In many ways, the idea of threat modeling exercises is both inimical to how a free press should work and also completely necessary given, well, reality. Rosen has ariculated not just a rationale but a very clear set of tactics for newsrooms to put in place.

I’m often reminded of Karl Popper’s paradox of open societies these days, and no place does that apply more than the press. In a functioning society with a robust free press, the idea of threat modeling would seem silly or even absurd — the journalist’s role is to help readers understand the world, to cover the stories that matter most, not parse through the metalayers and bad faith. That’s a fanciful ideal that was almost never really true, of course, but with every single institution we rely on either compromised or in shambles, understanding the threats is crucial.

Substack is leading the pack and offering services traditionally filled by a publisher, including editors and even health care. Worth revisiting Ben Thompson’s profile of Substack and interview with their founders from 2019.

I love the newsletter model, even as my inbox is overstuffed with more newsletters than I could possibly ever read. I tend to forward most of them to Feedbin to read via NetNewsWire, a bit of an interface hack that doesn’t really address the problem of abundance and a la carte payment. Solving for allowing niche publishers to work at scale, which has been a basic promise of the internet for decades, while packaging in a user friendly remains the publishing problem to crack.

Barton Gellman, from the cover story of the November 2020 issue of The Atlantic:

Let us not hedge about one thing. Donald Trump may win or lose, but he will never concede. Not under any circumstance. Not during the Interregnum and not afterward. If compelled in the end to vacate his office, Trump will insist from exile, as long as he draws breath, that the contest was rigged.

Trump’s invincible commitment to this stance will be the most important fact about the coming Interregnum. It will deform the proceedings from beginning to end. We have not experienced anything like it before.

This is, in many ways, a terrifying and depressing read. It’s also vital to understand and be prepared for what is actually coming. Read this now, be shaken by the conclusions, then be prepared for what you’ll do when the worst of it comes true.

Fantastic tool and really damning indictment of how we’ve just defaulted to letting third party tracking seep into just about everything we build online (nice behind the scenes writeup about how Blacklight was built). I have intentionally avoided embedding a single tracker on this site, but as of today (22 September 2020), there are three third party cookies on my homepage because of embedded YouTube videos. I’m reminded the original spec for browser cookies said pretty explicitly that third party cookies should be rejected by default, but Netscape and IE simply failed to implement that provision. I’ll also note at the time they were ignoring their own spec, Netscape was managed by future-Facebook board member Marc Andreessen; and the one browser that seems uninterested in implementing third-party cookie blocking today is Google’s Chrome.

A truly wonderfully written piece from a woman who lived an exceptional life.

I couldn’t tell you the first time I read this seminal piece from James Fallows, but it’s one I return to occasionally, especially during election years. As someone who, like Fallows, is generally sympathetic towards the media, it is as sharp a diagnostic of how to do better things as it was nearly 25 years ago.

Really excellent insight from Josh Marshall about why tech platforms are so difficult to control. The analogy to the negative externalities generated by nuclear (and fossil fuel) power is really apt. I’ve been trying to articulate a core argument against the current generation of what we call tech and it largely comes down to a refusal to be responsible for the work they do. This is a succinct and well put version of that.

Update: Marshall followed up with a full post on Talking Points Memo that includes a great bit at the end about why new management wouldn’t fix this.

Lotta good stuff here, starting with “the Library of Congress Labs”.

They used a bit of machine learning to make the images — drawings, maps, charts, photographs — from over 150 years of newspapers searchable.