Machines have a long way to go to replicate 'massive complexity' of natural world
Genomics Medicine Ireland CEO Dan Crowley tells our reporter that the challenge is to enhance human ability
It's almost 20 years since Kevin Warwick, then a professor of cybernetics at Reading University, had a chip surgically implanted under his skin and caused a global media frenzy.
The chip emitted a signal that a computer could recognise to operate devices such as room lights, a lift or door locks. It was, we were assured, the dawn of the hybrid human, where we would accentuate our natural abilities with the use of artificial intelligence and implanted devices.
But two decades on, we still use credit cards, use our phones to make cashless payments and still carry paper money.
It's a common human trait to overestimate the pace of future change.
In 1988, the 'Los Angeles Times' predicted that by 2013 we'd be using robot butlers in the home to clean the windows and change the bedsheets. To be fair, it got some things almost right, but it underscores the difficulty in predicting what the future will look like. Who would have imagined not just the advent of the smartphone, but the impact it has had in just the space of a decade?
At the Futurescope conference in Dublin yesterday - a gathering that promotes collaboration between entrepreneurial companies and large enterprises and multinationals - one of the panel debates was titled 'Machines Versus Humans' and asked if intelligent machines will destroy jobs or create higher value ones.
One of the participants was Dan Crowley, the CEO of Genomics Medicine Ireland (GMI).
Last year, GMI raised $40m (€36.8m) of funding to undertake an ambitious project to map the genomes of about 45,000 Irish people. The project is being backed by drug giant Abbvie.
The aim is to better understand the genetic make-up that predisposes people to specific diseases, and by identifying them, enable drug companies to eventually develop specific molecular treatments for them.
In the longer-term, it raises the prospect of families being able to home-test for various genetic markers that could cause concern, and thereby being able to avail of treatments or procedures to pre-empt them.
The long-term goal of the GMI project raises a whole lot of questions - from scientific to ethical, and also about the increasing advance of technology and its impact on our lives.
Crowley, a former director of the National Digital Research Centre, doesn't think that we're on the path to becoming cyborgs or living in a world where machines are superior in everything they do, however.
"What we don't specifically understand is the massive complexity of the natural world," he says. "When we talk about trying to replicate human intelligence, or simulate reality in profound ways, most of the time we miss the absolutely massive gulf in complexity between the simulated versus the reality that we're surrounded by.
"These are not things where in reality something is 10 or a hundred times more complex than in a simulation; the real system is trillions or a hundred trillion times more complex," he adds. "The growth in AI (artificial intelligence)in capability of computation is incredible. GMI and a lot of genomics wouldn't exist without those capabilities, but reality is so much richer."
Crowley points out that we don't understand how humans work on a whole lot of levels, so how we can expect to replicate that in a machine environment?
"We're a long, long way from being able to recreate intelligence in a profound way," he believes. "I think we can have things that are genuinely useful and do real work, but in terms of being able to replicate nature, we're a long, long way away."
And how far do people want artificial intelligence to develop? We might trust a computer to decide when to turn on our heating, order food for the fridge, or park the car. But who yet would trust a computer to perform heart surgery on a child, or fly an airplane completely unaided by humans?
"It's different when it comes to life or death stuff, for sure," says Crowley.
But AI has its place, evidenced by the computing power and systems needed to map the genomes that Crowley and his team at GMI are harvesting.
"The complexity of natural systems like genetics is so huge that it would be impossible to do the work we're doing without massive computing capability," he says. "We're inherently in a situation where the intelligence that we're using to solve these problems is a hybrid of human and machine intelligence already.
"We've got to change the argument. It's not just an argument of machines versus humans. It's actually machines and humans. It's really about enhancing our own abilities in an exponential way," says Crowley.