On Wednesday night, a number of Twitter's core features stopped working for many users, including tweeting, retweeting, following, and direct messages.
When attempting any of these actions, users were presented with an error message informing them that they had reached their daily limit. The platform remained online, but in de facto read-only mode. Some famous users were affected, such as Steve Martin, who tweeted out his own error message, and Motherboard's own account was prevented from tweeting and retweeting. The outage lasted for hours but appeared to be resolved by Thursday morning.
Such an outage was widely predicted by IT professionals on Twitter after Musk took over the site by saddling it with a huge amount of debt and gutting its staff to cut costs, taking it down from around 7,000 employees to just over 1,000. There are only about 500 engineers left on staff.
The company under Musk's leadership has also been experimenting chaotically with new features in order to drive revenue, including changes to core elements of the platform such as its API. Cracks were already appearing before the outage; a bug this week let users post tweets with thousands of characters, for instance.
Fortune reported that Musk sent an email to Twitter staff during the outage that implicated new feature experimentation and server migration. "Please pause for now on new feature development in favor of maximizing system stability and robustness, especially with the Super Bowl coming up,” Musk wrote in one email.
In a follow-up, Musk wrote that Twitter “should also pause on transitioning away from Sacramento, consolidating Atlanta and reducing [Google Cloud Platform] usage until at least next week.” Twitter has data centers in Sacramento and Atlanta.
Musk's erratic actions—like banning open-source flight trackers after saying he wouldn't—and desperate search for new revenue sources have caused chaos at Twitter that is not often seen at a major tech platform. Finally, it appears that it all broke under the strain.