9 Comments

Jeremy. Becoming a paying subscriber because of this post. This is what the journalism world needs - solutions. Not just more of what we have. (I also don't use Twitter, btw, and never ever will) Make sure you aren't leaving out people who don't use Big Tech stuff. We are many.

Expand full comment

Good note :)

Expand full comment

Hi Jeremy! I've been following your blog since you launched it (I think). I don't use Twitter, so I have to leave my comment here -- hopefully you'll see it. I just wanted to say that I love the idea of an online forum for participatory citizen journalism, and I'm excited to see where your project goes! I'm not sure if you're familiar with Aaron Swartz (he was a legendary programmer in the Internet's early heyday, and later became a hacktivist and advocate for open information). He did some interesting research into who writes Wikipedia, and the ways contributors contributed to it, that might be helpful while you're developing the site. Here are his blog posts about it:

http://www.aaronsw.com/2002/whowriteswikipedia/

http://www.aaronsw.com/2002/whowriteswikipedia/swartz2006

Expand full comment

Hey there. I knew of Aaron, and know a few folks who knew him. Tragic story. I’ve read a few of his posts, but not these. Will add to list. Thanks!

Expand full comment

That's so cool! I never met him, but he's a huge inspiration to me. When I saw your post, I immediately thought of his research on Wikipedia contributors. It was a simple study, but it could probably be helpful in understanding how people use wiki's and working out how to get people to engage more substantively.

Expand full comment

Another bit of useful history would be early Vox, which obviously didn't try your bounty system, but otherwise seemed really close your ideals. They focused on background-heavy "explainers" which appealed to the already-curious, and built their custom "card" system to handle continuously-updated stories. Many of the writers shared articles on Twitter, and the front page had enough traffic that just bumping articles to the front page with a new card in the stack was a useful signal boost for news. They were unabashedly center-left, partly I think because that's where they perceived the market for this kind of journalism to be, in addition to the obvious backgrounds of Matt, Ezra, and other main figures.

My understanding of what happened, according to the dribs and drabs of information mentioned by Matt, Ezra, and Vox's explainer on why they started a merch store and donation page, is that traffic never developed the way their VC-backers hoped, with Buzzfeed taking the low end (free site distributed by social media; obviously a massively more attention-hack approach) and the NYT the high end (more resources for deep reporting, good i-apps which is where the deep-pocketed readers were spending their time: see Stratechery for more on the NYT's biz model changes https://stratechery.com/2021/publishing-is-back-to-the-future/). Another factor, of course, was that after 2016 the market for center-left reporting contracted drastically (see again some Stratechery analysis https://stratechery.com/2020/twitter-responsibility-and-accountability/).

I think there are some important lessons from that history:

1. Who are the already-curious-but-underserved population you want to reach, and how does your intended editorial alignment fit with theirs? Bearing in mind, of course, that editorial alignment is more than just in-story slant, it also controls what counts as news.

2. Who is going to fund everything, and what kind of growth trajectory do they expect? Are there ways to significantly reduce your costs, like giving bounties for timely updates and corrections to wiki articles, relying on existing wiki mechanisms for dispute resolution, and just tweeting out the wiki updates, or having some kind of automated alt-frontend for wiki that foregrounds the updates (seems like this is the kind of thing GPT-3 could do).

Expand full comment

Yep. Matty addressed some of this on Twitter when he saw this: https://twitter.com/mattyglesias/status/1403347396373880833?s=20. Vox was definitely one of my conceptual influences, along with Wikipedia and Wait But Why. While I think Vox's product ultimately underperformed, there were a lot of smart intuitions behind what they did. (Though I've talked to Tim B. Lee about his time there and his experience with cards etc, I do hope he and Matty and others do a full oral history at some point about what worked, what didn't, and so on. It was a really interesting experiment.)

To your questions:

1. We're expecting two major constituencies up front: teachers/profs and folks who identify with the rationalist community (so a huge swathe of tech). Most of these folks are curious and not super tribal. And while they're not a huge number in aggregate, we don't really need volume in the early days. Being shared by those people (esp. teachers) will give us gradual mainstream penetration, as will the virality of really good writeups on hot topical things where we're just way ahead of the pack on accuracy/utility. But we won't care too much about distribution for a while as we focus on library-building. So long as we get just enough to ensure data about what's working / not working, we don't need to turn on that tap until later.

2. We're going directly for philanthropy dollars. Tech is full of billionaires. More than one of them already reads this newsletter, and we've already had a bunch of initial discussions. Money isn't going to be a problem. We're going to offer massive leverage per dollar invested. And these are very forward-looking people who see the applications here to the future of education, the importance of driving higher accuracy in journalism etc. So they're largely happy to lay out sums that are pretty trivial for them that give us the ammo we need. While it's possible that we'll take some reader donations too (a la wikipedia and this newsletter), I don't see us pushing that bit too hard.

2b. While every business/org wants to be smart about cost-control, our payouts for information flow are always going to be a not-small share of budget because that's what's going to ensure the quality moat. But we can transform that cost into value at a really high multiple as we scale, which is what readers/patrons will ultimately care about.

Expand full comment

That makes sense; glad you've thought about this and have thought explicitly about how your funding/readership models interact. News for rationalists seems like a useful niche: they have the money, they're curious and want a content-heavy product (not Youtube/Podcasts), they're unhappy with NYT et al, and a lot of them have studied more CS than history so the explainer history is going to be really helpful.

I think the teachers/profs side is going to be harder to penetrate since I think what they want is not so close to what the rationalist world wants, but as long as expectations are reasonable that doesn't seem like such a problem.

Expand full comment

Yeah the ed side is tougher. But I've gotten a healthy amount of interest over the years (re: pre-newsletter fact-checks / explainers published on Quora) from teachers/profs in subjects like journalism, media studies, and other liberal arts-y categories where they use them as case studies. So I think there's an interesting market there where maybe we can help them form a sort of go-to curriculum that introduces us to students at a really ideal age/place, where retention for those who vibe with the project can presumably be pretty decent.

We can also branch into subject primers, which is a weirdly massive gap in the market. While, to your point, some teachers/profs may not actually like what we have to say there (sometimes being at odds with convention), my intuition is that Wikipedia is a good precedent for us. If we can offer bundles of articles that work as really concise and engaging intros to a subject, we can get a lot of interest and peer-to-peer recs even if some instructors (as they did with Wikipedia) are actively telling students to look elsewhere. But we'll see when we get there.

Expand full comment