#046 - 2026/09/23
A selection of what I've read this past week.

My main newsletter, Complex Machinery, includes a section called "In Other News..." It's where I list one-liners about interesting articles that didn't fit into any segments.
You can think of this list as a version of In Other News, but with a wider remit than Complex Machinery's "risk, AI, and related topics."
Above the fold
- This piece is about how UK fish and chip shops adapt and survive in a changing world. The stories mirror the experience of other fields grappling with genAI and other upheavals. (The Guardian)
- Insurers are reluctant to provide coverage of genAI-related incidents. That's probably a sign. (The Register)
- Contractors hired to prepare training data for genAI systems are poisoning the proverbial well – some on accident, some on purpose. (404 Media)
- Last week we passed two anniversaries from the 2008 mortgage crisis: the Lehman Brothers collapse and the AIG bailout. This piece offers a glimpse into the fall of the insurance giant. (New York Times – from 2008)
- That genAI-based transcription may lead to legal troubles down the road. (Sidley Austin LLP)
- An analysis of the effect of drone strikes on noncombatant populations – notably, what this means for geographic dispersion of civilians. (British Journal of Political Science)
- Former DraftKings data scientists call out the way the company identifies potential problem gamblers and then tempts them with more gambling. (New York Times)
- The CEO of Shopify is unhappy with the results of his company-wide genAI adoption mandate. (Fortune)
- Dolphins are notoriously smart animals. Smarter than most genAI bots, I might argue. (The Guardian – from 2003)
- An intelligence analyst allegedly accepted a chatbot's fabrication. The US almost triggered a war with China as a result. (CNN, Gizmodo, Le Monde 🇫🇷)
- New York's Flatiron building gets a makeover into luxury condos. (Bloomberg)
The rest of the best
- Apparently, people inside Microsoft and OpenAI knew that genAI could cause harm to the news industry. And not just by hoovering up articles as training data. (New York Times, Der Spiegel 🇩🇪)
- CEO of controversial genAI startup Character AI lands a new job running tech for Disney. (Gizmodo)
- The FAA is implementing AI-based systems to address air traffic control concerns. (I write "AI" there and not "genAI" since it's not clear what flavor of technology they're using.) (Ars Technica)
- genAI hype and hand-waving are running headlong into reality. While this article is about the risks of using this technology in medical scenarios, the lessons therein equally apply to other fields. (FT)
- The term "AI safety" has a hazy definition, but thanks to recent incidents, we can all agree that the various flavors of this field have become more important. (The Verge)
- Apple's claims to end-user data privacy build a level of trust that other large tech companies – notably, genAI providers – do not get. (FT)
- Apparently, those Flock cameras have another problem: they're running outdated software, which makes them vulnerable to hacking. (Michah Lee)
- Semiconductors are required ingredients of many electronic devices. Given that, what would be the impact of a disruption in that supply chain? (Taylor & Francis Online)
- While running in a simulated environment, genAI agents exhibited dishonest behaviors. (Bloomberg)
- Warren Buffett once pulled Salomon Brothers from the brink of collapse. One arrow in his quiver: being flat-out honest. (FT)
Did I miss anything?
Have something I should read? Send the link my way.
Don't miss what's next. Subscribe to In Other News...: