Jackson, D., Courneya, M., Heys, T., 2024.

Listening In: How Audio Surveillance became Artificial Intelligence

The logic goes that you can close your eyes to stop seeing but listening is not as easily avoided. But what happens when the listening subjects are not just individuals with ears and societies with recording devices, but also systems that can perceive more abstract utterances such as those made by the inner voice or a voice reconstructed beyond death? And what happens when these listening objects are not just deployed by governments to surveil foreign bodies, but are invited into our homes in the form of AI companions? This book proposes that the urge to listen to everything that has ever been uttered is scored deeply into the technological operating system of Western culture; a rationale that results in audio surveillance being normalised from pre-birth (via ultrasound) onwards. From Thomas Edison's yearning to record "important speeches of men and gods, down through the ages", to Charles Babbage's conviction that "[t]he air itself is one vast library, on whose pages are forever written all that man has ever said or woman whispered", there has been a desire to capture, store and re-distribute massive audio datasets that somehow represent the entire story of human experience.

Spanning from 1945 and speculating forward, the book traces an arc from the Great Seal Bug invented by Leon Theremin hidden in a wooden rendering of the USA's national coat of arms through to the present day in which Pegasus spyware can record our calls, copy messages and secretly film us and into the near future where AIs can listen to things we have not yet said. The narrative arc of the book amplifies the ways in which singular models of audio intelligence in the 1940s were expanded into national systems of surveillance by the Stasi, NSA, and Echelon. It then reveals how mass forms of audio intelligence now inform artificial intelligence through speakers and appliances brought into the home such as mobile phones, children's toys and IoT baby monitors. The final section shifts focus again as listening even transgresses the internal/external divide of the body through AI powered technologies designed to capture the inner voice. The book offers multiple learning paths for the reader by discussing chapter topics through thematic lenses including voice, fidelity, process, and intimacy.