Hi all, and welcome to our new readers from Maine, Detroit, LA and Arkansas! This time my column digs into what comes after the filter bubble, plus we’ve got stories on how publishers should disclose AI use, the New York Times as a gaming company, and finally, The Onion is bringing Infowars Back.

We’ve just announced our next webinar, July 16 at 1pET. How Public Media Companies are Building AI Policies is our quarterly collaboration with Current and in this webinar, presenters from several public media stations will share their experiences developing AI policies from the ground up. We'll cover what drove each station to act, how they structured their policies, and the lessons that came out of the process. Panelists will also share their policies directly, so we can all leave with something concrete to build from. This one might include your friendly neighborhood public media newsletter author, and whether that sways you one way or the other, you can register now!

I spent a long time listening to podcasts this past weekend. Often, I use weekend time to read and write, but this weekend the large block of ice that (ironically) turned my fridge into more of a swamp cooler had other ideas. So, as I sat leaning against the struggling appliance, directing the outflow of a borrowed hairdryer into the back of the freezer, I had the perfect opportunity to catch up on my backlog of podcasts.

Listening to Semafor’s Mixed Signals podcast (which you can watch above), I was surprised to find out that Eli Pariser is back. Many of you may not remember Eli, but you have heard of “the filter bubble,” the term he coined in 2011 to describe the algorithmic isolation that was shaping our media consumption. Those were slightly more utopian times and algorithms seemed like they could be a force for good (remember NPR live tweeting the Arab Spring?), helping us reconnect with friends and more easily find the media that we all wanted. Pariser is a creature of politics who also founded MoveOn.org and Upworthy, so he is intimate with the social media space.

Of course, social media ain’t what it used to be, and how we interact online is shifting again thanks to AI. 

Flash forward, and Pariser has now founded New_ Public (everyone frown quizzically at the name) and has a new take on the world of algorithmic media. It’s a theme we’ve highlighted here in comments and articles, where information becomes increasingly mediated by AI. New_ Public argues we are entering the “agentic interface era” (AIE). In many ways, the idea is an extension of the filter bubble, where AI scopes down all of the world’s information to just the information that supports your beliefs and tastes (both positive and negative).

The difference, as I see it, is that in a filter bubble world, there is still action and intent on the part of the media consumer. You had to go out to search, or to social media, with some type of intent (even if that intent was, or led to, doomscrolling). With AI, they caution, we will become more passive recipients of information thanks to AI summaries (both the search kind and the agentic assistant reports, like the Gemini Daily Brief, that companies like Google want to provide you).

Definitely take a look at the deck, but here’s food for thought for public media:

You can see New_ Public’s deck on the topic here. The headline for me is that many of our mission statements point to how ripe our service is to thrive in the type of future New_ Public is describing. Maybe in some places you do have to squint to see it in this deck, but then as innovators isn’t that kind of what we do?

Okay, on to the links.

Public Media’s Promise Was Never the Problem (Julio Ricardo Varela - Pressing Issues) - I'm working on a column to be published sometime this summer about how the stakeholders for public media are much broader that we typically acknowledge amongst ourselves. So, this piece tripped a very particular wire for me. We don't tend to talk about the fact that when public media first solidified into PBS and NPR, there were no legacy institutions with entrenched power and responsibilities. Put another way, there was no one who had something to lose. Now we have an old guard with a power base and skin in the game that is somewhat tied to the religion of broadcast and universal service, but really mostly tied to a personal solvency threshold for retirement. No disrespect intended for those who have served public media for years. We just need an innovation bypass so that a system is waiting for public media when the old guard doesn't need to guard as much.

KRCW's Jennifer Ferro on building a community model (Brian Morrissey - The Rebooting) - If you tuned in to our 2026 or 2025 "Future of Public Media" webinars, you know that we aren't shy about looking to KRCW as an aspirational model for community activation. Nice to see some major media podcasters finally keying into what we've known for a while.

How should news organizations label their AI use for audiences? New studies suggest some answers (Mark Coddington & Tamar Wilner - NiemanLab) - The big upshot was that people first, people last is still a good mantra, especially people last. I thought this line was the heart of the 'Interviewees said there’s a vital difference between an article written entirely by AI (which is the origin they assume when labels say “generated” or “made by,”) and one just aided by AI (indicated by labels like “assisted” or “in conjunction”). They were concerned about AI’s capacity for hallucination and bias, so they saw human review as essential, and wanted labeling to address that need. Unlike in the qualitative study, participants had particular concerns about visual content, saying labels were especially needed here.'

The information ecosystem is being redrawn by AI. That might be good news (Shuwei Fang - Reuters Institute) - There are two lines from this piece that I want to surface as an enticement to read the whole thing. First, "Now AI systems are inserting themselves as active intermediaries....It is tempting to see this as another platform shift....But the mechanism is fundamentally different." And then, "Every information revolution in history has expanded who could participate.... AI could be the next jump, but its distinctive contribution is not access. It is comprehension." I'd bookmark this one or stick it in NotebookLM. You'll want to revisit it periodically.

The consequences of relying on AI for accurate news (Adam Conner-Simons - MIT News) - At first the picture looks bleak. But it turns out, it's how your use AI that matters. This line from the piece is key: "...the Media Lab team uncovered several strategies associated with stronger independent detection later on, even if the strategies initially slowed down performance during the interaction. This included the Socratic method of the AI asking guided questions, as well as so-called “deep probing,” where the system provides gently persuasive statements if the user appears to be veering away from the correct response."

In 2026, Less than One Third of Google Searches Still Send a Click (Rand Fishkin - SparkToro) - I see data like this as a call to better understand where media consumption is going in the future. While "Google Zero" is a pithy way to sum up the shifts happening in search thanks to AI, I don't think we'll ever actually see 0. What I think will happen before that is that the mode of consumption will change so much that referral traffic will cease to matter. It will no longer be the coin of the realm.

Google introduces Search profiles within Google Discover (Barry Schwartz - Search Engine Land) - More changes from Google for publishers interested in trying to shape how they appear in Search and Discover.

How to Track and Report on Traffic from AI Tools (ChatGPT, Perplexity) in GA4 (Dana DiTomaso - Analytics Playbook) - We've been wrestling with how to accurately gauge the digital performance of our websites in the era of agentic AI. The data has gotten very dirty. If you have suggestions, send them my way, but in the meantime our NPM's Director of Research sent this my way as a source they are using to trying and filter for more clarity.

Americans and AI 2026: Chatbots, Smart Devices and Views on Impact (Jeffrey Gottfried, William Bishop, Monica Anderson, Michelle Faverio, Eugenie Park & Colleen McClain - Pew Research Center) - This report is from a survey of 5,119 U.S. adults that was in the field February 17-23, 2026. The updated data is useful. I was especially interested in the fact that about half of surveyed respondents use chatbots, and 24% use it daily. But this is less surprising when you consider that 42% use AI to search for information, and search has pretty thoroughly baked AI into its experience. Interestingly, only 13% use it to get news.

Three ways to avoid being fooled by AI slop ( Silvia Montaña-Niño & T.J. Thomson - The Conversation) - Good tips here. I don't want to steal their thunder, but in brief, know how AI slop presents itself. Then look for details that feel off. Then consider the source, and what trusted sources say about the source.

Your AI habit is wasting precious resources. Here’s how to use it responsibly (Seyedali Mirjalili - The Conversation) - These tips closely resemble the responsible AI use appendix that I created for our updated AI policy (dropping next month). I worked those up a couple of weeks ago, and y’all know how I loves me some good confirmation bias. So, this one is worth checking out.

5 ways data centers endanger their local communities and the country as a whole (Neha Gour, Ed Maibach & Luis Ortiz - The Conversation) - This is a link heavy piece that I'm including for those who want to do a deeper dive into data center issues. Despite the alarmist headline, there is a small solutions section at the end that I found was sensible. So, the thrust of the piece is less 'shut 'em down' than it might seem at first.

Landmark German ruling declares Google's AI Overviews are Google's own words and makes it liable for false answers (Matthias Bastian - The Decoder) - It's worth keeping an eye on this one. While it's hard to imagine that anything like this would take root here, European regulations can have nudging effects on how we use global tech in America. And the idea that a corporation is responsible for its AI speech is an interesting twist on corporations having the right to free speech (since the Supreme Court has ruled that in the US, ‘corporations are people too’).

ChatGPT’s market share slips below 50% for first time (Ivan Mehta - TechCrunch) - I know I've been cooling on ChatGPT, though there have been a couple of instances lately for me where its efficiently outperformed Gemini. Guess I'm not alone.

US court rules Ohio can restrict children’s use of social media (Alina Maria Stan - The Next Web) - Interesting to see the issue of age-restricted social media use come home to roost in the USA. Australia has already banned social media use by young people, and the UK is making similar noises. Of course this is simply requiring parental consent for usage under 16, not a ban. But it's a start.

Substack is launching a sponsorship program (Neel Dhanesha - NiemanLab) - Substack has been making obvious moves to be more social media than newsletter platform. This move toward being an ad network is another plank in that platform.

Mastodon looks to newsletters to help revive the open social web (Sarah Perez - TechCrunch) - This type of move by Mastodon, a focus on the more private space of the inbox, is one that I think illustrates the possible future outlined by the New_ Public deck featured in today's column above.

Has the New York Times Become a Games Company? (Stephen J. Dubner & Theo Jacobs - Freakonomics Radio) - H/t to Adam Dylewski for sending this my way. Longtime PMI members know that the answer to the question in the title is "yes." As the quip goes, NYT is a games company with a news problem.

Paramount Skydance Launches New Video Game Studio as ‘Core Pillar’ of Content Strategy Alongside TV, Films and Streaming (Jennifer Maas - Variety) - I know Paramount's rep isn't a great one these days, but I offer as one more example of how the media landscape is tilting towards games.

Courts be damned, The Onion is starting its InfoWars next month (Matt Schimkowitz - AV Club) - And finally, this story still makes my heart glad.