from the so-much-missed-opportunity dept
Today, the new book by NY Times reporters Kate Conger and Ryan Mac, Character Limit: How Elon Musk Destroyed Twitter, comes out. If you’re at all interested in what went down, I can’t recommend it enough. It’s a well-written, deeply researched book with all sorts of details about the lead-up to the acquisition, the acquisition itself, and the aftermath of Elon owning Twitter.
Even if you followed the story closely as it played out (as I did), the book is a worthwhile read in multiple ways. First, it’s pretty incredible to pull it all together in a single book. There was so much craziness happening every day that it’s sometimes difficult to take a step back and take in the larger picture. This book gives readers a chance to do just that.
But second, and more important, there are plenty of details broken by the book, some of which are mind-boggling. If you want to read a couple of parts that have been published, both the NY Times and Vanity Fair have run excerpts. The NY Times one covers Elon’s infatuation with “relaunching” Twitter Blue as a paid verification scheme a week after he took over. The Vanity Fair one looks at the actual closing of the deal and how chaotic it was, including Elon coming up $400 million short and demanding that Twitter just give him the money to cover the cost of closing the deal.
Both excerpts give you a sense of the kinds of amazing stories told in the book.
But as I read an advance copy of the book, two things stood out to me. The first was Elon’s near total lack of understanding of the concept of Chesterton’s Fence. The second was how much the old regime at Twitter was already trying to do almost everything that Elon claimed he wanted to do. But as soon as he took over, he was so sure (1) that the old regime were complete idiots and (2) that he could reason his way into solving social media, that he not only ignored what people were telling him, he actively assumed they were trying to sabotage him, and did away with anyone who could be helpful.
Elon rips out some fences
If you’re unaware of the concept of Chesterton’s Fence, it’s that you shouldn’t remove something (such as a fence) if you don’t understand why it was put there in the first place. Over and over in the book, we see Elon dismiss all sorts of ideas, policies, and systems that were in place at Twitter without even caring to find out why they were there. Often, he seems to assume things were done for the dumbest of all reasons, but never bothered to understand why they were actually done. Indeed, he so distrusted legacy Twitter employees that he assumed most were lying to him or trying to sabotage him.
It’s perhaps not that surprising to see why he would trust his own instincts, not that it makes it smart. With both Tesla and SpaceX, Elon bucked the conventional wisdom and succeeded massively. In both cases, he did things that many people said were impossible. And if that happens to you twice and makes you the world’s wealthiest person, you can see how you might start assuming that whenever people suggest that something is a bad idea or impossible, you should trust your gut over what people are telling you.
But the point of Chesterton’s Fence is not that you should never do things differently or never remove policies or technology that is in place. The point is that you should understand why they’re there. Elon never bothers to take that tiny step, and it’s a big part of his downfall.
In Character Limit, we see that Elon has almost no actual intellectual curiosity about social media. He has no interest in understanding how Twitter worked or why certain decisions were made. Propped up by a circle of sycophants and yes-men, he assumes that the previous regime at Twitter must have been totally stupid, and therefore there is no reason to listen to anything they had to say.
It is stunning how in story after story in the book, Elon has zero interest in understanding why anything works the way it does. He is sure that his own instincts, which are clouded by his unique position on the platform with tens of millions of followers, represent everyone’s experience.
He’s quite sure that his own instincts can get him to the right answers. This includes thinking he could (1) double advertising revenue in a few years (when he’s actually driven away over 80% of it) and (2) eclipse even that erroneously predicted increased advertising revenue by getting millions of people to pay for verification. In actuality, as the book details, a tiny fraction of users are willing to pay, and it’s bringing in just a few million dollars per quarter, doing little to staunch the losses of billions of dollars in advertising that Elon personally drove away.
The stories in the book are jaw-dropping. People who try to explain reality to him are fired. The people who stick around quickly learn the only thing to do is to lie to him and massage his ego. And thus, the book is full of stories of Elon ripping out the important pillars of what had been Twitter and then being perplexed when nothing works properly anymore.
He seems even more shocked that tons of people don’t seem to love him for his blundering around.
Old Twitter was already planning on doing what Elon wanted, but way better
Perhaps this is somewhat related to the last point, but the book details multiple ways in which Parag Agrawal, who had just taken over from Jack Dorsey a few months earlier, was already looking to do nearly everything Elon publicly claimed he wanted to do with Twitter.
When Elon first announced the deal to buy Twitter, I suggested a few (unlikely, but possible) ways in which Elon could actually improve Twitter. First up was that by taking the company private, Elon could remove Twitter from the whims of activist investors who were more focused on the short-term than the long-term.
The book goes into great detail about how much activist investors created problems for both Dorsey and Agrawal, pre-Musk. Specifically, their revenue and user demands actually made it somewhat more difficult to put in place a long-term vision.
In my original post, I talked about continuing Twitter’s actual commitment to free speech, which meant fighting government attempts to censor information (not just when you disagreed with the political leaders).
But beyond that, there were things like further investing in and supporting Bluesky (see disclaimer)* and its ATprotocol. After all, Elon claimed that he wanted to “open source” the algorithm.
Moving to an open protocol like ATProtocol would have not just allowed the open sourcing of the recommendation algorithm, it would have opened up the ability for anyone to create their own algorithm, both for recommendations and for moderation. Instead, that’s all happening on the entirely independent Bluesky app, which really only exists because Elon threw away Twitter’s deal to work with Bluesky.
Furthermore, the book reveals that well before Elon came on the scene, Parag and other top execs at the company were working on something called Project Saturn, which was discussed a bit in Kurt Wagner’s earlier book on this topic, but which is explained in more detail here.
The book reveals that Parag very much agreed with Elon (and Jack) that expecting companies to constantly completely remove problematic content was not a very effective solution.
So they created a plan to basically rearchitect everything around “freedom of speech, not freedom of reach.” Ironically, this is the very same motto that Elon claimed to embrace soon after taking over the company (and after firing Parag).
But Parag and others at Twitter had already been working on a system to operationalize that very idea. The plan was to use different “levels” and “circles” in which users who were following the rules would have their content eligible to be promoted to varying degrees within the algorithm. The more you violated the site’s rules, you would move to further and further outer layers/rings of the system (which is where the Project Saturn name came from). This would lead to less “reach,” but also less of a need for Twitter to fully remove accounts or tweets.
It was a big rethinking of how social media could work and how it could support free speech. In reading about it in the book, it sounds like exactly what Elon said he wanted. A small team within Twitter, pushed by Parag’s vision, had been working on it since way before Elon purchased his shares and started the takeover process. According to the book, even as Elon caused such a mess in the summer of 2022 trying to back out of the deal, Parag kept pushing the team to move forward with the idea.
Once Elon took over, it appears that a few remaining people at the company tried to show him Project Saturn and explain to him how it would match the ideals he had talked about.
But Elon ignored them, tossed out all the work they had done on it, and just randomly started unbanning people he thought belonged back on the platform without any plan on how to deal with those users if they started causing problems (and driving away advertisers). He assumed that his new verification plan would solve both the revenue issues for the company and all moderation issues.
Even the idea that Twitter was too bloated with excess employees and a lack of vision seemed to be part of Agrawal’s plans. Before Elon had made his move, the book reveals that Agrawal had drawn up plans to lay off approximately 25% of the company and greatly streamline everything with a focus on building out certain lines of business and users. He did move to lay off many senior leaders as part of that streamlining, though it wasn’t as clearly explained at the time what the larger plan was. Elon’s effort to buy Twitter outright (and then back out of the deal) forced Agrawal to put the layoff plans on hold, out of a fear that Elon would view those layoffs as an attempt to sabotage the company.
It’s truly striking how much of what Elon claimed he wanted to do, Parag and his exec team were already doing. They were making things more open, transparent, and decentralized with Bluesky. They were decreasing the reliance on “takedowns” as a trust & safety mechanism with Saturn. They were betting big on “freedom of speech, not reach” with Saturn. They were fighting for actual free speech with legal actions around the globe. They were cutting employee bloat.
But the company was doing all of those things thoughtfully and deliberately, with a larger strategy behind it.
As the book details, Elon came in and not only tore down Chesterton Fences everywhere he could, he dismissed, ignored, or cut loose all of those other projects that would have taken him far along the path he claimed he wanted to go.
So, now he’s left with a site that has trouble functioning, has lost nearly all of its revenue, and is generally seen as a laughingstock closed system designed just to push Elon’s latest political partisan brain farts, rather than enabling the world’s conversation.
Of course, in the wake of all that destruction, it has enabled things like Bluesky to spring forth entirely unrelated to Twitter, and to put some of this into practice. Just this weekend, Bluesky passed 10 million users, helped along by Elon’s (again) hamfisted fight with Brazil, which (like so many other things Elon) may have a good reason at its core (fighting against secretive government demands), but was done in the dumbest way possible.
If there’s one thing that is painfully clear throughout the book, it is that Elon was correct that there were all sorts of ways that Twitter could be more efficient, more open, and less strict in takedowns. But he handled each in the worst way possible and destroyed what potential there was for the site.
Later today on the podcast, I’ll have an interview with Kate Conger about the book and Elon where we talk some more about all of this.
- As I’ve said before, I’m now on the board of Bluesky, which wouldn’t have been necessary if Elon hadn’t immediately cut Bluesky free from Twitter upon taking over the company.
Filed Under: character limit, chesterton fences, content moderation, elon musk, free speech, kate conger, parag agrawal, project saturn, ryan mac, social media
from the good-deals-on-cool-stuff dept
The Database Administration Super Bundle has 9 courses to help you go from data novice to expert administrator. You’ll discover how to build and manage databases with MySQL and MongoDB. Courses also cover Microsoft SQL Server, Informatica, Minitab, Tableau, and regression modeling. It’s on sale for $60.
Note: The Techdirt Deals Store is powered and curated by StackCommerce. A portion of all sales from Techdirt Deals helps support Techdirt. The products featured do not reflect endorsements by our editorial team.
Filed Under: daily deal
from the hurling-one-last-'fuck-you'-to-the-public dept
It’s been a whirlwind few months for NYPD Commissioner Edward Caban, at least in terms of press coverage. None of it has been positive, though.
A few months ago, reporting from ProPublica exposed Commissioner Caban’s efforts to make the NYPD even less accountable than it always has been. Public records and other information obtained by ProPublica made it clear Caban wasn’t interested in cleaning up a police department that was costing residents millions of dollars a year in the form of lawsuit settlements.
Earlier this month, things went from par-for-the-unaccountable-course to shit hitting the fan. The FBI spent a few days raiding homes and offices of city officials and direct subordinates of Mayor Eric Adams. Adams, a former NYPD officer, claimed to be unaware of any criminal activity happening under this nose — something he buttressed by stating he had repeatedly told staff and appointees to “follow the law,” which is the sort of thing you only need to say more than once when staff and appointees seem insistent on breaking the law.
One of the targets of the FBI raids was none other than Commissioner Edward Caban. It’s unclear at this point if it has anything to do with his brother, Richard Caban, who runs a bar catering to NYPD officials that has been repeatedly cited for violations of building and fire codes. Even if not, that bit of information couldn’t have helped.
With the FBI breathing down the neck of the most powerful officials in New York City, Commissioner Caban has decided to press the eject on this phase of his career, resigning his post before the city has a chance to punish him for his (alleged) involvement in whatever it is the FBI is investigating.
But he has deployed a golden-esque parachute of sorts. Not for himself, though. The officers he’s leaving behind have been given an undeserved parting gift: reduced punishment for misconduct, courtesy of a top cop who spent his short term in office doing little more than reducing punishments for misconduct. Here’s Reven Blau with the details for The City:
Fun stuff if you’re a misbehaving cop. And the NYPD has several of those. The matrix says less the bad stuff for bad cops, which means things that used to result in five-day suspensions are now subject to nothing more harsh than “additional training.”
The list of violations effected by the dilution of accountability ranges from the mundane to the “hey, maybe we shouldn’t let cops get away with this sort of thing.”
Some of these are violations best described as violations of employer policies. They should still be punished, though, and with more than some remedial training. Others are little more than “officers being assholes.” Even so, assholish-ness should be actively discouraged. Cops treating the people they serve disrespectfully can’t be handled with an extra PowerPoint presentation or two.
The worst part is Caban’s mandate also covers things are the subject of federal lawsuits:
This isn’t ticky-tack bullshit. This is the actual violation of constitutional rights. Treating this as a training problem is absolutely the wrong way to handle this. Certainly, re-training might be necessary but that should be on top of harsh punishments that make it extremely clear violating rights is never acceptable. And repeat offenders should be shown the door. Anything less than that is an insult to city residents, who not only have to deal with having their rights violated but are expected to cover the cost of lawsuit settlements with their taxes.
Caban’s exit and parting shot at accountability are par for the course, unfortunately. Officials who resign in the middle of outside investigations are also par for the course. Very few public officials are willing to take what’s coming to them, preferring to get out and under the radar before things get truly ugly. Caban’s exit is basically an admission of guilt. And it’s not enough for him to dodge his own personal accountability by taking the easy way out when faced with the consequences of his actions. He had to make it easier for the cops he left behind to do the same thing. These aren’t the actions of a leader. They’re the actions of a coward who loved the power but hated the responsibility.
from the it-never-ends dept
Just when you thought it was safe to go back on the internet, KOSA rears its ugly head once again.
The rumors of KOSA’s demise in Congress may have been overstated. Following a big push by supporters of the bill, including Senator Marsha “we need this to protect kids from the transgender in our culture” Blackburn, House Energy and Commerce Committee Chair, Rep. Cathy McMorris Rodgers, has announced that she’ll hold a markup of it and 15 other bills on Wednesday of this week.
This does not mean that KOSA is really going to get a vote. Lots of things could happen. But it does mean that KOSA (and COPPA 2.0, which the Senate combined into KOSPA — the Kids Online Safety & Privacy Act) are getting a bit of new life.
It’s possible the markup will be delayed or won’t actually happen. Markups get announced and delayed and sometimes shelved entirely. And what happens at the markup may matter. Markups are when other committee members can offer up amendments, and it gives everyone a sense of what people feel about a bill. It’s possible that amendments could change KOSA quite a bit, though the fundamental problems of the bill are unfixable.
I’ve also heard that House GOP leadership is still not a fan of the bill. So, even if it goes through a markup and passes out of committee, that doesn’t mean that House Speaker Mike Johnson would agree to bring it to the floor.
Since the House bill is still significantly different from the Senate version that passed, even if the bill went to the floor and passed, there would still need to be a reconciliation between both versions and another vote.
In short, there are still plenty of reasons why KOSA might not become law. But, the fact that the markup has been announced suggests that it could move forward and is not totally dead.
If you have a Representative who is on the Energy & Commerce Committee, you might want to call your Representative and point out the many, many problems with the bill. If your Rep is a Republican, I’d recommend Rand Paul’s thoughtful exploration of the problems with the bill. If your Rep is a Democrat, then just highlight how hard the Heritage Foundation is pushing for the bill, and how it sees it as part of its Project 2025 goals to have more power to stop speech it dislikes on the internet.