Good analysis here. Here are Luu’s reasons why people post on twitter or do videos instead of blogging:
Engagement
Just looking at where people spend their time, short-form platforms like Twitter, Instagram, etc., completely dominate longer form platforms like Medium, Blogspot, etc.; you can see this in the valuations of these companies, in survey data, etc. Substack is the hottest platform for long-form content and its last valuation was ~$600M, basically a rounding error compared to the value of short-form platforms . . . The money is following the people and people have mostly moved on from long-form content. And if you talk to folks using substack about where their readers and growth comes from, that comes from platforms like Twitter, so people doing long-form content who optimize for engagement or revenue will still produce a lot of short-form content.
Friends
A lot of people are going to use whatever people around them are using. . . . Today, doing video is natural for folks who are starting to put their thoughts online.
Friction
When people talk about [bad platform] being lower friction, it’s usually about the emotional barriers to writing and publishing something, not the literal number of clicks it takes to publish something. We can argue about whether or not this is rational, whether this “objectively” makes sense, etc., but at the end of the day, it is simply true that many people find it mentally easier to write on a platform where you write short chunks of text instead of a single large chunk of text.
Revenue
And whatever the reason someone has for finding [bad platform] lower friction than [good platform], allowing people to use a platform that works for them means we get more content. When it comes to video, the same thing also applies because video monetizes so much better than text and there’s a lot of content that monetizes well on video that probably wouldn’t monetize well in text.
Luu demonstrates with many examples.
I’m convinced by Luu’s points. They do not contradict my position that Blogs > Twitter (see also here). Luu demonstrates solid reasons for using twitter or video, even if blogging results in higher-quality argumentation and discussion.
Blogging feels like the right way to go for me, but I also like writing articles and books. If I’d been born 50 years earlier, I think I just would’ve ended up writing lots more books, maybe a book a year instead of every two or three years.
As for Luu, he seems to do a lot more twitter posting than blog posting. I went on twitter to take a look, and his twitter posts are pretty good! That won’t get me to be a regular twitter reader, though, as I have my own tastes and time budget. I’ll continue to read his blog, so I hope he keeps posting there.
P.S. I was thinking of scheduling this for 1 Apr and announcing that I’d decided to abandon the blog for twitter, but I was afraid the argument might be so convincing that I’d actually do it!
Today’s blog (about blogging) is dated May 28, 2024. Was there a blog dated May 27, 2024? I ask because I believe the Columbia University website has been unavailable for the past 24 hours or so.
Paul:
I moved the post scheduled yesterday to a later date.
Super interesting. Thanks for sharing.
Short-form vs. long form, or Twitter vs. Substack: The medium is the message!
Put slightly differently, various media have a different comparative advantage over messages. Marsh McLuhan was a student [and neighbor?] of Harald Innis’, who did teach such stuff.
Gonna plug Mastodon again for anyone who enjoys engaging with the kind of topics found on this blog.
I’ve been following a number of groups and people who tickle the interests I have. Including Shravan Vasishth and Richard McElreath every so often, but also some heterodox Economists like Kyle Montanio @[email protected] and the group @[email protected] (a.gup.pe groups are bots that act as boosters of anyone who tag them) as well as following hashtags like #economics #mmt #statistics #bayes #bayesian #julialang etc.
I’ve found that conversations stay civil, evolve throughout the day/week to potentially include multiple points of view, and it’s generally an intellectually honest crowd. If you find annoying people you can block them easily.
Fediverse people are much more likely to be atypical, such as neurodiverse, gay, transgender, or even (gasp) bayesian, but of course plenty of regular folks as well.
The group globally is fairly small, on the order of around 10M people with probably 2.5M active in any week, but they’re also incredibly engaged. People like Cory Doctorow who has 500k followers on Twitter barely get any Twitter engagement, and with 55k followers on Mastodon gets thousands of mentions a day (@[email protected])
The biggest barrier for most people is the first step, picking a server. This isn’t as hard as you might think. For readers of this blog they might want to choose https://bayes.club/ or https://fosstodon.org or https://fediscience.org but after you’ve established yourself it’s relatively easy to switch servers and have all your followers come along automatically. If you want to find a different server one way to do so is to use: https://www.comeetie.fr/galerie/mapstodon/ (which is based on well-out-of-date data, but is still somewhat useful)
After you’ve got an account, you’ll need to find people to start following, a good way to get started is first follow anyone you know (such as people on bayes.club) and then once you have a small follow group at least, use https://followgraph.vercel.app/ to expand that group to people related to who you know…
Also follow hashtags. Doing that and after a little while you can have a pretty decent set of discussions about science, statistics, academia, and soforth.
There is a wordpress plugin which would post this blog into the fediverse, if Andrew chooses to use it. It would potentially let everyone follow the blog posts here from there, and/or post here from there (not sure about details of how it works).
Daniel:
We do have a Mastodon feed! It’s on our list of 6 ways to follow this blog.
Yep, forgot to mention that. It’s a bot, and I follow it there.
I haven’t tried Mastodon as I’m unhappy with Twitter-style social media in general. But overall it appears to me structured to give a lot of power to the worst scolds in a community. If someone is sure they won’t ever offend those people, fine, it works for them. But I’d rather not have that over my head. For example, https://hub.fosstodon.org/coc/
“Do not tone police. Other toots may violate the standards of civility we set on Fosstodon, but you should not criticize users for their tone. If the user is a member of the Fosstodon community, you should report them; if not, you should feel free to block them. In neither case should you criticize their tone.”
This is a big red flag to me. It means if someone heaps racial or gendered abuse on you, and you reply by merely saying something like “There’s no call for that”, or “It’s unfair to say that”, then *they* get to report *you* for violation. And it’ll come down to who the moderators side with. Again, why do I need to deal with this? The scolds will surely know how to work the system, that’s what they do in life.
Yes, one can move servers. But the point is why should I need to have a detailed understanding of the community politics and hot-buttons, which are the obsessions of the most neurotic members?
I hear this complaint and I think maybe it’s relevant to some people, but it literally never, not once ever, has been an issue for me in about 2-3 years of use.
For the most part my interactions on Mastodon are almost identical in scope and style to interactions on Gelmans blog. So I like it a lot.