Q&A: Tech Journalist Zoë Schiffer Tells the Inside Story of Elon Musk's Twitter Takeover
And why an outbreak of pornographic Taylor Swift deepfakes could be a bad sign for 2024 election disinformation on X.
Zoë Schiffer is the managing editor of the technology newsletter Platformer. She is the author of a new book: Extremely Hardcore: Inside Elon Musk's Twitter. In 2022, the Tesla and SpaceX CEO purchased Twitter for $44 billion and immediately fired much of its workforce; he later changed the name of the company to X. Schiffer conducted hundreds of hours of interviews with over 60 people who worked at the company. With two major wars and a U.S. presidential campaign happening, X remains a crucial platform for the dissemination of information; Musk has adopted a laissez-faire approach to content moderation which critics say allows disinformation to spread uncontrolled. Previously, Schiffer was a senior reporter at The Verge, where she reported on the labor movement in Silicon Valley. Her work has been featured in New York Magazine, the San Francisco Chronicle, and Vox, and she has appeared on CNN, NBC, CNBC, and BBC. We spoke last week over Zoom. Our conversation follows, condensed and slightly edited for clarity.
Luke Johnson: Before we start talking about Elon Musk, can you describe the state of Twitter before he purchased it? It seems that it was successful as the center of culture and politics, but it was also not profitable.
LJ: Describe the moment that Elon Musk decided to buy it.
ZS: He first floated the idea in April 2022. It seems that it was kind of a lark: he was a power user, and he had felt for a while that the company was sliding in a bad direction. From selling Tesla shares, he had about $10 billion in his pocket. He was thinking about what to do with the money. He started buying up Twitter shares. When this became public, he floated the idea of joining the board of directors. Through conversations with the board and with then-Twitter CEO Parag Agrawal, he decided that was not going to be an effective strategy. If he wanted Twitter to change, he needed to buy it outright.
LJ: He offers to buy the company at $54.20 a share -- which is a marijuana joke -- below the company's stock price, and then the stock goes down a lot. Then, the deal is actually more expensive than what the company's trading at. What's going on then? Why does Twitter ultimately want this deal?
ZS: When he first floated the idea of buying, Twitter wasn't entirely dismissive of it, but the board of directors wasn't particularly excited about it, either -- as you said, just a few months earlier, the stock was trading at about $68. It didn't look like an offer that was made with a detailed financial analysis; it looked like it was a marijuana joke. Twitter's board had a fiduciary duty to entertain the idea, but they also wanted to see what else was out there.
In the weeks following the offer, the markets tanked. Russia was invading Ukraine, the Fed was raising interest rates. Suddenly, the tech industry was hurting largely because the advertising industry had bottomed out and it wasn't clear when it was going to bounce back. All of a sudden this deal that initially looked like a joke, looked really really good to Twitter investors and eventually they accepted, at which point Elon Musk tried to pull out. However, he had signed a very strict purchase agreement, which didn't allow him a lot of wiggle room to get out. Over the summer, there's a protracted legal fight where he sues Twitter to try and get out of the purchase agreement, and Twitter countersues to try and force him to honor the terms of the agreement. Finally, he buys the company in October 2022.
LJ: You describe the team of sycophants that Musk brings in around him, and then there are very experienced engineers who know how Twitter runs. Describe both sides of this dynamic.
ZS: Elon Musk comes in with this reputation of being a genius and someone who's very technically savvy, but the areas that he's been successful in are not the same as not the same type of engineering as what Twitter is doing. He comes in with a group of advisers, whose posture initially is that all of the Twitter employees are lazy idiots. On the Twitter engineering side, their posture is that the [new team] doesn't seem to know what they're talking about at all, and they don't seem curious to learn from us. Almost from the very beginning, they're at odds, both sides thinking the other one doesn't have a good technical understanding of what's going on; both sides feeling like they're the other side are making assumptions that aren't beneficial.
LJ: He ends up laying off about how much of the company?
ZS: About half the company within about a week of the takeover. The firings continue for months to the point where between 70 and 80 percent of the company has been laid off.
LJ: What's the effect of that on the functioning of the site? I remember the site having functionality issues then.
ZS: Two narratives are emerging simultaneously. If you like Elon Musk, you think that he's laid all these people off, but Twitter is still working. It's still online. If you're on the other side, then you might think Twitter stayed online because the engineering team has spent years trying to make it as stable as possible in the event that something like this happened, but it's much glitchier. It's prone to bugs: you can't post links one day, the next day, you can't see photos.
LJ: Is it true that [Musk] has to hire some of these people back because only they know how to work with the code, and nobody else does?
ZS: The cuts are done very haphazardly. The company has a stock cliff, where [employees'] stock is going to vest on November 4, 2022. Initially, Musk, according to sources, does not want to pay it out. He wants to fire people before that, which would violate their employment agreements. Finally, finance and HR employees convince him that it wouldn't be a good look, so he [relents].
Largely because when he buys the company he strapped Twitter with billions of dollars in debt -- the company is now hemorrhaging money -- he wants to cut costs down as quickly as possible. Many of the people were cut who were absolutely crucial to Twitter staying functional. Over a weekend, Musk relented and said we can hire some of those people back if we made a mistake. Then, managers started calling people who were cut and said [to the effect of] just kidding, we would like to have you come back.
LJ: It's interesting how much cultural cachet Twitter/X has lost since Musk took it over. What do you think are the causes of that?
ZS: You can think of Twitter as the celebrity users and then the regular users. When we think about what the celebrity users want, someone like Taylor Swift wants to go on the platform to share that she has a new album coming out but not get death threats or see non-consensual fake nude images generated by AI of herself. She doesn't want a free speech landscape; she wants some level of content moderation so that she feels protected.
Twitter had been working with [Swift's] team to figure out what she didn't like about it and why she had stopped tweeting. But for [celebrities], Elon Musk has made a decision to roll back content moderation policies and to fire many content moderators to allow a lot more speech on the platform as part of his free speech ethos. I think the definition of [free speech] is pretty shaky, since he does tend to ban people whose speech he doesn't like.
[The platform] is not a very attractive setting for a high-profile person. Those same dynamics are true, but in a slightly different way, for the average person. When you allow anyone to say anything, you're actually limiting regular people being able to speak because when they do, they could be harassed, receive death threats, or be bullied. It makes the situation less stable and safe for the vast majority of people.
Tying verification to Twitter Blue, and allowing anyone with $8 to buy a blue checkmark was a significant change as well. The algorithm rewards people who have that blue checkmark. You are seeing a lot of people who are Musk fans -- or aren't opposed to his viewpoint -- rising to the top rather than the voices of accredited journalists or news outlets, who are not buying the blue checkmark.
LJ: There's two major wars happening right now in Ukraine and Israel with considerable amounts of disinformation. Under Musk, what has happened to disinformation on the platform, and how does it potentially affect these world events?
ZS: We think disinformation has gotten a lot worse. One of the things that he's done is to limit access for researchers to study the platform. It's really hard to get accurate, real-time data about what is happening on the platform.
The outbreak of the conflict between Israel and Hamas was an important test case. Twitter/X had thrived in moments like that: a global war was breaking out that had a lot of importance to people in the United States, including politicians and news outlets. In the immediate aftermath [of October 7], the platform was awash in disinformation. Musk rolled back content moderation policies and fired content moderators; he was relying on community notes, this crowdsource fact-checking tool, which was completely inadequate for a real-time global crisis. He had limited access for researchers who were saying, from the little that we can see this looks like it could be the disinformation that's coming out that mirrors some disinformation campaigns coming out of Iran, but we don't know for sure.
The disinformation was being shared by blue checkmarks. In the immediate aftermath, it looks like not only was Twitter/X rewarding people with blue checkmarks for sharing disinformation, but because of the revenue-sharing model that he had rolled out, it looks like they would be financially compensated for sharing disinformation because those tweets would go viral, and they would get a share at the advertising revenue. In the days that followed, Musk did say that if a tweet had a community note attached, it would not be able to share in advertising revenue, but it was too little too late.
LJ: And, many reliable journalists and NGOs no longer have the blue checkmark. The [accounts] that have the checkmark because they're paying $8 a month.
ZS: Exactly, and the voices of actual journalists are being completely drowned out. What you see are the viral misinformation tweets from blue check marks that are completely unreliable.
LJ: How do you think this is going to play out in the presidential election? There was so much misinformation in 2020, about alleged election fraud in Michigan and other states. Where do you see this going?
ZS: I think it's going to be a really bad situation. The stakes are so high with the coming election, and X simply hasn't put the policies in place to be successful. Take the AI-generated pornography of Taylor Swift that the platform was recently overrun by. From what I could tell, they didn't change their overall policies; they blocked all searches for her name. It was a particularly newsy weekend for her. All of a sudden, they applied this very blunt tool to a situation that was preventable if they had the right policies and detection tools ahead of time.
I think we're going to see a lot more of that. If I had to guess [X will be] very reactive when these campaigns happen. They always happen in elections. They don't have policies or practices to prevent them ahead of time. They're going to be reacting with various blunt tools.
LJ: Is there any appetite to deal with [disinformation] more proactively? Or is this part of the free speech ethos that Musk has brought in?
ZS: I'm sure if you talk to [X CEO] Linda Yaccarino, or her son who is running political advertising, they would say that they have a huge interest in making sure that there is no disinformation about the 2024 election on the platform. But I think that you have to look at what is actually happening, not what they're saying is happening. There was a Senate hearing last week on protecting kids online. Before that hearing took place, [X] announced they were going to hire 100 content moderators. However, they fired so many of them. Saying you're going to hire a team of 100 content moderators is not enough to fix this problem. I think that is what we're seeing: a lot of promises and window dressing, but not a lot of commitment and follow-through to take the steps necessary to make sure that the platform has a robust prevention apparatus.
LJ: Lastly, what do you see as the future of X advertisers have fled? Twitter Blue isn't making much on subscriptions. Does Musk want to find another buyer? Does he want to renegotiate with his investors? Where do you see the company going?
ZS: My prediction -- and this is just a prediction -- is that X will be a feeder and an input for xAI, which will become the more important project and Musk's bigger focus.
LJ: What is xAI?
ZS: it's an AI company that Elon Musk started and it's creating a chatbot to rival ChatGPT. While he has blocked ChatGPT and other rival chatbots from using X as training data for their models, he is allowing xAI to use that data. I think that over time, X will be an input for these large language models that xAI is building, but it won't be as important as a standalone platform.
LJ: What will happen to the old X?
ZS: I never thought that X was going to die in a bang. I think the deterioration is slow. I think if his focus continues to shift to other companies, then we'll see the stalling of innovation like we did before he had taken over the company. We will see more bugs and glitches that are fixed slower, and the platform will continue to lose cultural relevance.
More Public Sphere Coverage on Twitter:
Public Sphere is 100% supported by readers. To receive new posts and help support independent journalism, consider subscribing or upgrading your subscription to paid. Paid subscribers get bonus posts, commenting privileges, and the full archive.