Guardian essay calls AI 'the pleasureless decadence machine'

A Guardian opinion essay this week did something that does not happen very often on mainstream culture pages. It named AI as the emblem of an era's malaise, and it named it precisely. Michèle Mendelssohn and Charlie Tyson argue that we live in an 'era of decadence without pleasure,' and that 'AI is the pleasureless decadence machine.'

The claim is not the usual one that AI is dangerous because it will take jobs or spread lies. It is that AI is the purest expression of a broader condition, abundance without enjoyment. They quote the philosopher Carissa Véliz for the reason: 'There is no pleasure or pain for an AI, no skin in the game.' An AI, they write, 'cannot care, and therefore cannot value.' The consequence, in their reading, is a stylistic flattening: 'the algorithm steers language away from the idiosyncratic and toward the predictable.'

Where this gets interesting for anyone building or selling AI products is the company the essay puts AI in. It sits next to Bryan Johnson's longevity regimen, GLP-1 weight-loss drugs (which the authors say 'reshape the felt relation between desire and satisfaction,' and which one in eight American adults are reported to take), and what they call 'looksmaxxing' culture. The unifying thread is optimisation as a substitute for enjoyment. Their line, 'against the pressure to optimise life, decadence issues a different command: feel it,' is designed to travel, and probably will.

Take the specifics as framed, not settled. This is an essay, not reporting; the 'one in eight' figure is the authors' framing rather than a fresh empirical release, and they do not sketch what a 'New Decadence' art movement or product would actually look like. The piece also lumps generative models, recommender algorithms, and pharmaceuticals into a single cultural mood, which is elegant prose and blunt analysis.

The forward-looking part is the one AI product teams should watch. If 'flattening' and 'predictable' become the sticky cultural framing for model output, the competitive edge probably shifts to whoever can credibly promise idiosyncrasy, human authorship, or embodied experience, from indie publishers to studios that sell 'human-made' as a premium tag. The critique will not slow model releases. It might slow how readily people are willing to let those releases speak for them.