We used AI to analyse three cities. It’s true: we now walk more quickly and socialise less | Carlo Ratti

Something is thinning in public space. Pavements are still crowded. Parks still bustle. But if you look more closely – or, better still, if you measure it – the texture of our interactions has changed.

Together with colleagues at Yale, Harvard and other universities, we used AI to compare footage of public spaces from the 1970s with recent video in the same locations in New York, Boston, and Philadelphia. The findings are striking: people walk faster, linger less, and are less likely to meet up. That’s no surprise in a world where phones, Netflix and AI companions are luring us away from real-world spaces and real-world friends. Yet, if technology is part of the problem, it may also be part of the solution. By using AI to study urban public spaces, we can gather data, pick out patterns and test new designs that could help us rethink, for our time, our modern versions of the agora– the market and main public gathering place of Athens.

The urban playground has always drawn curious minds. Among the sharpest was William “Holly” Whyte, who filmed plazas and parks in 1970s New York. He was fascinated by where people chose to sit, how they navigated space, and what drew them together. His findings, documented in The Social Life of Small Urban Spaces (1980), were at times beautifully simple: “What attracts people most, it would appear, is other people.” From his footage, Whyte turned his observations into data-backed recommendations: he said seats should be “two human backsides deep” and sung the praises of movable chairs that let people chase sun or shade. His analysis helped save New York spaces such as Bryant Park and shaped our modern approach to people-centred design.

Whyte’s experiments were revelatory, but hard to replicate. Why? Analysing the footage, frame by frame, took a team of assistants months. Now, finally, that challenge has been overcome, as we have invented non-human evaluators. Our team digitised Whyte’s original footage and compared it with recent videos – of Bryant Park, the steps of the Met Museum in New York, Boston’s Downtown Crossing, and Philadelphia’s Chestnut Street – collected by sociologist Keith Hampton. Then we trained an AI model to analyse the two sets of footage. AI allows self-driving cars to recognise bikes and pedestrians; that same technology excels at analysing footage of parks and plazas, and keeping track of hundreds of people at the same time. What took Whyte months now takes minutes.

So how have cities changed between 1970 and 2010? As we discuss in a recent paper in the proceedings of the National Academy of Sciences, walking speeds have increased by 15%. People stand still less often. Dyads – pairs meeting and then walking together – have declined. Downtown Crossing in Boston, once lively and social, has become a pass-through. Even in Manhattan’s Bryant Park – improved, according to Whyte’s vision…