Data scientist and Leiden University assistant professor Lauren Leek has sparked widespread discussion online with her research into how predictive algorithms are flattening everyday consumer culture. In a recent Substack article titled “Temperature Zero for Culture,” Leek analyzed how digital mapping, data analysis tools, and large language models are transforming everything from pop music lyrics to high streets into a homogeneous monoculture.
During a recent walking tour on a shopping street in southwest London, Leek demonstrated her findings in real time. Examining a traditional British pub owned by a massive international hospitality conglomerate, she pointed out how corporate data aggregation optimizes menus—removing local quirks in favor of standardized staples like burgers and fish and chips that can be ordered via QR code. Nearby, at an upmarket café and bakery majority-owned by an American private equity firm, she explained how specialized site-selection algorithms use public census data and proprietary insights to dominate prime real estate.
The core of this cultural homogenization, according to Leek, lies in the mathematics of next-token prediction used by artificial intelligence models. When an algorithm operates at "temperature zero"—always choosing the safest, most statistically probable outcome—it eliminates randomness and surprise. While tech commentators like Cory Doctorow describe this degradation of digital and physical spaces as "enshittification," Leek prefers the term "standardization," emphasizing that while individual items may not necessarily be lower in quality, the overall variety of consumer culture is steadily disappearing.
