El Podcast

The Eavesdropper Economy: How Surveillance Built AI (E186)

Episode Summary

From a battery-less Soviet bug hidden in a wooden plaque to Stasi-era data overload, this episode traces how “listening in” scaled from espionage into an automation problem that helped propel modern AI. We unpack how surveillance capitalism and always-on devices—especially around kids—reshape privacy, language, and behavior, even when no one seems to be listening.

Episode Notes

A lively tour from Cold War “The Thing” to today’s surveillance capitalism—showing how audio capture, too much data, and automation pressures helped turn listening into AI.

Guest bios:

• Dr. Toby Heys — Professor at the School of Digital Arts (SODA), Manchester Metropolitan University; co-founder of the AUDINT sonic research unit; co-author of Listening In
• Dr. David Jackson — Senior Lecturer in Digital Visualisation at SODA, Manchester Metropolitan University; researches AI’s cultural impact; founded the Storytellers + Machines conference (2023); co-author of Listening In.
• Marsha Courneya — Canadian writer/editor; teaches Digital Dramaturgy at the International Film School of Cologne; doctoral researcher in Digital Culture and Communication at Birkbeck, University of London; co-author of Listening In.

Topics discussed:

• “The Thing” (1945): passive bugging, resonance, why it went undetected
• Cold War escalation: normalization of listening, Five Eyes, PRISM/Snowden
• Stasi data glut: informants, dossiers, “collecting as mania,” behavior change
• Language under surveillance: cryptolects, slang, coded speech, hip-hop as evasion
• Surveillance capitalism: smart homes, smart toys, wearables, “data promiscuity”
• Kids + data: baby monitors/crib cams, school biometrics, “data twins” before birth
• AI training + intimate life: accidental recordings, human review, terms-of-service reality
• Future tensions: convenience vs autonomy, regulation lag, ownership erosion (“enshittification”)

Main points:

• Audio surveillance scales into an “automation problem.” Once you can record everything, the bottleneck becomes listening fast enough, pushing intelligence services toward automated analysis.
• Surveillance changes behavior—even when nobody is actively listening. The possibility of being overheard bends speech, jokes, and self-presentation (Stasi dynamics → modern smart devices).
• “Too much data” doesn’t make it harmless. The danger isn’t only what’s heard today, but the creation of a searchable “permanent record” that can be reinterpreted later.
• The home becomes the most valuable capture zone. People drop the public mask at home; that intimacy makes in-home audio uniquely revealing and therefore lucrative/powerful.
• Children are captured early—often via “safety” and parental anxiety. Baby tech, smart toys, school systems, and medical records create a data trail before kids can consent or understand it.
• Snowden shocked—but didn’t trigger lasting mass refusal. The episode argues leaks often lead to resignation/memeification (“the intel officer listening”) rather than sustained backlash.
• AI + ownership is the next front. Beyond privacy, the guests worry about erosion of ownership (you can’t fully “own” digital goods or refuse totalizing platforms as easily).

Top 3 quotes:

• Toby: “There was nothing to detect.”
• Marsha: “It ruptures language completely.”
• David: “data isn’t secure and safe.”