AI needs its slow-food moment

The recent wave of research on artificial intelligence and human cognition reads like an ongoing post-mortem of our attention spans. Papers track students handing drafting tasks to language models, monitoring cognitive load, memory retention and neural activity. The underlying worry is simple enough: if software does the heavy lifting, our own brains do less work, retain less context and atrophy.

Whether those claims hold up over decades remains an open question. Many of the studies are small, task-specific or reliant on self-reported habits from university cohorts. Some find drops in recall during short writing exercises. Others point out that structured prompting can force students to interrogate their own logic more closely. We do not yet have settled neuroscience proving that machine assistance permanently shrinks our capacity to think.

What we do have is an aesthetic and intellectual exhaustion that feels familiar.

To understand where digital culture is heading, look at how we handled industrial food. The mechanisation of agriculture and processing solved an urgent, historical problem: it fed millions cheaply and reliably. Caloric scarcity receded across much of the Western world. Shelf-stable packaging and high-yield monoculture replaced regional cultivation.

That efficiency carried an obvious cost. Flavour flattened into sweetness and salt. Varieties of produce disappeared because they bruised too easily in freight crates. Soil quality degraded under repetitive cycles of extraction. The supermarket shelves filled to bursting, yet eating became a detached, solitary act of calorie intake rather than an engagement with what was grown.

Cheap abundance always degrades taste before it restores it.

When fast food expanded across Europe in the 1980s, the reaction was not an immediate Luddite withdrawal into the woods. The reaction was the Slow Food movement, founded in Italy as an explicit protest against homogenised, industrialised dining. Its founders did not argue that people should starve for the sake of purity. They argued for provenance, careful preparation, local ingredients and the time required to sit down and share a meal properly. They wanted to protect the specific character of food from being engineered into uniform mush.

Industrial efficiency solves access, but it eventually destroys distinctiveness.

Generative software has reached its factory-farming phase. Writing, illustration, audio and code are produced at zero marginal cost and infinite volume. The feeds are stuffed with syntactic sludge: grammatically correct, frictionless paragraphs that say nothing distinct and cost nothing to produce. It is intellectual shelf-stable loaf. It keeps the screen occupied, but nobody finishes reading it feeling nourished.

The consequence will not be the eradication of human craft. It will be the re-evaluation of it.

When quantity becomes instantaneous, care turns into the only meaningful signal. The labour required to draft an essay by hand, to balance an argument through revision, or to photograph something with physical light becomes visible precisely because the alternative is so cheap. We do not value a handmade ceramic bowl because it holds water better than a mass-moulded plastic tub. We value it because an individual made decisions about its weight, surface and imperfections.

Abundance does not extinguish human touch; it turns human touch into the only scarcity that matters.

Does the human touch ever really disappear from culture? History suggests it retreats briefly while the novelty of a machine plays out, then returns as a conscious preference. Mass production never killed the baker; it created the bakery queue for sourdough with a charred crust and an identifiable grain.

The digital ecosystem is overdue for that exact reckoning. The novelty of pushing a button to generate five hundred words on any subject has already worn thin. Readers are learning to spot the synthetic cadence within two sentences. The reaction against it will not require abandoning technology entirely. Instead, it will create a distinct demand for slow content: essays that took a week to consider, arguments that reflect personal risk, and ideas that were not averaged out by a predictive model.

We spent decades learning to care about the soil our vegetables grew in and the hands that harvested them. The next step is simply applying that same discernment to the words and thoughts we feed our minds.