Last year, I felt it was suboptimal for tech platforms to outsource their governance issues, because giving permit/deny rights on a material portion of all human utterances to parties which rhyme with governance seemed unwise.

This year, it feels *extremely* suboptimal.

Thought partially inspired by the fact that, if you were to search Twitter for a particular blog post about a particular government's handling of a situation of pressing public concern, Twitter helpfully directs you to reliable information by the government agency in question.

"Pics or it didn't happen."

Pretty sure it happened; handily there's a way to demonstrate that without pics.

b3b7e835d16302c33b30684ca5b512f62a4db82c8e91632eb47bf2702dafdb76a5c321f4ca9d2b5f0da5d92a77a7712028f30302f41474a505fd571b301bbe45

This feeds another theme discussed recently.

No engineer and project manager got together in SFBA and said "OK, here's the Excel spreadsheet of all tuples of (government, issue) on which we will affirmatively endorse the government's perspective and de-prioritize others."

There are people who believe that that happened, that "the elite" work hand-in-glove with each other to stagemanage world opinion.

Those people are very miscalibrated.

What actually happened was simply "There is a team which had a meeting on how to reduce disinformation on the platform. They created some tooling and handed it off to a team whose name sounds something like Trust and Safety Ops."

That team has a runbook for issues of social concern. The runbook specifies a rubric for identifying reliable sources. That rubric includes, among other options, official pronouncements from democratic governments.

So when an issue of social concern came up, the team followed the runbook, added some keywords to the tool, checked on the rubric for who reliable sources w/r/t those keywords would be, Googled "Name of [agency type] in [democratic nation]", and there you go.