Doom Composer Mick Gordon Rejects Generative AI: It Only Averages the Past
Speaking at PAX Australia, composer Mick Gordon argued that generative audio models cannot innovate because art requires breaking rules—like turning chainsaws into guitar riffs.
In an interview with IGN at PAX Australia 2026, composer Mick Gordon offered an unvarnished verdict on generative machine learning in creative workflows: "Fuck AI."
The musician behind the defining soundtracks for Doom and Doom Eternal argued that generative models are structurally unsuited for genuine artistic invention. In Gordon's view, these systems merely ingest existing work to output an algorithmic average of past human output. Breakthroughs, he countered, happen specifically when an artist refuses the standard formula.
To illustrate the gap, Gordon pointed to his own production methods: running raw audio recordings of a chainsaw through signal processors to generate guitar riffs for Doom, or sampling a Tibetan ritual trumpet for the fighting game Killer Instinct. Predictive systems operate within existing distributions; they do not possess the creative logic to substitute power tools for instruments.
Gordon expressed optimism that listeners will ultimately gravitate toward authentic human craft, betting that automated saturation will push people toward live experiences. He is currently working on soundtracks for Warhammer 40,000: Boltgun 2 and the shooter Defect, and plans to perform his first live set of Doom tracks at the RADAR festival in 2027.
Comments
Help shape the next version: Add context or suggest a correction. AI review can add points toward a rewrite. Reviews and updates may take time; a full meter does not guarantee a new version.