AI is already accelerating the pace of medical and pharmaceutical research. It will also improve healthcare, and ultimately make it cheaper.
By Dona Suri
It’s only been about a year since Chat GPT, an AI algorithm designed to understand and generate human-like language, burst across our computer screens.
A “generic” description of Chat GPT is “large language model that is a type of algorithm employing deep learning techniques and massively large data sets to understand, summarize, generate and predict new content”.
Chat GPT was the first but not the last. Over the past year, other versions of generative AI have followed, all of them claiming to be improvements on the original and competing for market share.
From Day One, Chat GPT polarised opinions as few technologies have ever done before. Even people who are not knee-jerk techno-reactionaries said “Whoa … let’s think about this”.
The hesitations and desire to keep a judicious hand on the sluice-gate lever are easy to understand. Easy to understand
…and irrelevant
… and too late.
For decades now, AI in its various forms, has been oozing into every field of endeavour — commerce, entertainment, education, research, defence, medicine, logistics … everything.
Now, all of a sudden, it’s scary. Even the big brains whose work contributed to the emergence of AI are warning against it.
AI replacing humans in every sort of job – even jobs requiring creativity and imagination.
AI replacing humans in decision-making.
AI rendering human life meaningless.
AI taking over the world !
The world under the command of mankind: how has that turned out?
Q. What is disrupting ocean currents, destroying glaciers, forests, rivers, wiping out biological diversity, pushing the planet closer and closer to anihilation?
A Human activity.
Q. What violently snuffs out the lives of scores of people every day?
A. Human activity. Specifically human activity in Gaza, Ukraine, Chad … or every other war zone.
Humans place themselves at the top of the Intelligence Scale.
Intelligent ? Sort of. But obviously not so wise.
If mankind were so smart, would the planet be facing with global warming and climate change now?
Would we be seeing humanitarian catastrophes, large scale and small?
Such a strange species, mankind. Full of compassion for cows and cats and dogs; indifferent to exploitation, abuse and cruelty heaped on fellow man.
This is by no means an original thought: for centuries, innumerable thinkers – religious and secular – have pointed out the inhumanity of humanity.
Maybe a world with AI in charge would be a healthier, kinder, more just and more honest world.
But, to repeat: fear AI, don’t fear AI. It makes no difference. AI is here. Get used to it. Fortunately, when it comes to getting used to stuff, the human race has had plenty of practice … practice stretching back to the stone age.
This is a blog often focuses on advances in medicine and pharmaceuticals, so look back over the past century …
If you are lucky enough to be nearing your 100th birthday, it means that you came into the world in an era when insulin did not exist. The condition called diabetes was recognised but poorly understood and the only treatment for it was to keep diabetics on a near-starvation diet. The first insulin was produced in 1923 and it used insulin taken from animals. It would be another 55 years before synthetic human insulin was developed.
If you are 45 or older, synthetic human insulin did not exist before you were born.
If you are around 85, you were born before the era of sulfa drugs. They were called “wonder drugs” because they cured “incurable” infections.
If you are around 75 years old, penicillin did not exist before you were born.
If you are around 70, then your parents would have carried you to the nearest school to get polio vaccine on a sugar cube. Before that, hundreds of thousands of people succumbed to polio, or were paralysed for life.
You only have to be about 40 to look back to a time when there was no vaccine for HIV.
Even a kindergartener can play old-timer …”When I was a baby, millions of people died from covid; now I get a shot and stay healthy.”
AI equals data. If a limited amount of data is fed in, then AI learns slowly. For rapid learning, AI requires Big Data. That’s where you and I come in. Every digitally recorded bit of our lives is learning material for AI.
Again, think of medicine.
Mr X came out of a car accident with a broken collar bone.
The police filed a computerised report on the accident – what time, where, how, circumstances such as weather, etc, including previous accident/traffic violation history.
The hospital filed a computerised report on emergency room treatment and follow-up visits until the bone was healed. The report included a digitalized X-ray and complete list of medications.
Two insurance companies filed reports: the car insurance company computerized everything relevant to repair or sale of the car; the health insurance company computerised everything related to outlay for Mr X’s treatment.
Hundreds of thousands of similar reports are filed every year. Integrating the maximum number of such reports results in Big Data. Artificial Intelligence sets to work on Big Data and in much less time than an army of human analysts would take, it spins out conclusions deduced from the data, plus safety recommendations.
What about AI as Doctor’s replacement or at least Doctor’s assistant? About ten years ago Disney made a children’s film about a Japanese boy who builds and programs a nursebot named Baymax. In 1980, Star Wars viewers were introduced to medical droids , the 2-1Bs. As far back as 1934, sci-fi fans read the tale of a self-aware robot surgeon, hero of Harl Vincent’s story, “Rex”.
A big part of the apprehension about AI comes from the way it often looks, and feels: creepy.
Picture yourself rolling down the highway: car slowing down, speeding up, braking, passing, signalling. Your hands are NOT on the wheel. You are doing nothing. Is that creepy? No, you’re just on auto-pilot. It’s been a reality for years now. It’s not quite full-fledged AI yet, but it won’t be long.
Now picture a hospital or pharmacy in the not-too-distant future. The processes of inventory, e-prescription translating, record management, workflow, ordering and many other processes are all performed by systems that use AI. If something doesn’t look right in an individual’s prescription, a warning pops up automatically. AI is built into the pill-counting machine and it not only counts, it scans too. If a capsule of one drug accidently mixes with a batch of another drug, the machine “sees” it and sounds the alert.
In hospitals or at home, patients are monitored round the clock by machines that not only read and record their vital signs but contact medical staff the instant any critical signal is detected.
AI, programmed to quickly review millions of cases, will soon enable doctors to diagnose quicker and more accurately.
AI has tremendous relevance in the field of health and medicine. Healthcare professionals are in short supply all over the world, training them takes a lot of time and money, and this high cost is eventually passed on to the patient. AI will allieviate this shortage, stretch the capacity of human doctors, nurses, technicians and pharmacists and ultimately make healthcare cheaper. AI will free up doctors and nurses and give them time for compassion and empathy; they will be able to do human things better. AI gives medical researchers time to do do more complex higher-level thinking. In fact, AI has already become an invaluable tool in the process of drug discovery, drug testing and research into the biochemical, molecular level processes of the body.
So what has NOT been turned into a computer code and a collection of circuits by AI?
Us.
No matter how good an AI becomes, it’s never going to be there to hold a patient’s hand when they’re going through a hard time or to listen to their concerns. That’s a job for humans.