Artificial intelligence has the potential to greatly benefit humanity, but governments and experts need to act now to ensure its associated risks are mitigated, a lead researcher in the field said.
Michael Witbrock, the senior manager of cognitive systems at IBM, is in Melbourne this week for the IJCAI 2017 conference and says that Australia has an important role to play in testing new AI technologies and how they change societies.
The ethics of artificial intelligence and how to best regulate the emerging technology has thrust into the spotlight this week following a letter sent to the UN by Elon Musk and a group of 100 AI experts.
The letter called on the UN to ban the use of AI in weapons before it got out of hand.
“Lethal autonomous weapons threaten to become the third revolution in warfare. Once this Pandora’s box is opened, it will be hard to close. Therefore we implore the High Contracting Parties to find a way to protect us all from these dangers,” the letter said.
This is an important issue, Mr Witbrock said, and now is the right time to have the discussion about it to ensure the full benefits of artificial intelligence are realised.
“These are legitimate things to think about because this is going to change the future fundamentally, but I think for every risk there’s also an enormously balanced upside,” Mr Witbrock told InnovationAus.com.
“We should proceed deliberately and with thought, but this can be the best thing that ever happened to humans, and we need to make sure that it is.
“The purpose of governments is to understand the desires, needs and wants of their populations, and ensure people are able to realise their wants and desires in a way that is coordinated and effective, and that we’re pushing together as a group and helping each other, not harming.”
Ensuring these risks are balanced and the right regulations are in place early is crucial to reach the true benefits of AI, Mr Witbrock said.
“Artificial intelligence has the promise to be a uniquely powerful technology. There’s a good chance that AI is in the order of the invention of agriculture or language – it’s going to fundamentally change everything, and when you have fundamental change there are always risks and opportunities,” he said.
“We’re focused on making sure that the opportunity that AI offers for improvement to the human condition is realised, and in a way that contributes to the shared prosperity of humanity.”
Governments around the world are currently grappling with the technology and the associated fears of automation and job losses.
“Government’s’ role is certainly relevant in deciding how we should apply and deploy technologies, cope with the way work changes over time and cope with how healthcare changes over time. Governments are and should be engaged in dealing with societal changes,” he said.
“It’s never too early to start that process with any technology, and the earlier you start it the better decision one hopes we will make.”
Mr Witbrock is leading IBM’s team focusing on machine reasoning techniques and innovations and the application of AI to complex real-world problems in medicine, for example.
Its initial test case centres on using these technologies to better diagnose and treat patients with serious diseases.
“What we’re trying to do is to marry learning and these sort of AI techniques which have been developed over the last 17 years together to make systems that can carefully reason in ways that human beings do when they’re trying to solve important problems,” he said.
This could allow artificial intelligence to assist with the treatment of serious diseases in a way that humans can’t, he said.
“Imagine a system which can look at a patient with cancer, sequence the populations of cancer they’ve got and then plan a strategy for defeating each of these series of cancer before they can start to spread. That’s not able to be done at scale by people, but it’s clearly within reach with extensions of the sort of technology we have these days,” Mr Witbrock said.
“We’ll be able to fully understand how some diseases work so we can design treatments that are extremely likely to be effective in a particular person with extremely low unintended effects.”
IBM is currently testing some prototypes of the technology and running early trials of it.
In June it announced a partnership between its Watson system and Icon Group for a cognitive computing platform available to cancer specialists.
This will allow access to “massive volumes of cancer research and literature across the globe” and support the decision-making process by giving “evidence-based personalised treatment”.
Earlier this year the firm conducted research into the use of deep learning and visual analytics technology to advance the early detection of eye disease and found an 86 percent accuracy rate.
And last year the organisation signed two agreements to undertake further research to advance the identification of melanoma using artificial intelligence.
Mr Witbrock, who was born in New Zealand, said Australia and its surrounding regions have great potential to act as testing beds for potentially life-changing technology like this.
“Smaller countries actually have an opportunity to try experiments for the rest of the world – it’s easier to try things out in smaller countries with good developed infrastructure. Maybe Australia has a role in leading the thinking about what a future with AI should be like,” he said.
“Historically this area has been a relatively progressive area in thinking about how societies should be.
“It’s worth Australia thinking about what it can contribute to this debate in the world, and I’d encourage them to think about how they might help us find out how to produce the best possible future for people using artificial intelligence.”