Hi all. In this exploration: I’m writing about how I’m using AI for self-guided professional development, registration is open for PMI’s June webinar, Victor Pickard reimagines and reconstructs public media, Hank Green fixes the internet, we find out which AI models are good at Spanglish, bots have now potentially overtaken human traffic on the internet, 70% of Americans play video games each week, and, finally, a (satirical?) wiki made up entirely of AI hallucinations.

I’m excited to announce our next webinar, coming June 18 at 2pET. In PBS News Innovates with Reddit we’ll address the fact that public media audiences are already on Reddit, and they're already talking about our content. The question is whether we're in the conversation. PBS News has been answering that question with a thoughtful, evolving strategy: an active subreddit community, live Ask Me Anything (AMA) events, and a multi-platform approach that's helping them reach younger, digitally native audiences without abandoning the journalism that earned their trust.

This webinar expands on their “Steal This Idea” presentation at the PBS Annual Meeting (PMI is anti-FOMO). You'll hear directly from the PBS News team about how they built their Reddit presence, what the subreddit community looks like in practice, and how they've used Reddit AMAs as a genuine journalism tool, not just a promotional one. Register here to get a candid look at what's worked, what surprised them, and what other public media organizations can learn from their approach.

AI tools can be useful for generating potential answers to problems. But they can be even more valuable in helping us organize curiosity into a disciplined course of study.

Fun fact about me: I love used bookstores and thrift stores where someone behind the scenes has an eye for quality nonfiction. In a world where nearly any body of knowledge is instantly searchable (and even academic work is increasingly legible to lay readers) a finite shelf of books shaped by taste and timing feels like an adventure. And while I don’t believe books magically arrive when you need them, I do believe you become ready for a book when you can finally recognize what it has to offer.

Case in point: A couple of years ago I was in a used bookstore in Manchester, UK, and I ran across a copy of Marshall McLuhan’s seminal Understanding Media: The Extensions of Man. I bought it as a totem, as much as anything. But it unexpectedly kicked off my collecting books about media and its effects on people. 

I’ve now accumulated dozens of physical and digital books on media, technology, philosophy, and society. And the stack keeps growing. What I’ve realized is that, at the core, I’m trying to understand the purpose (and therefore the future) of public media. Reading white papers about public media, while valuable, wasn’t getting me where I felt I needed to go. To understand public media, I need to understand media itself: what it does to us, what it makes possible, and what it changes about being human.

And that was only the entrance to the rabbit hole. Understanding media meant understanding the importance of print (i.e., the Gutenberg parenthesis), which meant understanding writing, which meant understanding the alphabet, which ultimately meant understanding language. If language is the foundational way we humans organize and communicate information, then that seems the best place to start in understanding public media.

The main problem I needed to solve was: Where to begin? I work better if I have an external structure (even if it’s self-generated). So, instead of hunting and pecking my way through a reading list I knew I would keep rearranging and expanding (shiny object syndrome is my kryptonite), I decided to use AI to organize dozens of volumes of knowledge into a linear course of study I could pursue over the next year.

Step 1: Choosing the learning tool. I chose Gemini because I’ve been increasingly using it as my day-to-day model, and because its ecosystem—especially NotebookLM—fits this kind of media-heavy learning project. Claude is excellent, but Gemini’s multimodal workflow is more useful for my purposes here. And I didn’t feel comfortable committing to ChatGPT for a long-term learning workflow.

Step 2: Giving it the sources. For this learning project, I’ve got a mix of physical and digital books to read and typing in all the titles and authors would have taken a couple of hours. Instead, using my iPhone, I took pictures of the relevant sections of my bookshelves, I took screenshots of my Kindle library, and I took screenshots of my reading wishlist on Amazon.* I then uploaded all of these to Gemini (you’re limited to 10 images at a time) with the following prompt.

Me: You are an experienced philosopher of technology and media, and a communication studies professor. I am someone trying to develop a theory of public media in America for the mid-21st century and I need help developing a self-directed reading list to help advance my thinking in an efficient and productive way.

Attached are photos of physical books I have on my shelf, titles I have in my Kindle library, and titles of interest I have saved in my Amazon wish list. Please read and consider all titles and suggest an order in which they can be read for maximum benefit to someone who is thinking about the future of public media in American society. Feel free to group titles together that might benefit from parallel co-reading. For each title in the sequence explain why it appears in the sequence where you have placed it. Feel free to take certain titles off the list, though please group those in a ‘No Need to Read’ section of your response, noting either irrelevance or redundancy as the reason.

Before you begin, do you have any questions? 

Pro user tip: I always like to ask the model if it has questions, because it opens a door to further clarification. This is a good technique to use no matter what prompt you’re crafting. 

Gemini replied:

G: Does your developing theory lean more toward the structural and institutional (how we fund, govern, and protect public media from commercial capture) or the ecological and cognitive (how the transition from a literate to a digital-oral culture fundamentally changes what it even means to be a "public")?

Me: Both. Ultimately, I want to develop a theory about the structure of the institution, but I believe that the transition from literate to digital-oral culture could significantly shape media in the next 25 years. 

💬

Step 3: Correcting the first pass. The initial reply (not shown here) was fine, but not comprehensive enough. So, I had to give it some corrective instruction:

Me: Thank you. This doesn’t cover every title available to me in those photos that I shared with you. I’m looking for you to place all of the titles in a sequence or discard them from my reading list. For example, there are three titles in there by Lewis Mumford, which are of interest to me, but which did not make your list. Please build off the previous response but this time include all of the titles you see in the images already sent to you. Again, it is OK to take some of the titles off the reading list and place them in a “no need to read” section. 

💬

The new reply (not shown here) was more what I was envisioning, so I told Gemini so and tacked on a further request:

Me: That is much better. It is the type of list that I was hoping to get from you. Do you have any other authors or titles you would suggest for any of the phases you have articulated above? If so, please restate the list with your recommended phases and include a section at the end of each under the heading “I’d also recommend.” 

💬

Step 4: Preserving the context. After Gemini delivered on my instructions (not shown here), I went on to ask if its recommended order was sacrosanct or if I could jump around. Then I asked which 5–10 books it considered core to the collection. I also asked if I were to use a ‘parallel pair’ approach to how it might pair up the titles on the larger reading list. I was satisfied with its suggestions for each, so I finally had it convert its suggested six phases of reading into a Google Sheet for me. 

One of the issues that you can run into having an extended exchange with any of the major chatbot models is “context drift.” Basically, the bot begins to lose the plot. To address this, I asked Gemini for guidance:

Me: I'd like you to be my discussion partner for these readings. Should this happen in this thread? Or should I start a new thread? Or should I start a Notebook? What do you recommend?... [Also,] can this chat thread be exported in a way that I can upload it to a new "Media Theory Seminar" thread as part of the initial context setting?

Gemini then “compiled the essential context, your structural reading rules, and the complete phase breakdown into a clean, ready-to-copy block below. [Telling me] you can copy this entire section and use it as the exact script to launch your new thread.” It essentially crafted my first discussion prompt for me, importing the essential knowledge from the reading list thread. I’ll use this technique whenever I start a new thread to control for context drift.

My hope is that you can adapt this approach to whatever question you need to think through more deliberately. For those of us in public media, this may be one of AI’s more practical near-term uses: not replacing expertise but helping people build it with more intention. The real opportunity is turning curiosity into structure, structure into sustained learning, and learning into leadership.

In my next column, I’ll explain how I’ve started using NotebookLM, along with chat threads like this one, to add extra organization to my learning experience. 

For now, on to the links.

_{*When I’m ordering physical books I can’t find used, I often use ThriftBooks.com or Bookshop.org. Amazon is simply a convenient place to keep a visual reading list on my phone, since it shows cover thumbnails and ISBNs. I try to support independents when I can.}

Reimagining and Reconstructing U.S. Public Media (Victor Pickard - Knight First Amendment Institute at Columbia University) - There are a small handful of academics at America's universities that care deeply about the future of public media and are beginning to codify their visions for what we should become. The problem is that those visions aren't always anchored in the realities of actually running a public media institution. This "analysis begins to sketch out the structural conditions necessary for creating a new public media system anchored by what [Pickard] term[s] the 'Public Media Center.' Broadly speaking, three organizing principles—de-privatization, democratization, and localization—can help guide us [he says] toward reconstructing our public media from the ground up." It's a nice vision, but there's a ‘rip it up and start again’ subtext that will make it a non-starter for most of you. Can we build something beneath the surface that is ready to inherent what we already maintain once the status quo is gone? Or do we all really think that we can just get back to business as usual if we hang on a little longer. I mean, I know which camp I'm in. How about you?
Related: This On the Media interview with Pickard has been sitting in my open tabs since March. Combine his Knight piece above with the headlines about 60 Minutes this week, and it seems like a good time to share: The Century-Long Capture of U.S. Media

Public Media Doesn’t Just Need Defending. It Needs Redesigning. (Andrew Ramsammy via LinkedIn) - This piece reacting to Paula Kerger's keynote got a lot of buzz when it was posted immediately following the annual meeting. I've been wrestling with what to say about the PBS Annual Meeting because, unlike Ramsammy, I was there and got to hear the tone and cadence of the messaging. But all my thoughts have been cynical, so I haven't penned that column yet. I am broadly supportive of Ramsammy's take on the speech, but a lot of it could have been pre-written. Paula's keynotes do follow a formula, and that formula exists in large part to rally the station staff in the room. It’s not really the space to issue declarations of media revolution. I’ve seen PBS leaders do that. OGs will remember Pat Mitchell's "Declaration of Interdependence." It was good in concept (probably more relevant now than ever) …and the room hated it. The people in that room want comfort, not conflict. And that's what that speech is constructed to do. So, let's take as wrote that reacting to Paula can be a easy hook, but its Ramsammy's points about articulating a new value proposition for public media, moving beyond TV and radio, and celebrating entrepreneurship all make this a piece worth a read.

Before You Merge: A Blueprint for Public Media Consolidation (Yoni Greenbaum - Backstory & Strategy) - This is an interesting way of slicing and analyzing the public media systems. It's a long read, but if you're not in the rooms where a lot of these decisions are being made, you'll find the analysis informative.

How to Fix the Internet (Hank Green - Time) - Here's a good manifesto from friend-of-public-media Hank Green. This piece isn't about answers; it's about focusing on what matters. The good news is that, by just being us, we can help fulfill his vision. But we have to be more focused on putting out products (not just content) that project our values into society.
Related: Mark Tyson reports in Tom's Hardware ‘Bots have now passed human traffic online,’ Cloudflare boss laments — says agentic traffic wasn’t expected to eclipse real people until next year

Creator journalism is the most disruptive shift the news industry has seen, ex-BBC News head says (Laura Hazard Owen - NiemanLab) - This quote spoke to me because it echoes what I'm thinking about in my own studies (confirmation bias FTW!): "This moment of disruption is so potent because it goes to the heart of how the relationship between news provider and news consumer is shifting from institutions to individuals, from big media brands to personalities, from PSBs to independent journalists, all with dramatic consequences for where news consumption is collapsing, and where it’s growing at speed…If we’ve been wondering for years what would eventually replace the broadcast news mass media model, I think we’re seeing the answer now. These new forms of journalism are taking the time, the loyalty, and the trust that consumers used to invest in big, mainstream news providers, and they’re moving it to new platforms." What I hear there is that this isn't about FAST channels or getting into Samsung's ecosystem. That's fighting the last war when media was a universe, not a multiverse. What needs to happen next is co-creation with personalities and influencers.
Related: Jill Manuel's piece for TV News Check: What Counts As Journalism — And Who Gets To Decide?

The death of Google Translate Spanish (Ernesto Aguilar - OIGO) - Ernesto (friend of the newsletter) usually writes good pieces. I found this one especially valuable because we have a lot of good tools at our disposal, but not every tool excels where you need it to for your discipline. I especially like his ‘Spanglish test,’ so his recommendations are worth noting.

AI licensing coalition SPUR in huge expansion (Charlotte Tobitt - PressGazette) - Interesting project here. I wonder if they'd accept local US public media organizations?
Related: This piece in The Media Copilot underscores the need for initiatives like the one above: New report finds wide disparity in AI tollbooths for publishers

Google Search as you know it is over (Sarah Perez - TechCrunch) - Good headline, but if you touch base with the Google I/O announcements every year, you've seen Google slowly iterating its search products (and us) in this direction for a while. But even if this isn't a seismic moment, search is changing, and public media does need to change its discovery tactics and its content packaging to respond to those changes.
Related: For those not ready to go quietly into AI Mode, Perez also recently reported on how DuckDuckGo makes its ‘no-AI’ search engine easier to access as its traffic booms
Also Related: Barry Schwartz reports in Search Engine Land that Microsoft releases Web IQ, powered by Bing but designed for how AI-agents search. Seems worth noting, and I’m interested in how AI agents consume the web differently from humans, but it also sounds vaguely like what Google was doing a year ago.

New opportunities, control and insights for website owners (Mrinalini Loew - The Keyword) - Google’s dominance of search doesn’t completely rob us of agency. If you haven’t explored a campaign to get your organization designated a preferred source by your public, you should if for no other reason than that Google "recently brought Preferred Sources into AI Overviews and AI Mode and launched new subscription labels in these features, so people can choose the websites that they want to see more prominently."
Related: More on how consumers can set your public media organization as a "preferred source," also from The Keyword: New ways to find your favorite sources and original content in AI Search
Also Related: Meanwhile, across the pond the BBC reports that Publishers in UK can opt out of Google AI search results, a feature that Dominic Ponsford reports in PressGazette will soon be available globally.
Still Related: Matt Southern's piece in Search Engine Journal (h/t/ to Collin Berke): AI Crawlers Are Reportedly Draining Site Resources & Skewing Analytics

Cloudflare Reports Bots Generate Majority of Web Traffic (AI Summary - Let's Data Science) - Two things. First, even if the advent is debatable (see below), it's coming…and we need to think about what that means. Second, this report (and I presume most of the site) is AI generated. It contains a "Relevance Score, a "scoring rationale" at the bottom, and "sources" a la Google's AI Mode. It's worth clicking through to see how they are executing this...and we need to think about what that means.
Related: This discussion thread on HackerNews: Cloudflare CEO is lying to you about the bot traffic jump

AI Is Already Shaping Who We Are (Laura Manley - Time) - I remember, years ago, the CEO of Ford talking about how cruise control was the gateway drug to self-driving cars. The point, it seemed, was AI infiltrates your world in a myriad of tiny ways. And ultimately, all those ways add up until it shifts from luxury to necessity. This piece takes a similar notion and focuses it on media consumption, asking whether or not we even know our own minds anymore.

Eroding a virtue: AI trains people to expect instant answers – and that’s bad news for patience (Christian B. Miller - The Conversation) - Is patience maxxing the new digital detox? I will say, just using myself as the experiment, that I do find that when AI gives me lengthy responses to my prompts I regularly wish it were shorter. (This from the guy sending you a 4000+ word newsletter. 😳) I know I can prompt it to be more direct or make everything a summary. But the length of response is often the point. And yet.

What Leaders Get Wrong About the ROI of AI (Katy George - Time) - This isn't necessarily an exact prescription for public media, but her broad guidance for knowledge workers can be easily adapted by anyone who knows our business. Plenty of good thoughts in this one, but this line especially felt worth elevating: "The most important question isn’t where to deploy AI. It’s what outcome matters most to your business, and whether AI can help move it."

AI sticker shock hits corporate America (Madison Mills - Axios) - turns out the corporate money for AI isn't endless. Maybe this will finally pop the AI economic bubble, as the gold rush mentality of the last year comes up against the dry creek beds of token budgets. Going forward, your organization is going to need to budget for AI-anchored compute. So now is a good time to learn corporate America's mistakes and figure out how to appropriately allocate compute for public media projects.
Related: Bradley Olson's piece in WSJ: Corporate America Is Starting to Ration AI as Cost Skyrockets ($)

Report: Broadcast Employment Hard Hit by AI (Tom Butts - TVTechnology) - I'm including this one primarily because I think it might very well be BS. The math here seems really fuzzy, and while I don't doubt that just under 12,000 broadcast jobs were lost in the two-year period cites (that seems to be government data), attributing those to AI is lazy logic at best, and possibly blatant fear-mongering. However you might feel about AI's potential impact on employment, this piece is a great example of why you can't take headlines like these at face value.
Related: Benedict Evans has a good piece on his blog discussing the complexities of Predicting AI job exposure

Everybody hates AI (Brian Morrissey - The Rebooting) - I agree with a lot of Morrissey's views on the current AI landscape. That sentiment has swung to the negative was predictable, even if the timing was not. But as accurately as he maps that negative sentiment, this is the line that I think is the kicker: "The resistance is likely futile in the long run – there’s too much money and power to be accumulated – but the pressure building from various corners of society will have an impact on the diffusion of AI. This is how societies work: messily. The AI YOLO era is over."
Related: Charlie Warzell at The Atlantic had a good episode of his Galaxy Brain podcast focused on Why Everyone Hates Data Centers.

Nearly 70% of Americans Play Video Games for at Least an Hour Each Week, New Report Finds (Jennifer Maas - Variety) - I'll say it again, games are a legit medium for public media. And not just kids’ media. We are not serving our public if we are not making games.
But You Don’t Have to Take My Word for It: Download and read the ESA’s Essential Facts About the Video Game Industry (or drop it into your favorite Chatbot and explore the report) yourself.

"Nobody's making games for the retired people" – The growing yet underserved market for grey gamers (Lewis Packwood - Games Industry Biz) - Say it with me: "Our 👏 audience 👏 plays 👏 video games. 🎮" After every presentation we do at a public media conference about the inherent opportunity for public media in videos games, someone inevitably says in the session feedback that they don't know why stations would make games. This highlights one major reason. And I know those critics are probably not reading this newsletter, but if enough of us put our shoulder to the wheel of making games a legit form of public media, then the culture in public media will change.

A Deranged New Wikipedia Clone Is Made Entirely of Surreal AI Hallucinations (Frank Landymore - Futurism) - And finally, clever little idea here. I find it helps if you convince yourself it's an art project or bit of satire.