Elon Musk's latest company: Pravda

And there doesn't seem a good existing mechanism to deal with that. Individual journalists with an axe to grind, which if you don't follow their work, you'll not know to take with a grain of salt. Large misinformation campaigns, which the press happily falls for, and afterwards doesn't report on or learn from.

When the Zuma defense satellite deployment failed, it was very quickly clear that Northrop Grumman was at fault. But they ran a congress and press campaign, pointing at SpaceX, generating a press "frenzy" (WSJ's word). It was bizarre to watch the press, from NYTimes front page on down, simply getting the story wrong. For days. Oh, the NYTimes got "facts" right - as in "people are blaming SpaceX" - but with little hint that that was bogus. The campaign was successful in avoiding a public story of "Northrop just burned up a billion dollars!" And afterward? No negative feedback for any of the participants. Some quiet little corrections, some of which required reading between the lines to even recognize as such. No articles about the reporting. No backlash. No postmortems. No sign that next week any of them will do any better.

And Tesla is apparently the largest short on the US stock market. With the press coverage you'd expect from that.

So perhaps Musk encountered a societal failure mode, didn't see anyone else successfully addressing it, and is doing a trial balloon on one possible approach. Hopefully this will at least raise the visibility of the issue.

At best, it's useless, at worst it contributes to further suppressing any news story that challenges widely accepted narratives.

There's a clear need to crowdsource judgment. The judgment could be much more expressive than "this is true". A hypergraph database would allow collection of arbitrary metadata -- "this helped me understand plate tectonics", "this contradicts that", etc. And it would enable more targeted search: not just "show me true things", but, say, "show me articles about voting that are highly discussed among the economic faculty at Michigan State".

I've written about this here: https://github.com/synchrony/smsn-why/blob/master/noise-vs-s...

I also wrote a language -- very, very close to natural language, but machine-readable -- designed to facilitate such higher-order queries: https://github.com/JeffreyBenjaminBrown/digraphs-with-text/b...