Tuesday, October 9, 2018

A revolution is coming, warns emeritus professor


You're out of touch if you believe medicine will remain unscathed amid the rise of artificial intelligence, he insists
So you thought that My Health Record was complicated and risky? In the digital revolution, it is chicken feed. 
Steady yourself and gird your loins, because artificial intelligence (AI) is the big game that’s coming to town soon and it can be challenging.
How close are we? A statement in the IT world called Moore’s law observes that the number of transistors that can be placed on a single integrated circuit doubles about every two years.  
It was named after one of the co-founders of Intel, Gordon Moore, following his 1965 paper. It means that the same-sized circuit you were using last year has doubled its capacity this year. 
According to Google, your smartphone has enough computing power to fire a person to the moon.
How long Moore’s law will apply is unknown because the space on printed circuits is finite, but we do know that today’s computers have the same processing power as the human brain. 
In the August issue of Foreign Affairs magazine, Kevin Drum, a 60-year-old Californian political blogger and columnist who knows a lot about Silicon Valley, credits the immense social progress of the 19th century to the Industrial Revolution.  
“Without it, there’s no capitalist revolution because agrarian states don’t need one. Without it, there’s no rising middle class and no real pressure for democracy,” he wrote in an essay called ‘Welcome to the Digital Revolution'.
“The key drivers of this era were the steam engine, germ theory, electricity and railroads.”
And now? The computers to support AI are ready.
Their power is measured in floating point operations — known in the trade as ‘flops’ — which basically means that they work very fast. For example, one second is the equivalent of about 10-100 petaflops.  
The capacity of the human brain is said to be able to handle 100 petaflops per second. That is, it can perform 100,000,000,000,000,000 operations per second.  
According to Mr Drum: “A computer with this capacity, unfortunately, is the size of a living room, costs $200 million and generates electricity bills of about $5 million (a year).” 
Software development is critical and AI experts say there is a 50% chance that AI will be able to perform all human tasks by 2060, he adds.
“The digital revolution is going to be the biggest geopolitical revolution in human history”, he says, adding that PricewaterhouseCoopers has predicted that 38% of all jobs in the US are at high risk of automation by the early 2030s. 
The effects on human employment will be profound.
Within a decade, he says, long-haul truck drivers will be displaced by driverless technology and similar technology will knock out the jobs the displaced drivers might have taken up. We need new politics.
Anyone imagining that medicine and medical practice will not be profoundly altered is out of touch. 
Our eldest son Nick, a vice-president with Google, recently told me that the AI development of the driverless car was now sophisticated enough to engage in ethical reasoning.  
For example, how should an AI-driven vehicle respond to an impending crash where either the humans in that car, or the colliding vehicle, will sustain a fatal injury? With sacrifice, altruism or self-interest?
And, if ethical reasoning can be used by AI for driving, then why not in medicine?
The two most important developments for the 21st century will be AI-driven mass unemployment and fossil-fuel-driven climate change, Mr Drum says.
A glimmer of hope is that AI might be able to solve climate change by scaling up wind and solar power.  
But what about medicine? Now, there’s the challenge for us doctors. 
At the very least, our medical education should accommodate more about the interface between practice and AI.
This must go way beyond the simplicities of how to use IT to include debating and considering the implications for what we do as doctors in this brave new world. 
What will ethical practice mean and how will we relate to AI in this pursuit? 
It’s time for a lot of serious and creative thinking. 
Source: Foreign Affairs 2018, online .
Related reading:
Published in The Medical Observer 13 August 2018 https://bit.ly/2PnMULp

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.