If I ran X

How to transform the internet's most toxic platform into essential infrastructure.

If I ran X

“After two incredible years,” Linda Yaccarino tweeted, “I’ve decided to step down as CEO of 𝕏.” It was the day after the company’s AI chatbot, Grok, began calling itself MechaHitler and started spewing antisemitic slurs and advocating for Adolf Hitler in public until the company eventually had to remove them. Many of the users still posting to X greeted the antisemitism warmly; some tweeted, “finally”.

No new CEO has been named.

I’ve previously written about what I’d do if I ran Bluesky product, Mastodon, my own platform, and a fund supporting the open social web. I’d intended to stop the series there.

But as there’s an opening: let’s imagine I was handed the reins to the most toxic mainstream social network. What would I do if I was the new CEO of X?

I have some history with the people behind Twitter, the site that eventually became X. I consider multiple people who were there at the site’s inception, as well as many people who worked there in its later years, to be friends; Biz Stone sat on my first startup’s advisory board; I worked for Ev Williams at Medium for a little while. I’ve heard various pieces of founding stories from each of them.

And I really loved using Twitter. Long before Elon Musk bought it, I was an addicted user who posted many times a day. It opened up communities for me: I attended Twestivals (and hosted one), saw Ben Walker play his Twitter Song live, took part in countless hashtag games and still own a tote bag I won from the literary magazine Granta for writing a tweet-length horror story. Every time I went to an industry event or a conference, I’d be in the Twitter backchannel, making and reading observations, and building new connections.

And then, one day, Elon Musk walked into the building carrying a kitchen sink and began to refashion it around his need to amass power and influence to bring about his edgelord, pronatalist, Hitler-saluting view of the world. The name changed to X, Musk’s original name for PayPal, a company whose board voted to fire him as CEO due to mismanagement. Musk fired 80% of the staff, cratered the morale of the people who remained, reinstated Nazi accounts, hate speech surged — and the rest is history.

I’ve been given the reins. I’m CEO now. Can it be saved?

Throughout this tumult, X’s active user numbers have remained remarkably steady. 2024’s 335.7M monthly active users represented a 9% drop from 2022, the year Musk acquired Twitter. TechCrunch recently reported that its web views remain vastly higher than new competitors like Threads and Bluesky. Famously, many brands and journalists have been reticent to leave the site, despite its near-constant onslaught of white supremacy and hate speech. In other words: the traffic is there. These users may be a mix of bots, literal Nazis, and people who don’t mind posting on a site that is heavily used by literal Nazis, but there are hundreds of millions of them. So that’s something.

On the other hand, its revenue tanked following the acquisition, with some reports showing a drop as big as 66%. More recently, it saw 17.5% growth between 2024 and 2025, but it needs to claw back a lot of ground if it wants to return to its 2021 revenue peak — and even then, it posted a loss for most of the decade that preceded it.

In short: its online culture is toxic, its finances are decimated, and morale at the company is through the floor. But there is a great deal of remaining traffic, and many stalwarts have refused to give up followings they spent over a decade building.

If we want to turn it around, the only place it’s possible to start is the culture of the company itself. Everything else flows from there.

First, let’s address the ketamine in the room: there is no possible cultural turnaround at X while Elon Musk owns the site or has a significant stake. It must be acquired from him. That may seem far-fetched, but much of his wealth is stored in Tesla stock; in a world where his political machinations tank the price of his cash cow, he may need liquidity through other means. X’s valuation had already declined by 80%; his own xAI acquired it at a relative discount. As X’s culture continues to degrade, it may prove to be a drag on the valuation of his AI startup. The benefits of offloading it, in other words, may begin to outweigh the benefits to keeping it.

A coalition of hands-on investors would be needed to perform this acquisition, but consider the upside: a de-Nazified Twitter would be worth considerably more than its current incarnation. Musk had previously taken it private; the revived company could once again be floated on the stock market. The potential gains would be significant.

A newly-independent X would then be ripe for change.

Executives and managers who were loyal to Musk would need to be removed. It’s not a given that other employees were fully on board with his “extremely hardcore”, hard-right ideology: some may have needed to remain to maintain their visas or to ensure a paycheck in a tough hiring environment. Internally, diversity policies and mandatory inclusion training would return. Protections for work-life balance would return, and remote work would once again be allowed, with a staged roll-out. Work would be undertaken to establish an internal culture of empathy, which is a hard requirement for building non-toxic communities; experts like Carole Robin and Vernā Myers would be heavily drawn on for ongoing manager training and culture guidance, and experts like Dr Joy Buolamwini and Timnit Gebru would be drawn on to help the company be mindful the implications of ongoing work. This is the work of years, but the process would begin immediately.

To be clear, it’s not that employees would need to be left-wing. This isn’t an inquisition, and a company culture that only allows one political alignment is more harmful than good. But basic human decency would once again be a value: hate speech and bigotry would be explicitly disallowed, and it would be made obvious that the aim of Twitter’s platform is to create a community where voices from a diversity of communities should feel safe to post, which starts with internal cultural expectations.

Nazi accounts would need to be banned again, swiftly, with their posts preserved internally in a vault in a de facto legal hold. The messages would be made available as a dataset to organizations that research and investigate hate groups, although users would have the ability to have their content removed in accordance with relevant privacy legislation for their jurisdiction.

Significant investments would be made into non-profits and on-the-ground organizations that support the vulnerable communities harmed by the hate speech, doxxing, and other activities previously enabled on X. A commitment would be made for ongoing support: it couldn’t be a one-and-done gesture.

A permanent, inclusive trust and safety council would be established, modeled after the Supreme Court, with real, encoded power to affect company policies around content and moderation. This council and its permanence would be defined in the articles of association of the new company; it would be almost impossible to remove them. Its members would be guaranteed a salary (but could not own stock) and required to both adjudicate on content moderation issues and intervene during strategic lapses. The number of members of the council would increase based on active users but could never decrease.

Employees who objected to these measures would be free to take a buyout. Some would likely just quit. Meanwhile, the company would begin to aggressively hire across the board, including but not limited to engineering, product, research, design, partnerships, communications, and trust and safety. Public statements and heavy PR would signal the internal cultural changes in order to attract new applicants.

The name and brand would be symbolically restored to Twitter, indicating a clean break from the Musk years. It’s a trustworthy brand, and the change back would likely turn heads. But it would be a mistake to simply return the company to its old self. The new Twitter needs to be something else entirely.

It’s now clear that Musk’s acquisition was about amassing power, not providing value. As Dan Primack wrote in Axios last year:

Musk used the platform to help shape public opinion during the election, artificially amplifying his political messages and giving him influence with Donald Trump that money alone wouldn't have bought.

While this might seem like a selling point to a would-be autocrat, it serves as a weakness to Twitter as a business. Not only does this property make it vulnerable to takeovers from terrible people, but it elevates the company’s business decisions to a level where a mistake could significantly impact democracy or even public safety around the world. (This is also the lesson Meta failed to learn in Myanmar.) In turn, this could subject the company to greater regulatory scrutiny and oversight. It may seem like value at first sight, but it’s actually a business risk.

So how can the new Twitter be incredibly valuable but also safer from regulatory oversight?

The core value of Twitter is not its social network as such. That’s always been a means to an end. The vast majority of content on the site has historically been created by just 10% of users, but people don’t log in to post: they want to see what other people are saying. The point of Twitter is to find out what’s going on right now and get an insight into the conversation around it.

My new Twitter would retain the social network but expand upon the idea that people want to know what’s happening right now. The feed would expand to not just include Twitter’s own conversations, but every public conversation across the social web. Search, trending topics, and live dashboards would represent the entire addressable real-time web, using open standards and protocols like AT Protocol, ActivityPub, and RSS. Twitter would become the way to find out what’s happening everywhere.

Some social networks — like Facebook — would not be available via this feed. The addressable social web would consist of sites that embrace open protocols (best), or otherwise make a content deal directly with Twitter (less good). However, a Twitter that aims to provide the pulse of the internet and not necessarily a social network in itself is less of a competitor to closed networks; in fact, the tools may provide new engagement and users for them. The aim would be, over time, to make not being represented on the feed more of a business liability for closed social networking platforms than being included. Twitter would provide libraries, ecosystem support, and even funding for open social web protocols: its value increases as the the social web pie gets bigger.

Monetization would be three-prong. Advertising would remain on public feeds, and a renewed focus on trust and safety would improve brand safety, bringing advertising prices back up. Twitter Blue premium profiles on the social network would remain. But most differently, enterprise tools for tracking public sentiment and yielding unique insights from the global conversation would be made available for corporations, media companies, and governments. These tools would also be made available to public research institutions, with some restrictions, for free.

Using open protocols should help reduce regulatory risk. In this world, trust and safety is as much about de-listing external content as it is about policing the in-house social network. Aggregating across the entire social web is an enormous infrastructure task that requires real resources and expertise to carry out; the compliance demands across global jurisdictions are complex but fully within Twitter’s wheelhouse. The moat for the business is, in part, the combination of the difficulty of building the tools and their ease of use for end users. In-house expertise would create proprietary machine learning algorithms for sentiment and insight analysis, and integrations with third-party tools like CRMs and analytics platforms could embed the platform in existing enterprise ecosystems. It’s a very similar model to how Google originally became the dominant search engine: simplicity of use combined with the sophistication of its internal technology, all layered over the open web.

It’s not even particularly unprecedented: Blaine Cook, then CTO of Twitter, was fired for demonstrating interoperability with Jaiku, another social network. For this new Twitter, interoperability would be core to its value.

This transformation wouldn’t happen overnight. A still-aggressive but more realistic timeline might look like:

  • Culture change: biggest changes in the first six months, then ongoing
  • Platform redevelopment and rebuilding the team: 1-2 years
  • Building up aggregation and partnerships: 2-3 years

Once established, the new Twitter would set the tone for the social web itself. By implication, it wouldn’t just be the way that conversations were discovered; people would find platforms this way, too. By not hyper-focusing on its own social network, it has the potential to become a much bigger and more valuable company: one that people use to unlock the entire online social conversation. By using open protocols, the company ensures the whole internet isn’t locked into its services, but by caring for trust and safety, design, and the quality of its feeds, it provides best-in-class options for people to choose. And it fosters a bigger, more interesting social web in the process.

Does X or Twitter actually need to be saved? Absolutely not. There are plenty of new social networks that are more interesting, more fun to use, and show more promise. But there’s interesting potential here, both to remove a major source of toxicity on the internet, to diffuse a centralized source of power that can be abused by autocrats and oligarchs, and to create a way to elevate the entire social web.

So that’s what I would do if I was CEO.