step further and limit access to usernames that had been vetted in some fashion, but that would involve some time commitment and uncertain benefit. Perhaps during situations like FTX, you could allow posts to be limited to usernames created before the crisis happened . . . but that might give people a false sense of security as the odds of any mass communication leaking are non-trivial.
One caveat: any communication can become a "serious legal statement" if it's not legally protected from disclosure in discovery. And although the topic of record retention is far more complex than this sentence (or even Molly's backgrounder), as soon as a sufficient copy of the information exists, there may be an obligation to preserve it if relevant litigation is forseen. Any technology that allows recipients to view the information at their convenience will probably involve creating a sufficient copy. So this would not have helped with the FTX situation. I'm guessing that the people involved in that situation have been advised to rely on their telephones considerably.