������ý

Douglas Hines started out with what sounded like a nice idea.

In the early 2000s, the former Bell Labs engineer was busy caring for his elderly father and building his own technology business. That’s when he first came up with the idea for a companion robot: a machine that could look after his dad and keep him in touch with the outside world via webcam.

Hines started working on a prototype, but ran into trouble finding financial and legal support for the project. So he gave up, and instead turned his attentions to Roxxxy, a life-size sexbot dressed in filmy black lingerie (“always turned on and ready to talk or play!”). That gambit was far more successful. As Hines deadpanned in an interview with IEEE Spectrum in 2010, .

Hines’s story is a good allegory for the wider landscape of care robots: five years later, sexbots, though not yet exactly flying off shelves, have stoked enough cultural interest to inflame a covered . Meanwhile, care robots for the elderly remain stuck in sociocultural purgatory. They’re the flying skateboards of the service industry: always predicted, always trotted out as an example of the future, perpetually just out of reach. It’s time to admit that the problem with this vision isn’t the technology. It’s us.

On the surface, the fates of sexbots and carebots should not be so divergent. Both are mechanised stand-ins for roles that are typically undervalued and ill-treated in society, with neither ethically straightforward to replace. Neither will work without a robot that can move around on its own and do some heavy lifting. Both would work even better with some level of social or emotional intelligence built in, to better respond to human needs.

Where are all the robots?

It’s especially curious that the carebot revolution has not taken place, in light of how direly we need it to. In the UK, the number of citizens over the age of 65 is expected to by 2020; and the number of over-85s by 18 per cent. Reports have identified care for the elderly as .

It’s certainly not a lack of robots that’s causing the hold-up. A bevy of recent prototypes includes Toyota Research Labs’ , wheelie bot Zenbo, which can call for help in an emergency, and the seal pup Paro, which takes on the emotional labour of fuzzy companionship. In a demo video for , a project recently piloted in Italy and Sweden, pick up groceries and mail, relay video calls, take out the rubbish, provide reminders about medication, and take their owners’ arms as they stroll down the street.

But how well will these sell? Not very, if you believe surveys. It seems that people don’t like the idea of carebots looking after their vulnerable relatives. Of more than 25,000 people questioned in a 2012 survey of attitudes in the European Union, 60 per cent thought robots that care for children, the elderly and the disabled should be banned outright; and with one caring for their children or parents (though many more were OK with the idea of a robotic assistant and even a surgeon).

In a separate poll of people in the US, 65 per cent of respondents across all ages agreed that it would be a “change for the worse” if robots became the primary caregivers for the sick and elderly.

Why the squeamishness? We generally look forward to robots doing the chores for us, from answering emails to picking apples to defusing bombs, tirelessly, cheerfully, with uniform precision. (The word “robot”, in fact, is derived from the .) It’s quite all right for a machine to carry out such demands, from the trivial to the tawdry.

On the surface, carebots look like mechanised butlers, too. However, in difficult moments they flip the script – asking us to relinquish control, human connection and our fantasies about ourselves.

Complex dilemmas

Every day, carebots will run into hundreds of small moral dilemmas: their owner decides not to take today’s prescribed medication; she keeps leaving the stove on, or wandering out of the house and down a street heaving with traffic; or he commits a crime in full view of a watchful mechanical eye, as in the film Robot and Frank, in which an ageing thief recruits his carebot as an accomplice.

What mistakes will be acceptable, and which will be grounds for a recall? Will there be limits to a bot’s responsibilities? Or will their charges have to submit to their power?

In the paper “”, and at the University of Sheffield, UK, point out another drawback to life with a robo-caretaker: it’s lonely. Putting a carebot in place of a human might deprive many of one of their few opportunities for regular social contact. Such isolation is linked to poorer health outcomes, such as a greater risk of developing Alzheimer’s disease or dementia. It could also make people feel plain dehumanised – ripped of their dignity, a vulnerable object to be lifted, fed or prompted at intervals.

“If the human rights of the elderly are to be respected as much as the rights of other members of society, it is important to ensure that robots introduced into elder care do actually benefit the elderly themselves, and are not just designed to reduce the care burden on the rest of society,” write Sharkey and Sharkey.

Or, as one person put it recently in The Guardian, being left with a carebot is just “”.

There’s another reason that carebots might not sit comfortably with us: they don’t jive with our flattering visions of ourselves. Looking after another human being is hard work. It’s physically and emotionally taxing, occasionally messy, and can be boring and thankless. It’s also among our . There’s an expectation that this work is a kind of calling, performed out of love or a sense of service by a friend or family member, or at least a compassionate and conscientious worker.

The reality is a harsh departure from that ideal. In elderly care homes in the US, people are more likely than in the wider community to be subjected to  – one in 10, according to some reports.

Carebot dystopia

No one looks forward to a carebot dystopia, in which machines exercise dubious moral power over people. But the alternative, too, can be discomfiting: robots turning out to actually be preferable to human aides. It doesn’t reflect too well on us if our future seniors opt to live in a non-human ghetto, with whatever glitches and lack of contact, over the prospect of abuse by bitter and angry staff.

“We need to think of automation as a political question,” said Lucy Suchman at Lancaster University, UK, speaking at a White House workshop on artificial intelligence in New York City on 7 July. “What grounds are there to believe that a robot can engage in the work of care?” Work like this is difficult for a machine to master because of its nature: heterogeneous, open-ended, and often reliant on the ability to interact with others.

Rather than jump to robotic substitutes, we could think of other ways to sate society’s growing need for workers who care for the elderly, such as revaluing the work involved. “The fact that you get paid a huge amount of money to write code and you get paid nothing to take care of people’s children is not a reflection of the relative skills,” said Suchman, “but rather a reflection of the valuation that we make of those jobs within a particular political economy.”

We should ask whether there are really not enough people to do those jobs, or whether it’s just that those roles have been devalued, she added.

The problem closely parallels the idea of using robots for childcare. New parents are expected to extol the joys of parenthood and gloss over the drudgery, even though the . Tireless devotion is considered a virtue, one that the vast majority of us cannot attain; leaving a child with just a human nanny carries an undeserved social stigma of neglect, even though for many it’s the only practical solution.

What would the neighbours say if they heard that little Jimmy was left with a machine while mum went out for a well-deserved drink? It may not be fair, but it’s not unimaginable. That’s a tough norm for a shiny new robot to break down.

Leaving a loved one in the care of a machine will look tantamount to admitting that we have other things we’d rather do – that all humans have things they’d rather do. Like, maybe, spend time with our new sexbot.

So while sex robots already have enough of a built-in audience that people are fighting over whether we’ll or ban them, the future for care robots is looking a lot murkier. Unlike with sex robots, we don’t know what we want from them.