The robot takeover will start in the smaller cities. Towns and small cities have a smaller proportion of jobs that will be resilient to automation than larger urban centres, according to a new study.

By looking at the jobs that are most susceptible to automation and their distribution across different US cities, Iyad Rahwan at the Massachusetts Institute of Technology Media Lab and his team have found a trend between the size of a city and the impact we should expect artificial intelligence and robots to have on human workers. Roughly speaking, cities with fewer than 100,000 inhabitants are more at risk.

The East Coast cities are full of jobs that should be resilient to automation. Washington DC, for example, has many government-related roles that are hard to automate, and New York, with its population of 8.5 million, is able to support many specialist jobs too.

On the other hand, in Madera County in California – a wine-growing area with a population of just 60,000 –  many of the agricultural jobs can be done by machines. Nearby San Francisco with a population of 850,000 will be resilient due to its size and thanks to being a centre of innovation.