“A Subset of the Natural” explores our complicated relationship with human-made elements in the natural world.

Listen: on our site, PRX, Mixcloud, AudioPort (login required) — complete track list

This week on the Earth Eclectic Radio Hour, Bart Everson hosts a provocative tour through sounds that examine humanity’s synthetic footprint on Earth. Beginning with the Stop-Shopping Choir’s anti-war meditation on machines, the program flows through Tired Circuit’s sonification of oil spill data—revealing the shocking frequency of pipeline failures—before confronting our plastic-saturated oceans through songs from around the world. North Hive’s “Artificial Nature” introduces the concept of simulation through granular synthesis, while H.Flora’s mysterious composition blurs the boundary between actual frog recordings and electronic biomimicry. After Pollenate maps bird migration through human instruments, we confront an increasingly urgent question: in an age of generative AI, how do we distinguish authentic ecological expression from its algorithmic imitation? With examples of both unlabeled and transparently artificial music, Earth Eclectic faces both philosophical and practical challenges posed by emerging technologies. In a world where microplastics infiltrate our bodies and artificial intelligence infiltrates our soundscapes, this collection invites us to consider what remains “natural” when the artificial has become ubiquitous—and whether that distinction still matters for our ecological consciousness.

PSA: The ecological impact of artificial intelligence is concerning, particularly generative AI, but how does it stack up to existing habits many of us don’t think about twice? We’d like to draw your attention to a non-partisan calculator called “What Uses More.” It’s an attempt to provide a more objective comparison of the energy and water usage of various AI tasks versus other common digital activities like streaming movies or using a search engine. These calculations are based on academic papers, industry reports and expert estimates. Many thanks to John Ippolito of the University of Maine for helping to develop this resource. You can find it at What-Uses-More.com.