Should computers be allowed to make our health decisions for us?

Could computers one day make our medical decisions for us? A New Zealand professor says humans should still have input – lest we give rise to another HAL.

The stakes could not be higher: A high court judge must decide if a seventeen-year-old boy is to receive a blood transfusion to treat his leukemia despite the objections of his parents, who say the operation is against their religion and would damn their child to hell.

Such a choice forms the centre of the plot of Ian McEwan’s 2014 novel The Children Act. It also raises an interesting question: rather than a judge, could machines one day make such healthcare choices for us?

Dr. Tim David, who heads the University of Canterbury’s High Performance Computing centre, believes humans should still have a say on medical decisions – especially ones that could literally mean life or death.

“Computers are extremely helpful,” he said during a lecture hosted as part of Canterbury’s ‘What if Wednesdays’ series, titled ‘What if... computers could save lives?’

“But sometimes, they’re not.”

Image: Professor Tim David, director of the UC HPC supercomputing centre, a high performance computing service facility at the University of Canterbury 

David said he believes that, instead of being the sole decision-maker for our health, computers can be a powerful tool to assist human patients and medical professionals.

One advantage is the sheer amount of data computers can store – and which can potentially be accessed by health professionals and patients anywhere thanks to the cloud. Such remote access can be especially useful to patients receiving care at multiple medical facilities, and to hospitals and doctor’s offices with limited on-site storage capabilities of their own (like those in some rural parts of New Zealand). With IT budgets at hospitals often tight and storage and processing capabilities of computers increasing exponentially each year, cloud storage helps hospitals in, say, Balclutha and Gore stay on par with Wellington and Auckland in terms of storing patient records.

And it’s not just the desktops or laptops you might be reading this story on that are used for storage. Supercomputers are getting in on the action, too – including machines known even to non geeky-types.

In 2011, an IBM supercomputer named “Watson” made waves when it beat the world’s best Jeopardy! players at their own game. The quiz-mastering machine, which boasts 80 teraFLOPs of processing power, today hosts a massive health database that clinicians can purchase time on.

David’s description of Watson’s current usage is even simpler.

“It’s a computer that helps tell you what’s wrong with you.”

Another advantage is a computer’s usefulness as a diagnostic tool. An international collaborative known as interRAI – of which New Zealand is a founding member – includes an assessment of 250 questions, a patient’s answers to which doctors can then use to personalise treatment plans.

But all these advantages come with a cost, argues David.

“If you think a computer is a brain, I’m afraid you’re very wrong.”

Among the biggest problems, he claims, is depersonalisation, and the social isolation computers can create. With too much technology, humans could end up like the Solarians in Isaac Asimov’s “The Naked Sun” a society of hermits completely dependent on technology for everything.

“I think we need to have a debate about how much technology we actually need,” he says, adding that the interRAI assessment does not include a single question about whether a patient interacts with other people or has relationships with others.

The other danger is far more sinister: what if the computer making our health choices decided it wasn’t efficient to keep taking care of us? Such a nightmare scenario would not be unlike HAL 9000 in 2001: A Space Odyssey, where the (possibly?) malfunctioning computer decided accomplishing its mission meant killing people.

The idea of computers taking care of us has been a staple of futurist thinking for decades, but it would seem the future really is here – and advancing further all the time. Even Watson, powerful as it is, doesn’t even crack the top 500 list of the most powerful supercomputers.

But despite the risks, David says computers are an invaluable took we need to take advantage of if we’re to keep ourselves healthy and – ultimately – live longer. The idea of computers taking care of us has been a staple of futurist thinking for decades, but it would seem the future really is here – and advancing further all the time. Even Watson, powerful as it is, doesn’t even crack the top 500 list of the most powerful supercomputers.

“We need to keep the human interaction in health decisions,” says David, who adds there are some equations humans can solve that computers can’t thanks to our ability to think creatively.

“Humans and computers together can save lives.”