Elon Musk Wants You to Hear About Substack Notes
He's boosting the Substack competitor to Twitter that he wants to kill. Plus: Killer AI sounds fake to me.
You may have heard that Substack will soon be launching a feature called Notes, which will be a place to “post short-form content and share ideas… including posts, quotes, comments, images, and links.” In other words, the feature will be very similar to Twitter. You will even be able to “restack” another writer’s note.
The reason you may have heard is probably not that you saw Substack’s Wednesday announcement that Notes is coming. I mean, maybe you did — maybe it came to your inbox because you’re a big fan of Substack qua Substack and so you subscribe to Substack’s substack about Substack, On Substack. But probably not. I’m guessing if you heard, it’s because Substack’s announcement made Elon Musk mad, and he used Twitter to retaliate.
As The New York Times reports, “the changes by Twitter on Friday meant that Twitter users could still share links to Substack newsletters, but blocked other users from liking or resharing those links.” This actually isn’t quite true — it’s still possible to like or reshare tweets with links to this newsletter and any other Substack newsletter that uses a custom web domain, i.e., any newsletter that is not located at a .substack.com address. But even though the policy is not (yet) directly affecting Very Serious or Serious Trouble, it still bothers me a lot — it means Twitter is not useful for talking about content on many other Substacks; and I assume, if there are any engineers left at Twitter, Musk will eventually get around to blocking those of us with custom domains, too.
Note again that this kerfuffle was covered in The New York Times. The Times did not cover the initial announcement of Substack Notes — it wasn’t an important news story. But once Elon acted, he got a wide variety of outlets to cover his competitor’s new product. Matt Taibbi, one of the highest grossing authors on Substack and (until recently) a vocal advocate for Musk, has announced that he’ll be leaving Twitter to use Substack Notes. I can currently see inside the private beta for Substack Notes, and a lot of prominent Substack writers are already engaging there, much more actively than I suspect they would have if Elon were not taking action to reduce Twitter’s usefulness to us. It’s basically a tech version of the Streisand Effect.
And then there’s me. Like Taibbi, I am also withdrawing from Twitter in favor of Substack Notes, for at least as long as this policy is in effect.
I certainly understand that Musk needs to generate more revenue for Twitter. He’s under no obligation to let competing platforms gain advantage through interaction on the one he owns. Platforms often take steps to reduce the likelihood that users will be attracted to other platforms; one such move, which Washington Post reporter Taylor Lorenz pointed out yesterday in the Substack Notes beta, is that Instagram down-ranks Reels videos that display TikTok logos.
Still, I don’t think Musk has really thought through what he’s doing here.
First, Musk has consistently said his motives for purchasing Twitter are not purely financial. He wants to foster a better culture of free speech on the Internet, he claims. Even if you don’t take Musk at face value there — if you think he’s only interested in spreading specific kinds of speech that he agrees with — then the action still doesn’t align with that goal. In recent months, Musk has leaned heavily on certain prominent Substack writers (including Taibbi, Bari Weiss, and Michael Shellenberger) to share his narrative about the past and present management of Twitter. He has recognized some writers on this platform as important counterweights to traditional media outlets, and now he’s telling them to fuck off. This move is not aligned with any version of what you might take to be his goals about speech on the internet.
Second, the move is part of Musk’s overall self-defeatingly rude approach to the people he would like to be Twitter’s customers. He wants Twitter’s advertisers to come back, but he also threatens to “thermonuclear name and shame” advertisers who leave Twitter, which is a pretty strong inducement not to advertise there in the first place. Ad revenues are still deeply depressed from their levels before he bought the company. He wants media companies to pay substantial monthly fees to have their Twitter feeds verified, but he also fired the company’s entire PR office and set its media relations email to automatically respond to comment requests with poop emojis. And even his approach to individual users has been taunting.
The whole thing just has a weird air of entitlement coming from a guy who vastly overpaid for a business that was never that strong and never had a track record of attracting subscription revenue. He thinks he can insult people into giving him a reliable revenue stream, and I don’t believe that will work.
A lot of people have gotten weirdly emotional about Musk’s takeover of Twitter — you have his fanboys, and you have people writing eulogies to the site. Personally, I understand that I have no right to Twitter and Twitter has no right to me. Elon Musk can mismanage the company however he likes; it’s his $44 billion on the line. But I also don’t have to keep using his product.
I know some of you are on Mastodon. Personally, I am suspicious of decentralized open-source projects, since they have a tendency to turn into virtual HOAs. I haven’t joined Mastodon because I’m not any more interested in being at the mercy of a condo board full of MSNBC addicts than I am in being at the mercy of Elon Musk.
But I obviously have a positive view of Substack, a company that I think respects its creators, builds pleasant-to-use products, and has good instincts for how people want to read on the internet, and I’m looking forward to engaging with you on Substack Notes when it launches to the public in just a few days.
Once that’s available to you, I’ll let you know how you can join the conversation with me.
I don’t believe in killer AI. You’ve seen the AI-generated images I include in a lot of newsletter posts, created on OpenAI’s Dall-E engine. Dall-E has a competitor called Midjourney, and one strange thing about Midjourney is that you can use it to make satirical images of world leaders like Joe Biden and Vladimir Putin — but not Xi Jinping.
"I think we want to minimize drama," Midjourney’s CEO said last year. "Political satire in China is pretty not-okay and at some point would endanger people in China from using the service.”
This sucks in the way it always sucks when American businesses abet the censorship efforts of the communist Chinese regime. But that’s not why I want to talk about Midjourney and Xi. What I want to note is that this shows AI is nothing more than software: It does what humans tell it to do, and if humans decide the AI shouldn’t make mocking images of Xi, then it doesn’t.
There is a lot of talk lately about existential risk from AI, and it’s being driven — I use this term advisedly — by nerds. There’s always some nerd fad about a radical public policy view: A few years ago, it was that computers were going to get so good at our jobs that we wouldn’t be able to find work and therefore we need a universal basic income. Then, it was that blockchain was going to “disrupt” all of our traditional systems for money and contracts and trust, and eliminate the need for central banks. Now, it’s that computers are going to kill us all. One thing these trends have in common is that they’re advanced by people who have otherwise never thought deeply about public policy and who have no concept of how their prescriptions might intersect with broader political and economic realities.
Currently there’s a letter circulating calling for a six-month “pause” on AI development, signed by the likes of Musk and Steve Wozniak, positing that AI development on its current track means we “risk loss of control of our civilization.” They call for the government to impose a moratorium on AI development if the labs don’t agree to pause voluntarily. They do not engage with the practicalities of that — how do you define “AI development” for policy purposes? What if other countries don’t agree to the pause — wouldn’t we just be letting China get ahead of us on the development of an important technology? “AI Safety” expert Eliezer Yudowsky says — in Time magazine! — that the answer to that problem is airstrikes:
If intelligence says that a country outside the agreement is building a GPU cluster, be less scared of a shooting conflict between nations than of the moratorium being violated; be willing to destroy a rogue datacenter by airstrike.
I am sure that AI will pose problems — it’s already creating some, for example by facilitating cheating on school assignments — but this doomsday stuff just marks its proponents as bored dilettantes with no idea how policy or international relations work.
So I assume that in four years, this whole panic will be forgotten, just like how nobody talks about UBI anymore.
Very Serious is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.