Algorithms Decide What Goes Viral. Journalists Are Still Chasing the Truth.
At SXSW, I sat down with the man Elon Musk sued to talk about who really controls what we see online—and why journalism may be losing the fight.
Welcome back to The Red Letter.
I spent last week at SXSW in Austin talking about a question that sounds abstract but actually shapes almost everything we see online: Who owns the truth?
That question brought me on stage with Imran Ahmed, the founding Chief Executive Officer of the Center for Countering Digital Hate — the group that Elon Musk famously sued after it published research claiming hate speech surged on X after his takeover. A judge threw the case out, but the fight over the future of the internet — and who controls it — is very much alive.
The room was packed and the conversation we ended up having wasn’t really about politics. It was about power — specifically the quiet power of algorithms.
I came to the panel from the perspective of someone in my second year building an independent journalism company in real time. For the last year, I’ve been reporting, publishing, podcasting, and experimenting with new ways to reach readers. But the truth is something I didn’t fully appreciate when I worked inside legacy media institutions: even independent journalists aren’t really independent. We depend on algorithms.
As I told the SXSW audience, I didn’t fully grasp how much distribution had shifted from editors and homepages to recommendation systems — the invisible formulas deciding what appears in your feed. If a story doesn’t surface on those systems, it might as well not exist and that creates a strange tension.
Investigative journalism is slow, expensive and methodical. It requires verification, sources, documents, time. But the platforms where most people encounter information now reward speed, outrage, and emotional reaction — often before facts are fully established.
I can spend weeks reporting a complicated investigative story. Maybe it lands well on Substack for a few hours. But then someone with a massive following reframes the same story in a more emotional or sensational way and the algorithm rewards that version instead. And then suddenly the narrative belongs to them.
The incentives are obvious. Platforms are businesses competing for attention. The longer users stay engaged, the more ads they see. And outrage, fear, and conflict keep people scrolling longer than nuance or context ever will.
Imran described this bluntly: social platforms aren’t really public squares — they’re advertising billboards designed to capture attention. And that attention economy shapes politics in ways we’re only beginning to understand.
During our conversation I talked about interviewing James Fishback, a candidate for governor in Florida, who has essentially built his campaign on algorithmic outrage. He thrives on provocation. The more people criticize him, the more engagement he gets which pushes him higher in feeds and gives him more visibility. That isn’t an accident.
Extremists have understood this dynamic for years. One far-right playbook Imran’s researchers found essentially says the quiet part out loud: say the most outrageous thing possible, because even criticism will amplify you in the algorithm. Visibility leads to normalization and then normalization moves the fringe into the mainstream.
If that sounds familiar, it should because it’s basically the story of the last decade. But the conversation at SXSW wasn’t entirely pessimistic. There is also an upside to this moment.
The same platforms that reward outrage have also opened the door for independent journalists to build direct relationships with audiences. For decades, the public only encountered reporters through large institutions like newspapers, television networks, cable shows. But now journalists can speak directly to readers.
The tradeoff is that trust has to be built one reader at a time. That’s why I’ve come to believe the only sustainable model for journalism in the algorithmic age is community. Not viral reach. Not chasing every trending topic. But building a core audience that trusts your reporting and comes back for it.
As I told the SXSW audience, journalism used to rely on institutional credibility. Now the burden falls on individual reporters to prove their credibility directly to their audience — to show their work and maintain trust in real time. It’s harder and it has to be more honest and transparent.
The deeper question we kept coming back to is whether democracies can function in an environment where the information landscape is constantly manipulated by opaque algorithms. Right now, we don’t even know how those algorithms work. We just feel their effects.
Imran argues the first step is transparency, which means forcing platforms to reveal how their systems amplify content and shape public conversation. Without that visibility, we’re essentially navigating the information ecosystem blind.
I’m not sure regulation alone will fix the problem but I do know this: journalism that prioritizes truth over engagement is more important than ever.
The algorithms may decide what goes viral but they don’t get to decide what’s true. That part is still up to us.
What a fantastic interview. I only worry for you because I can't imagine 30% of any group of people hating you! We must push our politicians to make the algorithms transparent and maybe it's time for good old fashioned PSAs about the algorithms. Then maybe enough people wake up from hate scrolling and go be with people again. Thank you for all you do.
You have a common problem of "Independent" Journalists. You need to provide all of the services that were previously provided by staff at traditional media, such as editing and fact checking. Unfortunately, the arithmetic of subscribing to individual journalists makes it impossible for people to subscribe to very many. I would advocate forming journalist collectives were you pool your resources and revenues to hire the technical support you need and divide the income in an equitable, mutually agreeable way. Propublica and the States News Rooms may be models.
The Contrarian is another example of how to do something like that. The difference between the Contrarian and what you do, is that in addition to covering the news, they use their revenues to fund law suits against the administration.