Dad’s ‘third daughter’ always came to look after him. Then computer said no

It was one of the most unusual friendships – between a 94-year-old man and a 20-year-old woman. But, somehow, it just seemed to work.

He soon started talking about her as his “third daughter” – although to be more accurate it should have been his second granddaughter. In turn, she, who doesn’t know her own father, said she simply enjoyed his company, his stories about the old days, his wartime reminiscences and the extremely colourful way in which he expresses them all.

An elderly neighbour started sobbing when she was told she would no longer be seeing the care worker she’d got to know.

They’d meet up once or twice a week and, for both, it started proving an immensely fun time. Until, that is, AI joined the party, and ruined it for them both.

AI undoubtedly has a role in many businesses, undertakings and spheres of life. It can grasp complex situations in milliseconds, can digest, assess and analyse, and can often provide great solutions to almost impenetrable problems. Accordingly, by 2030, it’s forecast to have a global market value of up to $US4 trillion ($5.97 trillion).

But it has one huge, massive gaping hole in its operation: it doesn’t have a heart.

That’s something Australian Unity, one of the country’s biggest, and most award-winning, aged home care service providers, needs to hear. One of its core businesses is giving the elderly the means to stay in their own homes, with a number of carers who go out to help them with their shopping and their cleaning, visit them for a chat, take them out for a sociable lunch, or drive them to important medical appointments.

For the recipients, these carers can sometimes be the only people they see all week. As a result, they offer absolutely invaluable assistance that they come to rely on, as well as exceedingly precious friendships.

Accordingly, Australian Unity has always advertised on its website that it has “local offices” and employs “local people who deeply understand their community and what’s important”. Yes, it’s true that it has dedicated workers who get to know their elderly clients, their history, what they need, and their likes and dislikes. In my experience, these co-ordinators have always carefully matched their clients with the carers who most suited them and watched as they regularly forged strong bonds.

Until now.

Now, the company is experimenting with machine learning, a subset of AI, with algorithms borrowed from the aviation and fast-moving consumer goods industries. These now rewrite all the aged carers’ schedules to dictate which carers visit which clients to maximise efficiencies, regardless of how suited they are and the years they’ve been together.

As a result, my nonagenarian father has lost the young woman to whom he became so attached, and she has been banished elsewhere to get to know a completely different set of people. When she and my father, by chance, bumped into each other in a shopping centre one day, and had a teary reunion, it nearly broke my heart.

An elderly neighbour similarly started sobbing when she was told she would no longer be seeing the care worker she’d come to know, and become so fond of, over the past four years. And another carer has been so devastated to have her regular visits to another beloved client axed, she’s threatening to leave altogether.

Yet when we’ve all protested, we’re told, sorry, it’s now AI running the show, and it’s impossible to overwrite, or deviate from, what the computer says.

At the top, Prue Bowden, the company’s CEO, Home Health, insists this new reliance on algorithms to deliver care, which started this year, is simply necessary. “It takes courage to innovate in sectors that are in need of change,” she says. “While I hear and appreciate and acknowledge that there is some negative feedback, these changes have delivered real benefits to customers and staff and are all about preparing the sector for a very significant demography shift that’s coming.”

Of course, that’s a fair point. We’re confronting an ageing population and it is going to become exceedingly expensive to deliver more care to the Baby Boomer bump, which make efficiencies necessary.

As Bowden says, jettisoning the old one-on-one care for a business model of team-based care, where algorithms make the decisions, is about the necessary modernisation of the aged-care sector and its increasing complexity and preparing for greater scale for the future.

“There’s always resistance to change, but we’ve had some good feedback,” she says. “The reality is that people resisting the change are the loudest.”

I disagree. My dad, like his neighbours who are distraught about the loss of some of the most treasured figures in their lives, aren’t loud people. They’re elderly, they’re fragile, they feel absolutely helpless, and they express themselves in tears more than angry words.

There are only two hearts in this photo. Credit: Getty Images

Most of them, moreover, are afraid to complain, fearing that, if they do, they’ll lose whatever carers the algorithms deign to bequeath them.

Naturally, few have any idea what AI and machine learning actually are. But they do know that they’ve robbed them of some of their most beautiful friendships, which had conferred meaning and pleasure and, to what time they have left, a reason to get up in the morning.

Sue Williams is a Sydney-based freelance travel writer, author and journalist.