Aug 12, 2026 · 12 min
AI stories fool readers as black holes wander and pumas reshape roads
Can you spot AI-written stories? Listen and find out
The episode connects uncertainty about machine-made writing with discoveries showing how AI can aid astronomy and animal behavior can improve road safety.
- 1Readers often miss AI-generated stories and may prefer them when they believe humans wrote them.
- 2An AI algorithm helped identify a wandering supermassive black hole by recognizing light from a disrupted star.
- 3Pumas can reduce deer-vehicle collisions by changing deer behavior, raising questions about their natural return.
Don't miss
Mary Louise Kelly identifies the human-written koi-pond passage, while the surrounding research shows how easily readers can misjudge AI stories.
The brief
Short Wave hosts Regina Barber and Burleigh McCoy bring in Mary Louise Kelly for a science roundup that begins with a deceptively simple test: can she tell human writing from AI writing?
Kelly correctly identifies the human-written koi-pond passage, but the broader research is less reassuring: readers often cannot distinguish AI stories and may prefer them when told a person wrote them.
The episode then turns to astronomy, where an AI algorithm recognized the light signature left when a star was torn apart, revealing a supermassive black hole outside a galaxy’s center.
In the final story, researchers link active pumas with substantially fewer deer-vehicle collisions, because the predators change where and how deer move.
Together, the stories examine how expectations shape judgment, how algorithms extend scientific detection, and how predators can alter human safety without touching a road.
Featuring
Mary Louise KellyMentioned
Villanova UniversityListen to the full episode and explore every guest, topic, and moment on PodLume.