Should Robot Workers Have a Gender - Or Isn't Their Role Already Controversial Enough?

Robotic gender discrimination? JL

Emma Jacobs reports in the Financial Times:

When it comes to talking about the structure of work in the next machine age, gender has largely been left out. (But) assigning a gender to machines may reinforce workforce stereotypes
If you were to design a robot to assist in a kindergarten, handing out water, tidying and watching out for grazed knees, what would it look like? A bulky, shouty metal hulk? Or a gentle-voiced android that mimicked a carer?

Assuming you are not the Child Catcher, the answer is likely to be the latter. If robots are to provide support in the workplace of the future, they might prove less shocking if they attempt to blend in. Carers tend to be women, so perhaps engineers should recreate them in metal form: with female voices and bobbed helmet hair.To be human is to anthropomorphise. In an experiment published this year in Nature, participants were shown photographs of a human and a robot hand being cut with a knife. The volunteers felt empathy for the human and the robot.

The extent of this tendency to impute human characteristics to machines was brought home to me earlier this year when I visited Xchanging, a company in the City of London that provides business process services, procurement management and IT outsourcing. The machines that replaced the mundane and tedious work of data entry were no more human than a hard drive, yet they were dubbed Henry and Poppy. Even the French boring drills used in the Channel tunnel were given names.

There have been cases where a machine’s “gender” has caused problems. In the 1990s, BMW reportedly had to recall German cars installed with its navigation system because back then, it seems, male drivers objected to taking directions from a woman.

There is a danger that by assigning a gender to robots we reinforce stereotypes in the workforce: a problem not just for women but men too. Research by the University of Bielefeld, in Germany, published in the Journal of Applied Social Psychology, investigated attitudes to robots when they were assigned a gender. In the experiment, they showed two robots to a group of men and women. The robots looked the same except for their hairstyles and the shape of their lips. The one with longer hair and fuller lips was seen by the groups as “female”, the short-haired one was “male”.

Consequently, the male robot was perceived as capable of more “masculine” tasks — repairing technical devices, guarding a home. The female robot was seen as suitable for stereotypical activities such as housework and caring.

While humanoid robots are not about to take over our nurseries in the immediate future, such studies raise important questions about gender and technology. Wendy Hall, professor of computer science at the UK’s University of Southampton, notes that there is a tendency to see science as pure and objective. Yet even algorithms — a set of instructions to be applied to data — reflect human assumptions. They model the world and can be infused with unconscious bias, she told me. This was brought home in a study of Google’s algorithms, for example. Researchers at Carnegie Mellon University and the International Computer Institute found that male job seekers were more likely than their female peers to be shown ads by the search engine for high-paying executive jobs. When it comes to talking about the structure of work in the next machine age, gender has largely been left out. Hopefully, this will be redressed in part by the World Economic Forum, which is researching the topic ahead of its annual meeting in Davos in January. In a paper published in November, it stated: “What is missing from the debate is a gender lens.”

It is not yet clear how automation will affect men’s and women’s work. Martin Ford, the author of The Rise of the Robots , which won the FT and McKinsey business book of 2015 awards, believes that industries where men tend to work — manufacturing, finance and warehouses — are particularly susceptible to automation, either via actual machines and robots or software. “In the near-term automation seems likely to fall more heavily on men,” he says.

I doubt this spells “the end of men” to borrow the title of Hanna Rosin’s 2012 book, which argued that women would dominate the future workforce. The male-dominated technology sectors are in the ascendancy, after all.

Yet it does highlight that the future of work should not be viewed just through the prism of “man and machine” but of women too.