UK Edition
Will a robot take your job? These are the roles at greatest risk from the robo-revolution
An algorithm can calculate which jobs are most likely to be swallowed by computation and which will put up the most resistance
The robot uprising may not yet quite be at the levels of The Matrix or The Terminator, but some professions are increasingly being engulfed by automation.
According to a new study, barbers, bartenders, factory workers and models are among the vocations most vulnerable to being lost to the upcoming robo-revolution.
In contrast, members of the clergy, doctors, scientists and chief executives are the most well-placed to keep their livelihoods.
In years to come, it is conceivable that some jobs will be lost completely, replaced by cheaper, more efficient and more reliable robots.
However, scientists at the Swiss Federal Institute of Technology developed a computer algorithm to calculate precisely which jobs are most likely to be swallowed by computation, and which jobs will put up the most resistance.
The algorithm is called the automation risk index (ARI) and takes into account how many aspects of a job can be done by a robot, ranks their importance and accounts for how close machines are to being able to do these things.
A number is then generated between 0 and 1, and the closer to 0 it is, the more robot resistant it is.
“ARI does not correspond to automation probabilities but provides a measure of the relative automation level of a job with respect to all other jobs,” the researchers explained in their paper, published in the journal Science Robotics.
“ARI = 1 means that machines outperform humans in all the human abilities required by the job, and ARI = 0 means that robotic technologies cannot replace even a single human ability required by the job.”
The mark takes into account if a physical robot can replace the role, such as in car manufacturing plants, or if an artificial intelligence programme will make humans redundant.
Almost 1,000 jobs were put through the analysis, and the system deemed physicists, with a score of 0.43, to be the profession least likely to be seized by machines.
Neurologists, mathematicians, surgeons and epidemiologists all made the top 10, while astronomers, judges, immunologists and chemical engineers featured in the top 20.
At the other end of the scale, meat packers came in last place with a score of 0.78.
Car wash staff, cafeteria attendants, orderlies, janitors and maids were in the bottom 10, while fast-food cooks, pot washers, solderers and launderette staff came in the bottom 20.
Out of the list of 967, models languished in a lowly 927th place, with the computer system confident robot androids can do just as good a job on the catwalk as people.
The arts performed poorly across the board, with singers and actors also scoring badly, at 633rd and 480th respectively. Poets did manage to buck this trend, at 239th, as did film directors, at 257th.
Barbers were in the bottom 100, coming in at 869th, while bartenders fared little better at 722nd. However, bar staff were deemed harder to robotically replace than sewage pipe cleaners, who ranked 763rd.
Middling scores were recorded by veterinarians and police, ranking at 339th and 459 respectively. Journalists were deemed to be in the top third of hardest-to-replace jobs, with correspondents considered the 310th most hard-to-replace gig, and editors even safer, at 144th.
The research team also built a website which allows people to see how robot resistant their chosen profession is, and suggests three more resilient options which they may be able to retrain for should the mechanical takeover come to fruition.
Andrea Gentili, an economist at the University of International Studies in Rome, who was not involved in the research, wrote an accompanying article in Science commenting on the research.
She said that the retraining tool is the strength of this new research and may allow people to adapt and thrive in new robot-proof careers.
“Joining the technical analysis with the distribution of occupations among economic sectors may allow us to estimate not only the individual exposure to robotisation, but also the aggregate sectoral/territorial/national exposure to robotisation and, therefore, to design targeted policies,” she said.