Hi all. This time I’m reflecting on public media a year after federal defunding, we indulge in a little PBS nostalgia, Google continues to terraform the internet with AI, how journalists are adapting to AI, the battle over AI nudification continues, prediction markets shape reality, and finally, an embarrassing example of how little AI (and the US) knows about Africa.

On August 20 at 1pET/10aPT, we’ll host our next webinar: After the Feed: AI and the Future of Community. Here’s the description:

Social platforms are struggling to keep people engaged, and AI is quickly becoming the new way people find information and each other online. For public media organizations built around community and trust, that shift raises real questions: Will people still come to us, or ask AI instead? What happens to the relationships we've built with our audience and our members?

Angelica Quicksey, Managing Director of New_ Public, will share her organization's latest research on how AI systems are becoming the new gatekeepers of what people see and trust online. She'll walk through why audiences are shifting away from public, algorithm-driven feeds toward smaller, trust-based relationships with creators and journalists, and what that means for public media stations.

This session is aimed at station leadership, community engagement, and development teams thinking through what this shift means for our audience relationships. Read New_ Public's full report at newpublic.org/afterthefeed

There are implications across the station org chart, whether you’re in content, MarCom, community engagement or (especially) senior leadership. Register now!

And then our July webinar, How Public Media Companies are Building AI Policies, is now live to watch on our website.

It’s now been just over a year since public media was federally defunded (but not defeated). The defiance that materialized following that independence day was immediate, creative, and lucrative.

It also inspired reflection. At the time, the phrase “PBS is Punk” cut through a cluttered zeitgeist and had a moment. I asked in this newsletter, “Is PBS punk?” And I concluded, “Now, it must be.”

One year later, PBS is punk in the way Ben & Jerry are hippies. Well-intentioned, but compromised. Tony Moon’s inspired 1977 illustration in the ‘zine Sideburns (“This is a chord. This is another. This is a third. Now form a band”) has yet to be met with a public media equivalent.

Our values could (should) make us punk, but our system-level, status quo strategy undermines that. In the wake of CPB’s demise, the overwhelming financial support we received from our communities has largely been interpreted as a validation of that status quo. We are missing the window of opportunity to evolve in a way that maintains our values while building a 21st century public media.

Outside of traditional ‘towers-and-transmitters’ public media, the demise of CPB was a clarion call.

In the worlds of nonprofit media, journalism, academia, and the arts, public media stakeholders are not turning a blind eye nor deaf ear to that call. Multiple times this year I’ve sat in on focus groups, webinars, or conference breakout sessions where the topic of discussion was the future of public media. These opportunities for reflection, brainstorming, and organizing are being created, supported, and largely attended by external stakeholders. These are people who care very deeply about the existence of public media and what it means for the realities of their specific community. They are pointing their DIY (punk) energy at exploring what could (should) come next. These discussions have everything to do with architecting the future of public media, but traditional public media is poorly represented in these rooms, when it is represented at all.

The optimistic, DIY energy of punk infuses these gatherings. The discussions range from organizing better political representation, to building better infrastructure, to creating content that better reflects communities, to rethinking the entire purpose and direction of public media in a 21st century media landscape. Whether it’s indie journalists (LION, INN), indie filmmakers (Future Film Coalition), indie game studios (Field Day Lab), or community-focused media (NFCB), those punks respect our legacy, but that isn’t stopping them from imagining a future without us.

The definition of public media is expanding beyond “PBS” and “NPR.” Too often, our institutions have treated these organizations as occasional partners rather than as fellow architects of public media’s future, and we could do worse than work with non-traditional stakeholders to expand what public media can be in the 21st century. That means not only collaborating more broadly with nonprofit organizations whose values align with ours, but also embracing our ability to be both a public media center and a community hub. It also means exploring new forms of public media (like games) to turn data into information and information into experience.

A new generation of leadership is ascending within public media. My hope is that those leaders can reach past the status quo and form tighter alliances with the punks, leverage the privilege and momentum of our legacy, put our reach and resources at their disposal, and align our institutional mission with that punk spirit…before the second act starts without us.

Okay, on to the links.

I Don't Think Enough Millennials Talk About How Good PBS Was Back In The '90s (Rich Knight - Cinema Blend) - Following the anniversary of defunding, there is no shortage of mixed emotions. So, watching writers wax nostalgic about PBS from 30 years ago is a nice dose of fandom as we rethink what should come next.

Collaborations and primary elections boosted traffic to public media sites this quarter (Neel Dhanesha - NiemanLab) - It strikes me that clearly answering questions about "Why does this matter?" and "Why should I care?" could also be good tactics for appealing to AI bots scraping for answer engine content.

How public media is adapting after federal funding cuts (Alex Curley - Editor & Publisher) - After being aware of him for more than a year, I finally got to hang out with Alex at the PMGC conference last month (bright guy, great taste in TV). I like his notion of viewing July 18, 2025, at public media's Independence Day, but as someone who is keenly focused on the data I very much appreciate his caution that our data is dated.

Google’s AI search is rapidly becoming the default, new data shows (Sarah Perez -TechCrunch) - What I think gets skipped in all the handwringing about the shift from '10 blue links' to AI Mode is the fact that Google appears to be offering a product that people like to use. I admit Google is making the ability to use AI as frictionless as they can. But the fact that people are still using it indicates user preference on some level. Media publishers need to consider that as much as looking at these shifts through the frame of what Google is doing to us.

How YouTube is navigating an AI tightrope (Reed Albergotti - Semafor) - It's not surprising that Google's attempts to subject YouTube consumers to a similar product model as Search ultimately flopped. AI search is about information disentangled from sources (i.e., personalities), but YouTube is predominantly about the personalities. Search is trying to serve a more literacy-oriented society in a oral way. YouTube is decided oral in its orientation, and its consumers didn't respond well to a literacy-infused filter.

A new study looks at the skills journalists are losing (and gaining) because of AI tools (Andrew Deck - NiemanLab) - Increasingly, I'm of the mind that, yes, AI tools are going to help lazy people be lazy (see David Brooks’ piece in the next section). But laziness isn't contagious, and AI is not an infection vector. A lot of times, the objections that people front in relation to AI just feel like an unwillingness to change.

New research reveals AI answers only as trusted as the newsbrands they cite (Richard Reeves - PressGazette) - While this makes perfect sense, it also leaves me feeling optimistic. The fear regarding AI from 2023 forward has been that it will dupe us all into believing a false reality. Broadly speaking, people have shown they can consume AI- and AI-infused media while still accounting for the AI context of what they are consuming. That doesn’t mean we should let our guard down, but it does reinforce that we need to continue to give our audiences credit.

From Prompt Engineering to Intent Engineering (Daniel Messier) - There's a school of thought that prompt engineering was only really necessarily in 2023, yet that's how many of us learned to use AI "properly" so that's how we still do it. But AI has evolved so much since then that we're not really working with the same 'intelligence' under the hood, so it stands to reason that how we interact with these models should change. You talk differently to a toddler than you do to a teenager, than you do to an adult. The same principle applies.

Why Amazon hates 'human-in-the-loop' AI governance (Jessica Lyons - The Register) - The basic argument here is that people are messy and things run better when machines (which are less messy) make the calls. It's a contrarian-flavored argument in favor of agentic AI. But it also assumes the most elegant solution to any problem is a (near) frictionless agentic solution. Sure, there are situations where humans are necessary, but they are seen here as few and far between. "Human accountability" — meaning oversight versus being "in the loop" — seems to be the preferred approach for some in big tech. So, I guess humans still matter...when shareholders need someone to blame.

AI companies are reportedly shredding millions of books after using them to train AI models — tech giants outsource to middlemen to secretly buy up books for training material (Zhiye Liu - Tom's Hardware) - While I'm guessing their workflow is really just indifferent efficiency at play, it's also hard not to see this as deeply troubling. It's the kind of tidbit of information that gets dropped in a paragraph, voiceover, or dialog of exposition in a dystopian novel, movie or series.

How AI Will Make Art Worse (Tim Brinkhof - Time) - The headline is a little clickbait-y. My read of the essay is that art will have to change as a result of AI. To me that doesn't necessarily make it worse. Art constantly changes. But I can see where the period of adaptation will be rougher on artists across all media. Few expected the rate of change to be so extreme, but that has become the norm. So maybe the real headline is that AI will make life for artists worse.

Have Data Centers Raised Your Electric Bill? Causal Evidence from the United States (Asa Watten, John Bistline, Geoffrey Blanford - arXiv) - Two points to note here. First, the data studied runs from 2015 to 2024, arguably before the current AI-driven data center backlash. Second, data centers do not equal AI. They are conflated now, but 5 years ago data centers meant "crypto." That's not to say that we should look past the fact that AI has an intensive hardware component, but we need to be honest that much of our digital life for the past 10 years has relied on an increasing data center footprint across the globe.

Rising Emissions, Depleting Water and Vanishing Land—UN Scientists: AI Is Threatening Natural Resources for Billions (United Nations University Institute for Water, Environment and Health) - I'm linking to the press release above, but you can also jump to a PDF of the full report. Here's one sample set of stats: "The energy required to generate a typical AI image is enough to power a 10-watt LED bulb for 17 minutes, and the energy required for a high-complexity AI video is sufficient to run that same bulb for 42 hours. Similarly, the electricity-associated water footprint is about two tablespoons (29 mL) for a single image, but jumps to 4.1 liters for a complex video—almost equivalent to a two-day drinking water need for one person." My takeaway is not that we should stop using these tools but that we need to be aware of the impacts and self-meter our usage (just like we self-meter our use of paper or gasoline).

xAI is challenging a new Minnesota law banning 'nudify' apps (Mariella Moon - Engadget) - Good for Minnesota. Let this me a reminder that one of the second-tier models still thinks this particular form of violence against women is a feature. Also, seriously, is anyone in public media still posting regularly on X? If so, ask yourself why you're supporting the platform.

Google pauses AI satellite images, after fears of deepfakes in the sky (Geoff Brumfiel - NPR) -
This one gave me pause. It was pitched as a potential education tool, however I (and apparently a lot of other people) worry that the images generated will be taken as fact, instead of algorithmic conjecture. We know that there is a huge amount of bias in image generators based on the highly skewed dataset of historical imagery. And that doesn't even touch on those who'd like to generate images to re-write history. I think I'd maybe feel better about it if it were a distinct tool drawing on the same imagery as Google Earth, not one baked into Google Earth. But maybe that is a distinction without a difference.

Some people's chats with Claude AI found to be publicly available online (Kali Hays - BBC) - This feels like that time when Millennials learned that Snapchat's disappearing photos didn't actually disappear. The old adage of never typing anything you don't want your mom to read in the New York Times still applied.

Netflix used AI in nearly 300 titles this year (Aytun Çelebi - Dataconomy) - Not surprising, but an interesting data point, nonetheless. Netflix doesn't have any reason not to lean into gAI tool (especially in post-production). The art of visual storytelling isn't their business plan, it's regular, massive drops of content. This requires scale, and AI helps achieve that scale.

Will betting on wildfires lead to arson? (Kylie Mohr - High Country News) - Insider trading is fairly rampant on prediction market sites, so I'd say the answer to the headline question is, eventually, yes. Setting aside the overwhelming grossness of that eventuality, there are a couple of serious considerations for media and, especially, public media. Are you willing to utilize prediction markets data in your journalism? And do you have guardrails in place for editorial staff who might want to place personal bets on prediction market sites?

LinkedIn's new 'Seems like AI slop' button lets you report all those cringey posts (Jada Jones - ZDNet) - I'm very interested to see how this plays out (i.e., how long it lasts). I'm glad they aren't dinging all AI-generated art (as I use it regularly in my newsletter). But I've also noticed that if you upload AI images you can duck the Content Credentials flag if you do a couple of onsite edits using LinkedIn's image editing tools. Since "slop" is a problem of human perception over algorithmic screening, this approach probably makes the most sense.

What the World Should Learn from Australia’s Social Media Law (Ravi Iyer, Jonathan Haidt, & Zach Rausch - Time) - I've posted a couple of stories recently that were reacting to Australia's ban. It seems like the early data is that bans like these won't stamp out teenage social media usage, but they'll create a significant speed bump for teens (some of whom are looking for any offramp anyway). Into the time void left by the lack of social media, they are starting to fill that with reading, sports and offline socializing.

TikTok is no longer banned on US government devices (Mariella Moon - Engadget) - For those stations tied to state networks that have followed the federal governments lead on TikTok, a thaw might be in the forecast for you.

US Government Displays AI-Mangled Map of Africa That Gets Every Single Country Wrong (Frank Landymore - Futurism) - Since I share a name with a country in Africa, this one hits a little close to home. I'm guessing China doesn't pull **** like this.