Robots are a threat to human interaction

Last week, I wrote a column about the job-killing potential of more robots in the workplace. Some readers sent emails sharing their own concerns.

But it’s not just jobs that are affected. Other human needs are at stake. For instance, consider what might happen to patient care.

Computers can already scan laboratory test results and some diagnostic images far more rapidly and with greater accuracy than humans. Robots have been designed that can detect some neurological disorders merely by “listening” to voice recordings from patients.

Smart machines can diagnose some forms of cancer with more precision than oncologists. And the da Vinci robotic surgery system can assist in the removal of prostate tumours with fewer harmful side-effects for the patient.

I see little real danger in these developments so far. It makes sense, for example, to augment the work of elderly surgeons whose hand control has diminished with robots that suffer no such ill effects.

However, if we begin to see robots gaining a more extensive presence in hospital wards and long-term-care facilities, I believe there is cause for concern. Patient care in its most essential aspects centres on compassion and human contact. Machines might learn to mimic some of those interactions, but they are still machines.

Yet the temptations will be enormous. On average, acute-care nurses in B.C. earn about $105,000 per year, with pay and benefits included. But at least some of their work, and conceivably all of it, could be robotized.

It would take no great advance to design intelligent machines that can give injections, detect fever or other symptoms, read physician instructions and carry out other basic elements of patient care.

Indeed, researchers at Rutgers University have constructed a “venipuncture robot” that uses near-infrared and ultrasound imaging to locate a usable vein and draw blood. This new technology makes it easier to get samples from difficult patients, such as small children.

And here another factor comes into play. Nurses are more likely to take sick days or go on extended leave than most other classes of employee.

The reasons are obvious. Patient care, particularly in intensive-care wards such as burn units, is immensely stressful. And this stress has only magnified as hospitals have struggled to keep costs down.

In one sense, that effort has been successful. In British Columbia today, there are 30 per cent fewer people in hospital beds than there were 15 years ago. And the province’s population has grown over that period by nearly 20 per cent.

But that means those patients who remain in hospital are, on average, much sicker. The “walking wounded” have been sent home with painkillers and a roll of bandages. And this, in turn, increases the stress on caregivers.

But robots don’t get stressed. Neither do they take sick days or expect extra pay for overtime.

In short, as artificial intelligence expands in scope, there will be growing pressure on hospitals (and long-term-care facilities) to replace human care with robot care.

Yet this would be disastrous. Hospital patients who are seriously ill are often in great discomfort and fear for their future. Many elderly residents of longterm care homes have no loved ones close by to offer support and consolation.

We must be able to rely on the sense of personal contact that only human caregivers can provide.

Here, then, is another risk factor embedded in artificial intelligence and smart machines. The day is not far off when tasks we thought were uniquely human — such as caregiving — could be automated.

In that respect, the threat of robotization is not just that it might kill jobs, but that it might undermine some essential aspects of human interaction.