From The Guardian:


The automated city: do we still need humans to run public services?

From driverless buses to an AI council worker called Amelia, municipal services are becoming increasingly automated. But what does that mean for the future of our cities – and the jobs market?

Scientific advances and new technologies often enable dramatic improvements in public services and urban life, eradicating some jobs while creating new types of employment. But the next chapter of urban automation might be more profound than any previous one. In fact, it’s already begun.

“Smart cities” offer a seductive vision of a world where everything runs as smoothly as the latest iPhone. Need a parking space? An app will tell you where one’s available, and notify you (and your friendly neighbourhood parking inspector) when your time is up. It’s the kind of technology many cities are trialling, embedding sensors in streetlights, curbs and buildings to monitor parking, traffic and air pollution – even crime. …

Still, it’s easy to guess where old municipal jobs might vanish. In London, driverless buses that can run bumper-to-bumper would make the new Routemaster seem as antiquated as a Morris Minor – and potentially supplant London’s 22,500 bus drivers. …

In October 2014, Transport for London (TfL) also unveiled 250 “driverless” tube trains which should come into service from 2022.  … Initially the new trains will still have drivers, though in the longer-term they may be redeployed as “train captains”, performing a similar role to attendants on DLR trains.



A ‘world first’: driverless minibus in Lyon

Jarmo Eskelinen, chief innovation and technology office at Future Cities Catapult, argues future transport networks will react faster to emergency situations. “At the moment they adapt based on the reaction speed of people, and their reaction speed is often very slow. In future this will be automated so that they respond in near real-time, stopping transport flows to the emergency zone.”

Could cities even become self-repairing? Leeds University is leading a £4.2m project to create a fleet of robot repair workers that can spot infrastructure problems before they become disruptive – including drones that perch on lampposts to change bulbs, automated machines that fix potholes without digging up half the road, and robots that live in utility pipes and patch cracks. Professor Phil Purnell, who leads the school of civil engineering research team at Leeds University, likens these machines to a city’s “white blood cells”, repairing damage before it requires a major intervention.

He says putting road workers out of work is absolutely not the objective of the exercise. “What we need is the people who are doing tasks that are fairly dull and don’t need much skill freed up to attack the real infrastructure problems, of which there are hundreds upon hundreds that we’re burying our heads in the sand about.”

If dull tasks are the target of automation, then many back-office council roles are susceptible. Spurred on by the austerity mantra of “doing more with less”, councils are beginning to apply robotic process automation, which mimics human interaction with computer systems, to repetitive tasks such as signing people up for council tax direct debit payments. In theory this should free up staff for more demanding strategic work. Machines do the boring data entry tasks, which they generally perform faster and more accurately than human beings. And human beings use their time, empathy and creativity to improve frontline services. …

Robinson says that, while artificially intelligent chatbots could have a role to play in some areas of public service delivery: “I think we overlook the value of a quality personal relationship between two people at our peril, because it’s based on life experience, which is something that technology will never have – certainly not current generations of technology, and not for many decades to come.”

But whether everyone can be “upskilled” to carry out more fulfilling work, and how many staff will actually be needed as robots take on more routine tasks, remains to be seen. Carl Benedikt Frey and Michael Osborne’s influential 2013 paper The Future of Employment: How Susceptible are Jobs to Computerisation?, estimates that 47% of US jobs are “at risk” of being automated in the next 20 years. Another report by Deloitte found that in London, 29% of admin and support services jobs, and a whopping 72% of transport and storage roles, are at “high risk” of automation.

However, a report Forrester published last year was less pessimistic about people’s future employment prospects, suggesting that only 9.1 million US jobs will be automated by 2025. Robinson is more inclined to believe Forrester’s estimate. “It’s inarguable that as technology develops, it will automate certain tasks. But ‘tasks’ are very different to ‘jobs’. I also think some reports are hugely optimistic about what technology will be able to accomplish in [the] future.”

If Google or another tech giant does eventually manage to create an artificial general intelligence that can successfully perform any task a human can, the job losses would dwarf anything we’ve seen before – and not only among the 1.5 million people employed by local government in England.

A universal basic income, which would provide everyone with enough money to maintain a decent standard of living, is often cited as a solution to this problem. But in the medium term we might find robots still need our help; that there are things we simply do better than machines.

For example, humans working cooperatively with machines are generally regarded as the strongest chess-playing entities, each drawing upon their skills to beat opponents. Identifying what those skills are in public service terms, and how best to combine them with automated systems to improve urban life, is a challenge no city can afford to ignore. Otherwise more people might find, as London’s lamp lighters once did, that their services are no longer required.