Twitter's ragebait champagne v Threads's sparkling annoyance

Plus a little sprinkling of AI stuff

Last week I was too busy enjoying the election to look properly at the news that had come out from Meta: that there were 175 million monthly users of Threads, a year after its launch. (I was also too busy to put a proper title on the article, but the holding one worked well enough.)

In the footnote about why I wouldn’t write about it last week, I said

it would be too dull. It’s got 175 million users, but every time I go there it feels like people are discussing junk like “why do British electric plugs have so many pins?” It’s ratio food, but also the lack of violence makes it pointless.

I thought, with the election out of the way, that it would make sense to look more closely at what Meta is doing with this (relatively) new social network, which has certainly reached some sort of critical mass in just 12 months—at a growth rate that wouldn’t have disgraced the mid-2010s.

How big is 175 million users? The context is important. An FT article earlier this week looked at growth in eX-Twitter, based on numbers released by the company, which said that the number of “daily global active users” was 251 million in the second quarter of 2024. That’s only up by 1.6% from the year before, though to be honest I think it’s impossible to say anything specific about the numbers: they might be up or down, depending how you count all the spambots, but the broad figure is probably around 200-250 million. The lower figure is what Twitter used to have before Musk bought it, the bigger one what is claimed now. If you buy advertising because you believe the larger number, more fool you. The only ads I notice there now are crypto idiots, scams and dropshipping stuff.

So 175 million is OK, though it only scrapes in near the bottom of the Wikipedia list of social networks with more than 100 million. So what are people going there for?

Meta says most people are coming to Threads for text rather than images, noting that 63% of all Threads posts are text-only. But, that doesn’t mean images aren’t shared on Threads, as one in four posts include at least one image.

The company says more than 50 million Tags have been created, with the top three being PhotographyThreads, BookThreads and GymThreads.

While the metric of monthly active users alone isn’t enough to fully gauge how well Threads is doing, it does show that the platform has a notable amount of users checking it out. Part of Threads’ success is likely due to the fact that the platform is somewhat integrated and promoted within Instagram.

I think that “somewhat integrated” point is understating it. Any time I go on Instagram I am told that various people are waiting for or expecting to see me on Threads. (I ignore this message, because the account I use for Instagram is a completely different one from the Threads one.) But it’s probably effective on a good number of people.

Note the two figures that aren’t in the Threads number: the 100 million that it hit in the first week. This is actually bad news: it means that in the 51 weeks since, there have only been another 75 million users added , ie only a little more than a million per week, culled from Instagram and Facebook, which have more than a billion users each (though with a lot of overlap). In other words, perhaps one in a thousand users peeled off each week to join the people already there.

The other number that isn’t being revealed is daily active users—which you’ll note is the number that eX-Twitter does give. The difference between daily and monthly active users is very, very significant in the social network world. A monthly active user could be someone who checks in on a single day, and then not for another 30 days. A daily active user is in there every 24 hours. That’s a colossal difference in loyalty.

So what is Threads doing to try to convert those monthly users to daily users? My perspective is: by loading its algorithmic timeline with bait. Specifically, ragebait.

I noticed this about Threads when I dropped in (which I do, er, about once a month, compared to my daily eX-Twitter habit) the other day. What used, a while back, to have quite informative threads had instead transmuted into collections of posts which felt calculated to inspire dunking, either in the replies or in quotes.

That is two columns made up from one single scroll. Out of those eight posts, I’d say that six of them are ragebait, which seems like a hefty ratio. And sure, you can switch to your “following” timeline, if you can possibly figure out what to find it. (Don’t forget, 95% of people don’t change defaults.) And, no doubt, Twitter’s For You algorithmic timeline is also full of ragebait.

The problem for Threads is that Twitter has more of this stuff to choose from. It can pick from the very finest champagne ragebait. Whereas on Threads, it’s just sparkling annoyance.

What, if anything, is going to change the arc for Threads? At the very moment on Thursday night/Friday morning UK time that Joe Biden was arguably torpedoeing his candidacy by screwing up at the press conference at the Nato summit, Twitter was all over it; Threads was showing me a post about someone’s worst Tinder date. OK, Threads, you said that you aren’t going to do politics. But sometimes, politics is where the nexus of interest is. The fate of the free world seems pertinent to a lot of your users.

At some point in the future Threads is going to start having advertising, which will no doubt annoy some users, while not making it any more attractive. Threads is where the Twitter refugees go—as the reliance on text indicates—but Twitter meanwhile is becoming more and more video-heavy. Its inbuilt first-mover advantage and all the formerly done work that even Elon Musk’s cost-cutting can’t wreck puts it far ahead.

All of which is to say that I can’t see Threads catching up with Twitter, certainly not in engagement numbers. It’s just going to keep on trying to annoy people to attract them.

---

Glimpses of the AI tsunami

(Of the what? Read here. And then the update.)

• Early Apple tech bloggers are shocked to find their name and work have been AI-zombified. The Unofficial Apple Weblog (TUAW) was sold in 2015, and now it’s been turned into AI-generated slop with fake pictures.

• ChatGPT advises Tony Blair Institute which jobs AI is good at doing. Please don’t do this. If it said “Certainly ChatGPT is very good at flying planes”, are you going to strap in to a seat? Don’t be daft.

• Google researchers on how malicious actors are misusing generative AI. And that’s not even including the Tony Blair Institute saying “what are you good at?”

---

• You can buy Social Warming in paperback, hardback or ebook via One World Publications, or order it through your friendly local bookstore. Or listen to me read it on Audible.

You could also sign up for The Overspill, a daily list of links with short extracts and brief commentary on things I find interesting in tech, science, medicine, politics and any other topic that takes my fancy.

• Back next week! Or leave a comment here, or in the Substack chat, or Substack Notes, or write it in a letter and put it in a bottle so that The Police write a song about it after it falls through a wormhole and goes back in time.

---