Media is a famously myopic and sclerotic industry. The big changes that take place within it often go unnoticed, at first, by the people who are paid to set its future course. Sometimes, the stuff that we in the industry miss out on is obvious to the rest of the world. We were not the first to notice, for example, that features and news stories were being cannibalized by social media, slowly at first, and then thoroughly. Many other changes start small before quickly catching fire—until suddenly you’re looking around and everyone you’ve ever met is working on a nine-episode narrative true-crime podcast.

So what’s happening now, and what’s going to happen next? In the spirit of year-end reflection, here are five stories, trends, and ideas that I think will be quite influential going forward.

5. Ryan Lizza takes “Serial” to Substack

Beginning in mid-November, shortly before Thanksgiving, the journalist Ryan Lizza turned his Substack newsletter, “Telos,” into a tell-all account of his messy relationship with his former fiancée Olivia Nuzzi, who was about to publish a memoir. He began with the story of Nuzzi’s alleged affair with the former Presidential candidate Mark Sanford before getting to her relationship with Robert F. Kennedy, Jr., and he presented everything in serialized form, with a series of cliffhangers that promised big revelations for anyone who was willing to pay ten dollars for a subscription. (A lawyer for Nuzzi has stated that the relationship with Robert F. Kennedy she describes in her memoir marks the “only instance in her long career as a journalist in which she had an improper relationship with someone she was covering.”) The series eventually ran out of steam, as Lizza seemingly exhausted his supply of venom and scandalous divulgences. (“Part 6: Bobby was behind the whole thing” appeared yesterday.) But the experiment—a serialized newsletter that asked its reader to wait patiently by their inbox for the next installment—more or less worked. I cannot remember the last time such a large group of people (in this case, mostly gossipy media professionals) were waiting around impatiently for a journalist to publish the next part of a written story. Lizza, who previously wrote for The New Yorker, turned the messiness of his personal life into something like a sporting event that needed to be followed in real time, with a crowd.

If this form takes off, what will it replace? The irony here is that the multi-episode narrative-podcast form that became popular after “Serial” is moribund now, in large part because the direct-to-consumer advertising that helped drive the podcast boom has proven insufficient to fund shows with high production costs. What Lizza did was much cheaper; it did not require a producer or hours of recorded audio. The subscription money he generated was presumably not eaten up by overhead. The serialized narrative, of course, is not a new form; it’s not even a new form on Substack, where a host of fiction writers, especially, have tried similar things. Still, I think that the high visibility of what Lizza did might lead to a small but significant migration away from audio and back to text.

The true-crime genre has been a cornerstone of the podcast market for years, and we could very well see a proliferation of newsletters about cold cases, wife murders, or gangland rackets. Sadly, this is a form that could easily be mimicked by ChatGPT, which can pull information off Wikipedia and other websites and stitch together stories that feel suspenseful.

4. An A.I. scammer cons her way into print

ChatGPT brings me to the next item on my list. In September, Nicholas Hune-Brown, a Toronto-based journalist and editor at The Local, put out an open call for stories about the privatization of health care in Canada. One of the better pitches he received came from Victoria Goldiee, a freelancer who boasted a résumé of publications that would intrigue any editor at a small but prestigious outlet such as The Local. Through some straightforward due diligence, Hune-Brown figured out that Goldiee had fabricated quotes in previous stories—sometimes from people who did not seem to exist—and concluded that she had likely used A.I. to write not only her articles but also her pitch. She did not appear to live in Toronto, as she claimed when she pitched her story to Hune-Brown. She had been deceptive in her other work.

Goldiee seems to have duped a long list of publications; the Guardian, Dwell, and the Journal of the Law Society of Scotland all retracted articles she had written for them. I do not think the editors in these places were naïve, nor do I think they made obvious mistakes that reflect widespread incompetence in the industry. And this does not necessarily augur a flood of A.I.-slop freelancers duping editors around the world—mostly because journalism pays terribly and there are better grifts to pull. But we are approaching a time when it will be hard to tell the difference between a daily feed of news generated by humans and one generated by a large language model. What happens when that line gets crossed?

Or perhaps an anxious and financially strapped media industry will simply cross that Rubicon itself, deliberately. Last week, the Washington Post launched an audio product called Your Personal Podcast, which will allow users to custom-build a daily summary of the news. According to an internal e-mail, users will be able to pick their own hosts, select their areas of interest, and even “ask questions using our Ask the Post AI technology.” Presumably, these answers will be derived from the paper’s own reporters and stories, but when you replace the names and faces that gather the news with a soothing robot voice, how will readers and listeners begin to think about the news?

3. Streamers get incentivized to talk about politics

I’ve written about this plenty already this year, so I’ll keep it short: streaming, like all disaggregated social-media phenomena, is much less democratic and independent than it might seem. The algorithm is the great determiner of success and failure, and the people who are always trying to game its secrets tend, ultimately, to do the same things. This past year, we saw something that I’ll call, in a term coined by the internet, “politicsmaxxing.” Content creators such as Adin Ross and the Nelk Boys, who only recently have demonstrated an interest in politics, began talking about the news—most notably, about Gaza. I imagine that many of these people will stop talking about Palestine and politics the instant the algorithms change; still, given the influence that these new-media forms have on young men specifically, it would not be surprising to see this switch get turned on during every major election cycle.

2. News traffic continues to decline

In October, Northwestern University’s Medill School of Journalism released its annual report on the state of local news. During the past four years, according to the report, monthly page views for the hundred largest newspapers in the country dropped by forty-five per cent. The other stats in the report are no better. The number of “news deserts,” defined as areas that don’t have consistent local reporting, continued to grow, as more than a hundred and thirty newspapers shut down in 2025, about the same number that shut down the year before.

Nobody seems to have much of a plan for what to do about any of this. Certainly, no one seems to know how to fill the need for local news—despite many efforts, which have had varying degrees of success. One possibility is that there is less demand for local news than journalists would like to believe, and that we now live in a world in which what people care most about are updates regarding Donald Trump. But I believe that the public is a bit sick, at this point, of endless Trump coverage, and that people will support local news efforts that try to meet them somewhere in their regular rounds, through the internet.

Many local journalism outlets function either formally or informally as nonprofits, funded by grants from a wide variety of donors—including, in the case of the excellent San Francisco Standard, by a venture capitalist who used to be a reporter for Time. One problem with this approach is that the direction of these publications can be subject to the whims of politicians or donors. And, in the era of the news desert, local news typically gets branded, and presented, as something essential and necessary—the vegetables in your daily diet of slop. But perhaps these outlets can find more readers by indulging in a bit of salacious crime and scandal reporting. One of the more successful local news ventures where I live is a one-woman effort called the Berkeley Scanner. Emilie Raguso, the veteran journalist behind the Scanner, almost exclusively reports on crime in the city. You can argue that focussing on crime promotes fear or can lead to other unfortunate political outcomes. But Raguso has identified a gap in local reporting and filled it.

1. Twitter is no longer the media’s village square

During the past decade, Twitter, now known as X, had an outsized and frankly grotesque hold on the journalism business. Nearly every outlet felt downstream from the social-media discourse; writers were hired on the basis of their tweets, and the coverage of the culture war that came to dominate American politics always seemed to dutifully follow what was happening in the mentions of big accounts.

This past year, the fever broke. Twitter no longer feels essential or expansive; the platform has become balkanized, fracturing into a hodgepodge of esoteric and oftentimes anachronistic conversations about housing policy, candidate polling, Marxism, and whatever else. It’s true that many people have left the platform, but I don’t think that’s why the discourse on X feels so stale. Rather, it’s more likely the product of online herding effects: everyone eventually finds a tribe and conforms to its norms.

Pew, which is on the short list of polling and survey outlets that I trust, recently put out a report on social-media use showing that women, in particular, have been leaving X. In 2018, Twitter had about equal participation between men and women; since then—and especially in the years following the company’s acquisition by Elon Musk—the platform has steeply tilted toward men. (Reddit, for what it’s worth, has had the opposite trajectory, going from a mostly male-dominated space a decade ago to something much closer to gender balance today.) I imagine there’s a feedback loop at work: X’s algorithms amplify shouting men, which, in turn, causes women to leave the platform and leads to more shouting men who believe their tribal concerns are more important than everyone else’s. X, in 2025, feels deeply self-referential and largely irrelevant.

I am not one of these traditionalists who say that we don’t need unruly public-discussion sites, because I would rather have some unpleasant chaos than a return to fully centralized media gatekeeping. Streaming, which is undeniably the ascendant form in media and commentary, is not as democratic as peak Twitter; it doesn’t allow previously unknown posters to turn themselves into the stars of an argument or a news story. Peak media Twitter was terrible, sure, but I imagine we will miss it more than we think. ♦