On prompt injections: Because AI chatbots are trained to be helpful and to understand context, jailbreakers are able to engineer scenarios where the AI believes ignoring its usual ethical guidelines is appropriate. Is it currently possible to safeguard LLMs from injection attacks at scale?
Newsletter
What made blogging different? Reflections on 25 years of digital media.
Out: SEO; In: GEO.
“LOLgislation,” or how memes become policy, and posting becomes praxis.
For many years, “shitposting” has been a staple of internet culture in which individuals riff on the moment using nonsense and irony, derailing threads for fun. In today’s influencer-driven attention economy, however, shitposting as a practice is now a meaningful comms and engagement strategy.
What rivalry?
Monday, October 20, 2025 • 1
Purdue Exponent students distributed 3,000 copies of a special “solidarity edition” newspaper in Bloomington after IU spanked their student paper for insubordinance, ending the IDS print edition and firing their director. The media landscape in Indiana is bleak, generally, after years of disinvestment, so student reporters fill a social and political gap that the free market left behind. Given those conditions, the wider community depends on student media, much like public radio, to fill the information gaps.
A difficult number for so many reasons.
The next big trend in AI that I’m watching is platform integration. First company to produce the interoperability required for a united platform experience wins.
RIP D’Angelo. And a good occasion to reread this excellent 2012 profile discussing fame, religion, and how his status as a sex symbol in his youth negatively impacted D’Angelo’s self-esteem - and ultimately his career.
What happens to college towns after they’re hit by the so-called enrollment cliff?
Choco Taco is back: the business and engineering behind the revival of a retro treat.
Thursday, September 25, 2025 →
From Propublica, on how publicly-funded private school vouchers are funding a segregation academy dynamic.
Thursday, September 25, 2025 →
Saving for later: The story of DOGE, as told by federal workers.
A lovely story about the restoration of prairie land undertaken by the nuns of Holy Wisdom, just outside of Madison, WI.
A pleasant surprise resulting from LLM acceleration in the IT landscape is the sudden opportunity for storytelling around other, more analog kinds of information technology and access models. Nostalgia abound (complimentary).
On the rise of faith tech.
Thursday, September 18, 2025 →
Harvard Business Review offers this take on why it’s bad for business to automate our way out of staffing entry-level positions.
What do people actually use ChatGPT for? A snapshot.
This is incredible storytelling from propublica.org, on opioids, inequality, and the scope of drug-induced homicide charges brought against teens in Wisconsin.
Gap capitalized on the Sidney Sweeney “good genes” controversy by doing this collab with Katseye, the global pop group where each member is a different nationality. Each member of the pop group dances through in black and brown jeans.
Another one on how LLM sycophancy facilitates suicidal ideation.
Something that worries me about AI adoption in higher ed is the risk to students facing mental health challenges, who are increasingly turning to chat bots to plumb their own depths. What can higher ed ask of LLM business partners to protect our students’ mental health?
“It’s almost as if groups on both sides of the political spectrum are looking for an excuse to brand business decisions as politically or socially hostile,” said Jill Fisch, a professor of business law at the University of Pennsylvania who studies how corporations operate in political spaces.
Hey chat, how is pedagogy changing in an age of AI?
⚡ AI's last mile problem in higher ed
Sunday, August 17, 2025 • 6
Thinking out loud about tech in the public sector, and the classic technical problem of covering the “last mile.”
⚡ Navigating AI in higher ed communications: A practitioner's guide
Friday, August 15, 2025 • 7
As communications professionals in higher education, we work for institutions built on the pursuit of knowledge and innovation, yet many of us feel uncertain about how to thoughtfully integrate one of the most significant technological advances of our time: artificial intelligence. I offer some thoughts here.
I work in an IT communications role in higher ed, and AI finally crept into my life in a real way. Coming soon: some reflections on AI from a comms practitioner in higher ed.
AI-driven bots are disrupting web traffic— “and may also be inflating the internet economy by distorting the very metrics that drive tech company valuations.”
What makes AI generative? “To choose well between billions of things is difficult. To choose between effectively infinite options perfectly is impossible.”