Twitter is hundreds of millions of authenticated messages a day, delivered with an expectation of instantaneousness and simultaneity, to hundreds of millions of clients.

There are images and video embeds and URL previews on tweets. There are fraudulent and abusive users who are trying to spam the user base. There are users whose threat model includes ‘will be attacked by the concerted cyber-offensive capabilities of sophisticated nation states’, as well as ‘will be cyber-bullied by the collective force of 8chan’, or ‘will be cyber-stalked by legions of k-pop stans’. Not clear which of those is more dangerous.

The SEC considers statements made on Twitter by corporate officers and company accounts to be material disclosures. Tweets are embedded in New York Times and bbc news articles. Courts take Twitter messages as evidence. Law enforcement organizations throughout the world have frequent interest in finding out what activity users have had on the platform.

And on top of that, it’s trying to run an ad business that takes that firehose of craziness and turns it into a marketable stream of attention that can be sold to advertisers.

Don’t downplay what Twitter does as ‘just moving 280 character messages’. It’s way, way more than that.