Notable links: May 8, 2026

Most Fridays, I share a handful of pieces that caught my eye at the intersection of technology, media, and society.

Did I miss something important? Send me an email to let me know.

An Invitation to Build the Civic Information Economy

This feels like a vital exploration to me:

“What does it take to fund civic information? We often focus on “more money,” but we limit the field’s potential by ignoring better capital design. Today’s landscape often sees dollars concentrated in intermediary and national plays—often with good reason—but without deliberate examination, we risk stifling the imagination of a developing field.”

This is part manifesto, but also part call to action and invitation to participate. Funding for news has often been reactive, filling gaps, but what does it mean to intentionally design a genuine ecosystem with dynamics that support production of, and access to, the civic information we need that is a prerequisite for democracy to function?

I love how genuinely participative this is: rather than a bunch of people trying to be smart inside institutions, this requires that the people who struggle the most to find information are active co-designers. That feels non-negotiable to me. There’s been a pushback against inclusion and equity in news and everywhere recently, but there’s no other way to build an ecosystem that genuinely serves everybody. We’ve all got to own it. We’ve all got to take part. Everyone needs to be represented.

I have high hopes for this, and I love that the effort exists. It’s something we should all support.

Some Rationalization May Finally Be Coming for Newsroom Intermediaries

In his latest post, Dick Tofel talks about a need for consolidation in organizations that support newsrooms (and, in fact, in newsrooms themselves).

“Devoting limited resources to competing services where one offering is superior not only leads those using the inferior service to poorer results, it also subsidizes entirely unnecessary administrative costs at the inferior service. And in circumstances where competing services are roughly equivalent, mere duplication can also be inefficient—and, as noted above, may place an administrative burden on already over-stressed client newsrooms. Time is one of the scarcest resources of all.”

I’m worried.

This isn’t a criticism of Dick Tofel: he calls out the benefits of competition and the difficulty of determining winners in a market. But I do think there are two more things to consider.

The first is that I don’t believe any intermediary service designed for newsrooms is optimal. That’s not a criticism of them, either: every service has room to grow. Any time you remove competition from a market and hand it to a single privately-owned player (nationalized services are another thing entirely), the offerings stagnate because the driver to improve has gone away. Just ask anyone who remembers the web’s Internet Explorer wilderness years before Firefox disrupted them and forced widespread standardization.

The second, and probably most important, is that funders are a narrow group of people with a narrow set of perspectives. Unless they’ve done the work to be representative and inclusive in their work and culture, they may miss how one service serves a community better than another and erroneously mark them as duplicative. Or to put it another way, if there is any consolidation in any American market, I don’t trust that organizations run by women and people of color won’t be the ones to lose out.

This isn’t anyone’s intention, but reducing competition at any level — funders, intermediaries, newsrooms, distributors — has the potential to create monopolies that become gatekeepers for vulnerable communities who need more support, not less. I don’t think that’s what the moment we’re living through needs. We need more ideas, more approaches, more funding, more communities served, and more diversity. The people who want to shut down an effective, independent press want to create a monoculture. The way to combat that is not to create another one.

Report on Burnout in Open Source Software

This bleak report on burnout in open source software communities from last year has been doing the rounds. I think it’s clearly indicative of where open source is at (and its trajectory), but the solutions aren’t immediately clear — we know this because similar concerns have been anecdotally highlighted by various people for well over a decade. However, this is the first formal research report I can remember reading.

It’s pretty stark: 60% of open source maintainers work unpaid, 60% have quit or considered quitting, and 44% cite burnout specifically.

From the author:

“In my report, I draw upon a combination of academic literature and OSS community discussion to identify 6 factors that contribute OSS developer burnout: difficulty getting paid, workload and time commitment, maintenance work as unrewarding, toxic community behaviour, hyper-responsibility and pressure to prove oneself. I then make 4 broad recommendations for how to address it: pay OSS engineers, foster a culture of recognition and respect, grow the community and advocate for maintainers.“

The thing is: who is going to pay OSS engineers? Every attempt to get downstream users to pay out of the goodness of their hearts has failed at scale. There are certainly corporate sponsors of OSS maintainers already, when there’s a clear link between an open source project and a company’s bottom line. That could certainly be a broader standard, but there are also a ton of open source projects that tie less obviously into bottom lines, or are useful for communities outside large corporations.

Beyond the lack of direct compensation, it’s also a pretty thankless job. Downstream users will often make demands of maintainers that don’t take their contexts into account; people who are unpaid and overworked find themselves treated as if their users had paid them large sums of money. Open source users communities can sometimes also be wildly detached from reality; I’ve had people tell me that the solution is to move beyond a money-driven society or to have a revolution to overthrow capitalism. I mean, sure, whatever, but is that going to happen tomorrow? In the meantime there’s rent to be paid and food to buy.

The answer may be that, actually, open source doesn’t work very well except (1) in mutualistic co-operative communities (2) as a strategic move by corporations who want to change the markets that affect them in some way (3) as a self-promotional way to gain more widespread recognition for your abilities. If that’s the case, maintainers should go into it with open eyes: if they’re successful, their work will be used by people who make a bunch of money without compensating them in any way.

If new maintainers understand that they’re voluntarily signing up for extraction, I think it’s uncomplicated. If they think they’re going to “win” open source and make a living by giving their work away for free, they may be sorely disappointed.

I can never talk to an AI anonymously again

On first read, I found this a little bit sobering:

“But soon, the entire debate over internet anonymity will be as anachronistic as an iPod Touch. That’s because Claude Opus 4.7 is here, and last week, I discovered it could identify me from text I had never published, text from when I was in high school, text from genres I have never publicly written in. And if it can identify me, soon, it will be able to identify many of you.”

Kelsey Piper plugged in 125 words from an unpublished essay and asked Opus 4.7 to identify the author. It listed her as the most likely answer.

I was skeptical, so I did the same: a few paragraphs from an essay, which I plugged into an account that isn’t connected to me. Here’s what Claude said:

“Based on the style, themes, and specific concerns in this snippet, my best guess is Ben Werdmüller.

Several signals point this way: the framing of the open social web as a movement rather than a technical project, the explicit linkage between Mastodon/Bluesky, journalism, and broader liberatory politics, the critique of newsroom demographics ("very white, very coastal, wealthy"), and the historical framing of the early web as an anti-gatekeeping movement. Werdmüller writes frequently about all of these topics together — he works at ProPublica on technology, has a long history in the IndieWeb and federated social space (he co-founded Elgg and Known), and his personal blog regularly weaves these exact threads.

The voice also fits: declarative, slightly polemical, comfortable making sweeping movement-level claims, and willing to name class and racial dynamics directly rather than hedging.”

I’ll take that as a compliment! But it’s impressive that it was able to figure it out without searching.

But those are topics I write about regularly. When I gave it a snippet of my short fiction piece Sharing is Caring, it wasn’t able to find a match:

“The passage has hallmarks of contemporary speculative/dark YA or literary sci-fi short fiction — body-swapping or consciousness-transfer premise, a class-inflected ethical horror (poor family selling a daughter's body to a wealthy disabled girl's family), tight close-third POV, terse interior monologue in italics-style cut-ins ("Sure. No. Not a problem."), and the bleak ironic kicker about an eighteenth birthday. That cluster of moves shows up in writers like Ted Chiang, Ken Liu, Carmen Maria Machado, Alice Sola Kim, Karen Russell, or in anthology pieces from venues like Lightspeed, Tor.com, or various YA dystopia collections — but none of those is a confident match.”

Again, I’ll take the comparisons as a compliment, but there isn’t enough of my fiction out there to make a real suggestion. That’s going to be true of most people’s writing — unless they’re publishing really regularly on the web, and have been for a while, the models might not know about them. Kelsey had a slightly different result: a school progress report was still identified as hers. But clearly, the reliability here is shakier.

In other words, we’re a little way’s off from AI being able to identify most people quite yet. But if you’re publishing regularly, out in the open, you should expect your work to be fingerprinted. That does have implications for dissidents and people writing anonymously who have previously done so under their real names.