I'm hard-pressed to think of any changes that, at least in my experience, have qualitatively improved the experience, but the bleeding appears to have slowed, app reliability is reduced but not apparently an existential threat, much-feared competitors have had little effect on market share (particularly Meta and TikTok's dismal attempts at capturing Twitter users), and, while Bluesky has picked up a great many high-profile posters, Twitter has replaced that volume to some degree by becoming what's essentially a more socially-acceptable Gab.
The only question is whether, given the predictable failure of Musk's attempt to become a subscription-based rather than ad-based business, its cash flow can support a balance sheet that's loaded up with debt based on an unsupportable valuation. Twixtter may be stable, but "stable" doesn't necessarily mean "surviving."
I suppose I could make more effort to mine other sources like Medium or decentralized Twitter clones, but haven't yet done so.
I'll put up with the dreariness and annoyances of X for now, but the instant I find a better alternative, I'm gone.
bsky's small in absolute terms (about 1.5mm users), and lacks the global audience reach that Twitter still has, but it's got better "stickiness" than, say, Threads. This, oddly, makes it a bit more of a threat than Twitter's large-scale competitors, because Twitter has always been reliant on a small number of posters generating the bulk of its widely-shared content, and a small number of high-usage users generating the bulk of its ad views. Losing a relatively small number of the most dedicated Twitter users hurts more than a larger number of low-usage/low-ARPU users.
Where bsky goes from here is an open question (complex systems are notoriously hard to predict, and as they add more users without concomitant improvements in their other systems -- think T&S -- things could go pretty badly pretty quickly), but by damaging Twitter's content generation engine they are probably doing more damage to Twitter's P&L than Threads or Lemon8 ever did.
twitter.com - 5.8B visits (~ user sessions)
threads.net - 45.4M visits
bsky.app - 21.6M visits
mastodon.social - 4.5M visits (caveat: this is just one of instances, though one of the biggest ones)
This reads oddly because both of those have considerably more users than Twitter - Meta actually has four services which are larger:
https://www.statista.com/statistics/272014/global-social-net...
I call it PP-kat (platform previously known as twitter).
I'm not sure that "stable" is the correct term, as Elon Musk's Twitter is dealing with lawsuits because the company is so screwed they can't even afford to pay for their offices.
MAU might be flat, but I can almost guarantee the bot or inauthentic traffic is what are keeping those numbers flat. Heck, a non zero amount of that inauthentic traffic has paid for twitter for purposes of spam or propaganda.
Protip: you can generally tell a sites actual user base by the quality of ads and their ad revenue. They are running non brand safe ads and reports are their ad revenue is a fraction of pre-musk levels. I half expect them to start running straight up taboola ads at some point.
Obviously the so-called 'exodus' never happened and their top predictions have failed to come true [0]. But it just shows that even most HNers at the time fell for the media exaggerations because it was the click-bait trend. Happened with the so-called collapse of Meta and once again with Twitter / X.
The deaths of both Meta and Twitter / X, have been greatly exaggerated. It is time to admit that such predictions like this [1] and this [2] aged extremely poorly 1 year on.
[0] https://news.ycombinator.com/item?id=37295543
[1] https://news.ycombinator.com/item?id=36580669
[2] https://www.technologyreview.com/2022/11/08/1062886/heres-ho...
And the X rebrand is the same.
