This issue isn’t so much that Musk didn’t ask for “as many 9’s” in confidence that turning off microservices wouldn’t break things as he should have, but that he apparently didn’t ask for any 9’s. He just had them all turned off without discussion. If anything, any 9’s involved would have been confidence that things would break, not that they wouldn’t. Like, the two-factor authentication ought to have been super-obvious on any sort of real evaluation of dependencies, for example. It’s true that none of this is particularly high stakes because it’s just Twitter, but “just Twitter” is what he just paid tens of billions of dollars for. Musk’s, shall we say, somewhat erratic leadership in the past couple weeks has already turned a lot of advertisers away, and those advertisers represent the vast majority of the actual monetary value of Twitter at the time of purchase. The current trends are not promising.
The entire premise of Musk’s initial offer to buy Twitter was that it was anti-free speech, and needed to be freed from woke censorship of “controversial” (i.e. racism and other bigotry, antivax conspiracy bullshit, etc) views. Anyone with passing familiarity to the internet knows that no moderation at all leads to toxic, 4-chan-esque hellscapes and worse, and given half a chance all sorts of bad actors will use your platform to engage in actual criminal behaviour like distributing child porn. Given that Musk is at least passingly familiar with the internet, he either knew that or should have known it, and therefore should have known that to achieve his goal - transforming Twitter into a platform that was free of “woke censorship” would need a new moderation regime to replace the old one, and should have known that the new regime needed to be designed and rolled out before turfing the old one, instead of throwing everything out and then making shit up as he goes along.
It’s all very well to try to implement the break stuff to find out what doesn’t work and iterate fast model that’s been so successful at SpaceX, but you don’t see SpaceX trying out C4 as an ablative heat shield material because hey, who knows, maybe subjecting high explosives to insane pressures and temperatures will work. Explosive Reactive Armor on tanks is a thing, after all!
I started out on alt.fan.cecil-adams in the way early days. It was pretty good at the time. Then the first version of The Straight Dope online was rolled out as an AOL site. Then it migrated to the web. Then a few years later I returned to the usenet group to see what was going on. Traffic was down, and of course it was unmoderated. (Most of the posts were pretty random and not really in the spirit of fighting ignorance.) I posted, asking why people were still using the usenet group when the web site was so much easier to use. I was immediately called a “fuck stick.” It was also noted that anyone using Google Groups to post instead of a newsreader application was an idiot.
You don’t have to believe what he’s said, but if you don’t, you probably shouldn’t believe anything else, making the whole discussion moot.
I don’t disagree that this will pose a serious problem, and think Spice_Weasel’s link about “speed-running the content moderation curve” is a good one. Musk has not said there will be no moderation–on the contrary, he’s emphasized at the least that they will follow local law, which basically compels some degree of moderation. It wouldn’t shock me if they end up re-hiring some of their contractors.
I think a verification scheme that works would be a great help. Yes, the blue check thing we saw so far was not even half-assed, but it doesn’t seem like an impossible problem. Bots have a hugely distorting influence.
There are reports that 80% of employees, in at least some groups, are choosing to take the severance rather than stay and be exceptional. Mac-mini in a closet or not, Twitter must need at least some employees to stay operating.
I suspect that a large percentage of employees are calling his bluff – that is, not signing the commitment to give up their souls to the company, but not resigning, either, and just waiting to see what Musk will do. And my guess is that, as with many other of his hare-brained decisions that he back-pedaled on, he won’t do anything. Still, in addition to his earlier mass firings, a large number of employees has reportedly already quit. It all seems like an excellent strategy if your objective is to get rid of your best people and be left with the least marketable ones.
This sounds much like part of the Peter Principle. If someone reaches the top of a hierarchy without reaching his level of incompetence, he moves to another hierarchy where he can achieve it.
Musk is like the dog chasing a car who doesn’t know what to do with it once he catches it.
He’s using his experience running other businesses without understanding Twitter is different.
