Twitter is Back. Shhh… Magic Dance!
X, the artist formerly known as Twitter, is having a moment. Which is a weird thing to write because the service is technically 20 years old. Still, it feels like it nearly died a few years ago, so perhaps this moment is best viewed as a resurrection.
But it's also not really a surge in usage, and certainly not on the business side – it's more just a vibes-based resurgence. For weeks now, people have been talking about coming back – Federico Viticci of MacStories wrote about his return a couple months ago, and Casey Newton of Platformer more recently. Both did so sheepishly because their stances are clearly that while they'd prefer not to use X, the realities of their businesses dictate that they should if not must.
Over the weekend, John Gruber of Daring Fireball gave a more full-throated endorsement of the X comeback:
I see more interesting posts on X than I do on Threads, Bluesky, and Mastodon combined. I wish that weren’t so but it is. Most people don’t want to think about politics so they don’t. Those people, when they want to participate in a Twitter-like social network, go to X.
And I can’t blame anyone. The only network that compares to X, size-wise, is Threads—but Threads’s algorithm is weirder than X’s now, and the content I see there is relatively boring. My Threads timeline is full of engagement bait without being engaging. And if you want to judge a platform by its owner, “I’m grossed out and/or appalled by Elon Musk so I’m going to Mark Zuckerberg’s joint” isn’t a way to avoid being grossed out and/or appalled by the owner. Bluesky and Mastodon are too small for most people to care about—and, worse, they are the networks where politics are, in my experience, unavoidable. You can have a discussion about cool new use cases and tools for AI on X and people just discuss the tools and use cases. On Bluesky and Mastodon it’s difficult to have such discussions without arguments from people who (a) believe all use of AI is unethical, and (b) will shove their viewpoints into your replies.
I never left X,1 but I did debate it a few times over the past few years. Certainly, there were times where I was using it far less. Moral issues aside, it just felt like the product and experience kept getting more and more shitty after the acquisition by Elon Musk. The algorithm felt weighted towards propaganda-esque nonsense (if not outright propaganda) while the infrastructure and scaffolding was clearly buckling under the weight of mass layoffs and employee brain transplants.
Musk, famously, was forced to buy the service after trying to back out and it just felt like he was hell-bent on bending it to his will, and if that meant destroying it, so be it. Again, it feels like that nearly happened. The business clearly flatlined, and Linda Yaccarino's strategy as CEO seemed to be simply "Yes". Should we do this? Yes. What about that? Yes. How about this? Yes.

Whereas Twitter used to be the key pipeline for content discovery, as the service tried to make itself into more of a destination, this vital role in the ecosystem got broken into a million little social pieces. It sucked.
Meanwhile, Meta set their sights firmly back on the space with Threads. When Mark Zuckerberg does this, he tends to win; eventually if not immediately. At the same time, Bluesky seemed to rise out of the Twitter ashes as an heir-apparent. And Mastodon was there to scoop up the more open-focused zealots.
But Threads was clearly built too much in the image of Instagram, including leveraging that network to start. This led to fast growth, but a sort of soulless product that just felt more like an ongoing growth-hacking experiment. While I think the focus on replies – and thus, "conversation" – was smart, they clearly dropped the ball on real time information in favor of their tried and true algorithms. So it was sort of just like an Instagram without pictures, rather than a Twitter replacement and X killer.
Every time there was some sort of breaking news you would see this in action. X would roar back to life with information flowing in, while Threads felt like a parallel universe focused on updates from three days ago.

Threads is far better at all this now. And the product itself is quite good – better than X in a number of ways. It feels like they're continuously shipping the features Twitter should have built years ago. As such, they may actually be bigger than X at the moment in terms of actual usage and user base – and if they're not in every metric yet, they will be soon. But it just hasn't really mattered. X maintains this hold over the zeitgeist. Both the memes and the means of information distribution.
Gruber mentions the Apple execs using X for promotion, but it's obviously far beyond that – even just recently. NVIDIA CEO Jensen Huang joined in June to start his push for open weight AI models (and push back against any idea of banning some of them). David Ellison signed up earlier this month to announce the completion of the Paramount/Warner Bros merger. Microsoft CEO Satya Nadella is busy trying to thought-lead left and right on the platform. And wait, who is that regularly using the service these days, decades after signing up? Mark Zuckerberg.
Yes, the owner of Threads has become an active poaster. Not quite to the level of his AI lieutenant Alexandr Wang, but yeah, he's back. When asked about this recently on Alex Heath's Sources podcast, Zuckerberg acknowledged that it was as simple as wanting to get the message out everywhere, and some people – the AI community in particular – is on X. Yes, it's the same basic rationale as Viticci and Newton.
Meanwhile, Bluesky and Mastodon have seemingly become even more insular and yes, extremely anti-AI, which is perhaps a problem if the technology really is the future of everything and not just some hype cycle that will pass.

At the same time, Musk finally figured out the right business model for X. Which is to merge it with a company that can actually bring in money at scale. At first, this was xAI, which rode the AI funding bonanza to quickly be valued more highly than Twitter ever was as a public company. And then when he merged both into SpaceX, he backed both into having an actual business model, even if X wasn't meaningfully contributing.
To be clear, I don't think anyone would be happy with X just burning money forever, but it's also relative. The amount xAI is burning must dwarf whatever the social network could ever hope to burn. And so as long as it continues to provide some level of value to the overall cause, it almost feels more like a marketing expense.
As such, and with Elon's eyes clearly on bigger prizes, it feels like X was finally given the breathing room to simply become a pretty good product again. Hiring Nikita Bier clearly helped and hopefully as he now steps back, the trend in the right direction continues...

X still has plenty of issues, namely around spam and self-promotion, both of which still pollute the feeds. VC Twitter, in particular, seems out of control in recent months, with every post now some big thought piece about the future of AI, undoubtedly written in part or in full by AI. I don't know who wants to read such stuff – especially given X's still shitty and cluttered reading experience – but X's algorithms are clearly geared towards promoting it.
But that too will pass. So the question going forward may be as simple as keeping X small enough in the grand scheme of things to be on the fringes of Elon's plate while not being so small so as to be shoved off it entirely.
What role might the network play in our AI future? Sure, the data can help feed Grok, but again, the network isn't at a scale that can match what Meta or certainly Google can use to feed their models. Is there some sort of role for X with personal agents, just as Muse is leveraging Threads and WhatsApp?
I still have a hard time believing they're ever going to make things like 'Chat', 'Calls', or 'Money' happen – as such, it's hard to see X as the true "Everything App", but if it can maintain the "town square" aspect of the internet, that's seemingly something money can't buy or be killed off. Many have tried.

1 I still hate the fucking name 'X'. It's just annoyingly aggressive and yet generic. 'Twitter' was a great name and brand, albeit perhaps too cute to be taken too seriously, X is the opposite. The Cybertruck of social brands.