Kiwis need to think about what they want from the age of AI, report says

Monday, 5 August 2019

Emerging technologies such as artificial intelligence present new, ethical dilemmas. It's partly up to us to be vigilant about sharing our personal data, says a new report from Royal Society Te Apārangi

Being more vigilant about who we share our data with, along with regulation, will help keep artificial intelligence technology on the right track, a new report says.

The report, from the Royal Society Te Apārangi, came out of a project led by the Australian Council of Learned Academics called 'The Effective and Ethical Development of Artificial Intelligence'. It examines what the growing use of AI means for New Zealand and how to ensure the technology benefits everyone. Few sectors will be unaffected in the coming decades, so safety nets and AI-aware labour laws need to be in place, it found.

While it's difficult to predict the changes AI will bring, there are clear risks that need to be managed so all Kiwis can thrive in an AI world, the authors said.

They defined AI as the term used for a whole collection of new computer methods and techniques that would otherwise require human thought: 'AI doesn't mean that computers will begin thinking for themselves, although philosophers, biologists and computer scientists are debating this possibility. Rather, AI helps us to process information much faster than is humanly possible, and select and weigh options.'

* Algorithms are everywhere but the public sector seems scared to use them

* Data for sale: the value of our digital lives**

Digital humans, such as those created by New Zealand company Soul Machines, are changing the face of the customer service industry

The technology is already being used to digitally identify biosecurity threats; triage hospital patients; deliver one-on-one interactive tuition for students; optimise energy production, consumption and storage. On a more personal level, it also drives Netflix recommendations and match-making on dating apps.

But its application is raising questions about the future of work, personal privacy, consent, and bias. The success of AI systems depends on large and broad data sets, the authors noted. 'This prompts questions about the legal, ethical and economic implications of data collection that people are often unaware of.

'Another consideration is the right of indigenous people to have control over their data, such as information about their DNA.'

(In New Zealand, Te Mana Raraunga, the Māori Data Sovereignty Network, argues Māori data are living taonga, and should be subject to Māori governance.)

There's a sense we're on the cusp of the fourth industrial revolution, says Professor James Maclaurin of the Department of Philosophy at the University of Otago. Maclaurin, also co-director of the university's Centre for Artificial Intelligence and Public Policy, worked on both the Australian and New Zealand reports. 'A feature of those big changes is they're unpredictable,' he says.

The report put forward broadbrush recommendations such as 'building robust systems for the capture and use of data that people can trust'. People need reassurance their data will be handled appropriately and will contribute to their wellbeing or the wellbeing of society overall, not just profit a small number of companies.

But people should also be vigilant and only share their data when the above conditions have been met. That's where regulation comes in, Maclaurin says.

While Australia is focused on winning the so-called 'AI race', New Zealand isn't big enough to think along those lines, he says. 'We're thinking more about niches where we could be valuable and what we could be doing now to regulate or organise to make the country more attractive to developers of AI.

'Most countries — including New Zealand — have very little regulation in this area which isn't actually what businesses want. They want certainties.'

Last year, the Artificial Intelligence Forum of New Zealand published a study of the potential impact of AI with a key recommendation for a coordinated, national AI strategy. That hasn't happened and the role that might have driven its creation, national chief technology officer, was scrapped by the current government.

Mclaurin says whether it's through 'one big document or lots of little ones, we do need to focus on it more than we currently are'.

'At the moment, almost all of New Zealand's focus is on research into the development of AI, rather than on research into its impacts. That has to change.'