The very idea of AI is courageous. And exciting. And scary.

I’m a fan of that classic parody of parliamentary life Yes, Minister. Particularly the acerbic adviser Sir Humphrey whose response to his minister’s ambitious vision and bold plans was inevitably “courageous decision Minister”. Delivered dead pan. Subtext: you haven’t thought this through.

When I talk to people about the amazing opportunity of Artificial Intelligence (AI), the profound changes it promises for how we live and work - and how we are already using AI at ANZ – it is Sir Humphrey I hear.

Courageous. The very idea of AI is courageous, after all we barely understand the nature of our own intelligence. AI is exciting. And scary.

But we already know AI can produce remarkable results. The accuracy of AI in diagnosing cancer from scan data in some cases is already vastly superior to humans. Google has used AI with its mapping app and has been able to identify neighbourhoods with higher incidence of obesity – and the reasons why.

In China, at least seven patients who doctors said had “no hope” of regaining consciousness were re-evaluated by AI. It predicted they would awaken within a year. They did.

These developments are truly amazing, the benefits clear. But I am equally cautious of the unforeseen and unintended consequences.

For example, a Canadian company developed a simple test to diagnose emerging Alzheimer’s with more than 90 per cent accuracy. However it turned out it only really worked for Canadians – because the data the AI worked with was sourced from Canadians. The Chinese AI system predicted some patients would awake who didn’t. We can’t know if there were some who neither AI nor the doctors predicted to recover who might have.

These are some of the more obvious challenges with AI: AI is a coded reality and so the biases of those doing the coding or the data they use will ultimately be crucial.

We had a fascinating evening recently at ANZ where we brought in five experts with five different perspectives on AI to talk to some customers and our board and executives.

They were Dr Elanor Huntington, an expert in quantum computing and Dean of Engineering at the National University of Australia; Daniel Petre, co-founder of Air Tree investors, former Microsoft executive and venture capitalist; and Greg Cross from Soul Machines, who is on a mission to humanise AI and has worked with us on Jamie, our new digital assistant in New Zealand as well as widely in Hollywood.

We also had our chief economist Richard Yetsenga and our head of digital Maile Carnegie.

It’s fair to say the evening canvassed the spectrum from Utopian to Dystopian visions of AI and while I’m not concerned by a Terminator-style takeover by robots, it is clear this latest phase of the digital revolution is at least as seismic as the agricultural and industrial revolutions.

It will change the way we work, live and think. AI will introduce some extraordinarily complex challenges. Think about how we might program a self-driving car. Elanor made the point we don’t even have legal consistency around liability across borders. Who is responsible? The car manufacturer? The software designer? The network? The passenger/owner?

How do we program such a car to react if it faces the horrible dilemma of either hitting a group of pedestrians or crashing and probably fatally injuring its owner?

Elanor added in some cases AI might make a decision but we could never understand why – what was behind its “thinking”?

Yet the potential for improving the lives of people and communities is truly immense. Having AI rapidly and accurately perform repetitive tasks or research vast amounts of data frees up medical specialists or lawyers or teachers to concentrate on where they can be most valuable as humans.

There is of course the concern robots will take our jobs. Daniel reminded us we had 50 years for the labour market to respond to the industrial revolution. This one will be on us much more quickly – the challenge of speed, Maile reminded us, will be critical for companies to cope with.

I know there will be disruption but I am actually less concerned about robots taking our jobs because when you look more deeply at the implications of AI, it is not about robots taking over but about how humans and robots work together, in new ways.

The final question of the night came from one of our customers who said he sometimes went fishing with his banker, they swapped salami and homemade wine, the relationship was deeper than business.

“No robot can replace that,” he said, to much applause. He’s right. And we wouldn’t want one to – but we would want his banker to have much more insight, based on vastly better data analysis and deeper understanding, driven by AI, to share.

After all, while I believe AI will play a more and more significant role in not just our business but our lives, I can’t imagine a robot ever being able to write a satire like “Yes, Minister”

If you’d like to hear some of the key insights from our discussion, here’s the link: