3 ms·
What evidence do you want? The majority of the staff have left. Tangible fact. The site is having outages. Tangible fact. Mastodon usage is significantly up.
by wokwokwok 4y ago
What evidence do you want?
The majority of the staff have left. Tangible fact.
The site is having outages. Tangible fact.
Mastodon usage is significantly up. Tangible fact.
People make vague hand wavey gestures about Twitter having better than ever engagement or usage or whatever… but there are no stats to back it up, so who knows?
Come on! Compared to 6 months ago the service is empirically worse.
It’s unbelievable to me someone can stand here and go “where the proof dude??” with what’s happening.
Is it going to be gone in a year? Who knows? No one’s knows the future.
Is it reasonable to say, given the recent state of the platform is, the probability of the platform being gone in a year is now significantly greater than it was six months ago?
Probably, I would say, yes. The empirical evidence suggests so.
I’m not saying there’s a 100% chance of that; no one is saying that; but it’s greater than it was.
You have to really make an effort to believe otherwise.
- MiguelX413 4y agoYou're also omitting the various core feature breakages as Elon cuts infrastructure in an attempt to recoup his investment. 2fa users were locked out of their accounts because he turned off a bunch of microservices.
- raydev 4y agoA week has passed and it seems like they’ve fixed it up. Similarly for the last week I’ve had increasing issues loading tweet details and it peaked yesterday, but today all seems fine. We’ll have to wait months before we can truly know if this diminished and demoralized crew is able to keep the site running with minimal issues, but for now they do seem to be (slowly) recovering.
- luckylion 4y ago> I’m not saying there’s a 100% chance of that; no one is saying that; but it’s greater than it was. Okay, but going from 0% to 0.0001% is also "greater than it was". What are you saying? Is it 90%, 50%, 20%, 10%, 1%?
- wokwokwok 4y agoIf you want odds visit a betting site. Twitter has graduated from “don’t believe it’s plausible that Twitter will fail” to “anecdotally things seem to going badly” to “empirical evidence of service issues and mismanagement”. I don’t care what you believe, those are the facts, right now. I think then next month will give us a reasonable ability to forecast as we get spikes in usage during the World Cup and we see if the service goes down and if people keep using it. Until then… who knows? More than 0.001%; that’s just random variance. Don’t be ridiculous. I’m not waving my hands and saying stupid meaningless shit like 0.0001 > 0 so there’s some meaningless increase in the chance Twitter will fail. I’m saying, there is a tangible increase in the possibility of failure, and already a concrete notable decrease in value of the entire platform.
- raydev 4y agoI spend hours a day on Twitter, and today has been by far the best in the few weeks. Things deteriorated temporarily but I’m missing where the concrete decrease in value is.
- ModernMech 4y agoTwitter is still having trouble serving Tweets. I'm getting "Tweet unavailable" often enough that's it's becoming a noticeable problem.
- philipov 4y ago0.0001% == 0% The only operable probabilities are "Never", "Sometimes", and "Always". Since Musk took over, the chance of imminent nonrecoverable catastrophe has gone from "No" to "Maybe".
- luckylion 4y ago"No", "Maybe", "Probably", "Yes"? That would put the "Maybe" at ~25% for "imminent nonrecoverable catastrophe"?