Essay

Killing time

AI and the future of memory

The days of our lives are seventy years;
And if by reason of strength they are eighty years,
Yet their boast is only labor and sorrow;
For it is soon cut off, and we fly away.

Psalms 90:10

‘HOW TIME FLIES!’ my elderly relatives used to say when I was a child. ‘You’ve grown so fast.’

Even nowadays, when my children, already in their twenties, join me on a visit to my ninety-year-old uncle Zelick in his nursing home, he beams at them, his thumb and forefinger darting out to execute a polished pincer move to the cheek: ‘When did you get to be so tall?’

None of them have grown an inch since reaching puberty, yet every time he sees them he wonders at how they have ‘suddenly’ blossomed.

Zelick’s days are marked off by the familiar signposts of mealtimes and visitors. Recently, I found him trying to make a beeline for the elevator. When I asked where he was going, he answered: ‘Home.’

‘Where’s home?’

He smiled. ‘Zhetl.’

He wasn’t referring to the house in suburban Melbourne where he raised his own family after immigrating from Poland in the early 1950s. In his mind, he was headed home to the Eastern European village where he lived as a child, until the Nazi invasion murdered most of the Jewish population, including his mother and sister. When Zelick was only twelve, he fled to the surrounding forests, spending the rest of the war years hiding in makeshift dugouts. He often returns to Zhetl in his dreams, crying out in terror during the night. With short-term memory loss, Zelick is trapped in the moment: for him, time seems to stand still.

I wish there was a magic wand I could wave to help mitigate the impact my dear uncle’s ghosts have upon his memory and mental health. Zelick would laugh at me; he never was one for the supernatural.

And yet a sorcerer may arrive soon, not draped in wizard’s robes, but encased within a miniscule piece of technology; it may even cast spells that could help my uncle navigate the time warp created by his ageing brain. Artificial intelligence (AI) is being touted as a new magic tool that will not only enhance quality of life, but also help delay the ageing process. With a lifespan of forty years being the norm for millennia until the nineteenth century, what will be the fallout of us all living to become centenarians?

Only time will tell.

 

RESEARCH HAS SHOWN that our perception of time is heavily influenced by our emotions. Whereas all our other senses – such as smell, taste and sight – have specific receptors in the body, it is the nervous-system wiring in our brains that functions as a kind of time machine. The very young and the very old don’t seem to have a true sense of time, perhaps due to the brain’s decreased ability in these populations to rapidly combine sensory information. The awareness of time passing very much depends on the functioning of the pre-frontal cortex, a part of the brain responsible for short-term memory. Elderly patients with dementia are seen to deteriorate rapidly when their disease is compounded by the effects of depression; they can become disoriented and confused, and this can affect their true sense of a time continuum.

Dementia is a group of disorders – including Alzheimer’s disease, vascular dementia and Lewy body disease – that currently affects approximately forty-four million people around the world. It mainly targets the elderly, although younger populations are not immune. It has become one of the leading causes of death in the US, UK and Australia.

If the fear of growing old has always been part of the human condition, what has become uppermost in most people’s minds – more than worrying about cancer – is the terrifying prospect that memory loss and dementia will be an inevitable feature of ageing. In part, this speaks to a fear of losing agency and control as we age, the very notion of our own self vanishing, alongside our sense of time.

This is where AI, or machine learning, can come into play. Computer programs specially developed to react and ‘think’ like humans, analysing information with sensitivity and speed way beyond a mere mortal’s capability, have the potential to detect subtle changes that occur early on in dementia. In this way, wearable technology and devices with inbuilt sensors are able to measure everything from blood pressure to body temperature and gait – no matter where we are. This information can be automatically integrated so that any untoward changes are flagged, and medical staff and family are immediately alerted to potential safety risks such as falls, infection and even a gas burner left on accidentally. Monitoring behaviour in the home or in an aged-care facility can also track changes in cognition. AI’s ability to learn and correct itself as it performs tasks, contributes to both the accuracy and safety features of different devices that can be of enormous benefit to people with dementia and those who support them. It may also be a way of shaking up our perceptions of how we live through the end of our years on this earth, enabling us to monitor so many changes in ourselves and understand what is happening to us.

 

HUMANITY HAS BEEN searching for the elixir of eternal youth ever since Adam and Eve were banished from paradise to prevent them eating from the Tree of Life. Had they munched on that elusive fruit instead of an apple from the Tree of Knowledge, they would have found themselves living forever. The ancient Greeks longed for ambrosia, the so-called nectar of the gods. Ancient Egyptians interred their mummified dead with provisions for an expected literal journey to the afterlife. European alchemists in the Middle Ages tried to distil the philosophers’ stone, which carried the promise of immortality. And only a few years ago, Russian scientists injected mice with 3.5-million-year-old bacteria found in the Siberian permafrost in an attempt to prolong the rodents’ lives. They reported that the elderly mice – grandmothers in the generational scheme of things – not only began to dance, but also produced viable offspring. One scientist, Anatoli Brouchkov, head of the geocryology department at Moscow State University, was so enamoured by this project that he even injected himself: two years later he claimed to be healthier than ever – but time will tell. There is a long history of people purporting to have found such miraculous substances, who have nonetheless dropped dead afterwards.

As average life expectancy has doubled over the last century, the foods, creams, supplements and miracle drugs that promise longevity have become a billion-dollar industry: business is booming in the death-denial racket. Meanwhile, our bulging planet is straining to accommodate almost eight-billion humans: World Health Organization predictions estimate that the number of people aged eighty and older will triple from 143 million worldwide in 2019 to 426 million in 2050. This will have a huge impact on both resources and demands for supportive care.

Can AI help avert this potential health crisis? Existing technologies can be enlisted at every stage of the ageing process, from preventing the high-risk factors associated with developing an illness to assisting with early diagnosis and monitoring disease progression. This can be acquired by using relatively inexpensive, uncomplicated gadgetry that can have a significant effect on quality of life. Changes can appear in the brain up to two decades before a disease such as dementia becomes apparent. Via specially designed apps – wearable technology and smart-home systems with sensors that track vital measurements such as blood pressure, heart rate and even body movement – the number of unnecessary emergency visits and hospital admissions can readily be minimised. Prompts that remind people to take their medication or avoid risky behaviour, like climbing a ladder, can raise an alert. This can help the individual in real time and augment caregiving, further reducing the burden on families and the need for costly aged-care services. Keeping medical and ancillary staff as well as families updated and informed of an elderly person’s wellbeing in real time has enormous potential benefits. Although person-to-person care cannot be replaced, AI can be used to enhance care services and help support the elderly, enabling them to live independently at home for longer.

Recent studies seem to indicate promising results in early detection of cognitive brain problems, with AI or supervised machine learning also becoming a potential game changer in ongoing patient care for those diagnosed with dementia. These specially designed computer programs can rapidly learn tasks such as face or voice recognition, or analyse patterns in diagnostic scans, often outdoing the accuracy and speed of their human counterparts. Early detection of subtle changes can be life saving.

Robots are already deployed in hospitals, taking over menial tasks such as distributing medication or zapping potential airborne pathogens with UV rays. The aim here is to harness technology in order to free up time for staff so they can attend to more complex and challenging tasks. Robots may be faster, cheaper and more accurate when it comes to performing certain jobs, and they never need to take a coffee break, but can they develop compassion and empathy?

US toy company Hasbro launched its therapeutic purring catbot pets in 2015, followed by barking pupbots in 2016. These offer comfort to elderly people who are no longer able to care for a real furry friend. It’s uncanny how lifelike they look. PARO, the fluffy white seal developed by Japanese company AIST, has been used in paediatric intensive care wards and aged-care facilities around the world and responds to stimuli such as touch and light. It can even turn its head towards a voice, adapting its behaviour intuitively according to patient needs. Just like a real pet, PARO learns that flapping its tail gets it stroked, so will flap it more often.

Taking the lead from these robopets, our next option for in-home carers may be a personal companion robot, able to wash the dishes, help us in and out of the shower and share a game of chess. Home systems such as ElliQ, which remind us to take our medications or go for a walk, are being marketed as ‘gizmos that get you’. With a large percentage of senior citizens in developed countries now living alone, and most adults wanting to stay in their own home as they age, such gadgetry can be an attractive option. AI companions like these are increasingly being touted as an aid to empowering people to maintain their independence. They also address common feelings of loneliness and isolation among the elderly: a 24/7 option to not only help us, but to help us pass the time.

 

MEANWHILE, IT SEEMS nothing can prevent my uncle Zelick from blurring the lines between his past and present experiences. There may be hope around the corner, though likely too late for him. The advent of neural nanorobotics suggests a way that patients with diseases such as dementia might regain agency over their distorted perceptions of time. Medical nanorobots applied to the human brain are able to emulate neural synapses, the region where nerve cells (neurons) communicate with one another. While some researchers predict that these technologies will facilitate accurate diagnoses and cures for many conditions that affect the brain, via direct monitoring of its electrical activity, another potential specialised application promises a different kind of relief. Referred to as ‘transparent shadowing’, this technique could allow us to engage in fully immersive experiential or sensory experiences of episodic segments from another willing participant’s memory: we could inhabit someone else’s life. This may sound more like science fiction, or Harry Potter inside the swirling memory-filled waters of a pensieve – but I wonder where Zelick would choose to go, or if I might visit his cloudy memory to better understand the way he feels.

Beyond the potential of any of these specific devices, the advances in AI – together with the availability of databases that generate amounts of information hitherto unheard of – can already contribute to developing a new healthcare ecosystem that may be able to predict and prevent disease years before it develops. Chatbots such as Ellie the virtual online interviewer (designed to identify symptoms of depression or post-traumatic stress disorder) already exist; the use of digital therapeutics is also growing, with virtual reality programs such as gameChange (developed to treat psychosis in a supervised environment) exposing patients to scenarios they may otherwise find threatening in order to help them learn they are safe and reduce their overall anxiety. The Internet of Medical Things already connects medical devices to networks – such as heart monitors, hospital beds and even pills – that can improve the delivery of quality, cost-effective healthcare and allow remote monitoring of performance issues before they become a problem. One example is Abilify MyCite, a smart pill used to help patients with schizophrenia remember to take their medication. Once swallowed, a sensor attached to the tablet sends a signal to a smartphone app to confirm it has entered the stomach.

The integration of such new advances raises important concerns, and crucial ethical considerations would need to be ironed out before widespread adoption of these potentially revolutionary technologies. If physicians use AI software as an aid in diagnosing conditions, what are the consequences of incorrect management decisions being made from the resultant information, and who bears the responsibility for any mistakes? At the same time, home-based and wearable technologies open up issues of potential infringement of an individual’s privacy. And what is the likely impact on medical insurance? We are on the brink of a revolution in healthcare, but these issues must be publicly addressed and debated.

 

THE REGULAR SENESENCE of cells – of every biological organism – may be one of life’s givens. If there is a birth, the future holds a death, irrespective of the length of time between these two events. But the nature and timing of the pathway between them is individual and unpredictable for everyone. It’s like the xylophones and horns that Terje Isungset, a Norwegian musician, carves out of ice. Each instrument is unique, so he can never know exactly what it will sound like when he plays it. ‘It’s never possible to plan in detail how things are going to be,’ he says.

Ageing is a bit like that. As a GP, I’ve seen patients who haven’t needed to visit a doctor their entire life suddenly present with an out-of-the-blue symptom that brings their mortality into terrifyingly sharp focus. Others, by contrast, are ravaged by disease from a very young age, suffering years of chronic pain and mental anguish as a result.

Taking our health for granted is part of our tendency to avoid thinking about the natural processes of ageing and death. Yet, as the line between humanity and machines becomes increasingly blurred, the arrogance of perfect health could well become a future reality for all. We may even dispense with what has, since the beginning of time as we know it, been the biological inevitability of death.

In a recent experiment that brings Dr Frankenstein’s animation of an inert creature far closer to reality, scientists at Yale School of Medicine demonstrated that a pig’s brain can be partially restored after a prolonged post-mortem interval. By essentially reversing cellular death, the brain can theoretically be kept functioning endlessly. Thirty-two pig brains – dead for hours after being removed from an abattoir – had their circulation and oxygen supply restored via BrainEx, a system similar to a dialysis machine. This result may be promising for future study of diseases affecting the brain, but it also opens up more ethical and philosophical conundrums about our current definitions of what constitutes ‘brain death’. This kind of disembodiment from the brain brings to mind the haunting lyrics of Nick Cave’s song ‘Death is not the end’. He may well be right, although I wonder what Uncle Zelick would have to say about the possibility of this kind of reincarnation and revivification. Some days when I visit, we watch old black-and-white Yiddish movies. Zelick’s favourite is The Golem (1915), about a clay statue brought to life by a learned rabbi to ward off anti-Semitism. The golem is rediscovered four centuries later by an antiques dealer searching a ruined temple, and soon falls in love with the man’s beautiful daughter. When she spurns him, the creature turns into a bloodthirsty monster. Each time we watch this story, my uncle remarks that the poor golem never should have been woken up: ‘It’s too dangerous to meddle with the unknown.’

I take his point. The AI revolution could well backfire, with rising concerns around bioethics, unemployment, breaches of privacy, and undue corruption and political influence. If popular movies like The Matrix and Terminator have any predictive value, nightmarish rogue AIs that gain self-awareness might end up biting us all. But despite the naysayers, this scenario is highly unlikely.

Instead, AI has enormous potential to help us grow old more gracefully. It may even be able to help us lead more fulfilled and meaningful lives for however long each of us has on this Earth, instead of spending our sunset years just killing time until time kills us.

Get the latest essay, memoir, reportage, fiction, poetry and more.

Subscribe to Griffith Review or purchase single editions here.

Griffith Review