Sean Smith: Businesses must prove artificial intelligence’s benefits rather than all-or-nothing approach

Gee whiz, how good is artificial intelligence? Heck, it can do almost anything — drive scientific advances, deliver better healthcare, write jokes, mitigate climate change, plan holidays or write an assignment. World peace and a winning West Coast Eagles team surely must also be within its power. AI is already part of our everyday lives; it has been for years. But businesses are now falling over themselves to expand its uses. It may well be that arguably the world’s biggest technological changes since the Industrial Revolution will overhaul working practices for the better, improve living standards and usher in a new era of prosperity, just as its advocates say. However, learned experience dictates that AI’s rising adoption be accompanied with caution, some more guardrails and not a little cynicism. Australia needs to do everything it can to ensure AI has national benefits, that it doesn’t just enrich a few at the expense of entire communities, that it doesn’t create more inequality. It is a rapidly evolving, increasingly intelligent and sophisticated technology, parts of which are still not understood. Also, ethical, privacy and transparency concerns abound and there are fears about developing an over-reliance on AI systems. That alone should be enough to not rush into every AI initiative that comes around the corner. However, that’s not the human way. It’s now a race to see how far AI can be developed by businesses, despite reservations by workers and customers. Want to deal with a real person on a customer service inquiry? Sorry, here’s a chatbot. The technology is already reducing critical thinking as humans entrust it with tasks that, the argument goes, will enable them to concentrate on other, more meaningful work, at least until the technology also takes that on. And the biggest supporters are naturally those conflicted by self-interest, from the LinkedIn spruikers to the growing number of companies that see it as a means of slashing costs, usually by cutting jobs, while spinning the line that employees will “partner” with AI. Employers are failing in their obligations to explain to increasingly nervous workers how AI can benefit them, or their children. Instead they are just imposing the technology on their workforces. Australian Securities and Investments Commission boss Joe Longo nailed it during a speech to a banking audience last week. “Trust is what seems to be missing,” he said, citing research by RepTrak and KPMG that shows that trust in AI over the past 12 months has declined. “Australians have lower levels of AI use, training and confidence than their peers around the world. Five out of 10 Australians use AI — but fewer than four out of 10 trust it,” Mr Longo said. “So the perception is a bit of a mixed bag. On the one hand, we have business leaders across a range of industries singing its praises, while on the other hand, the number of Australians who distrust the technology appears to be growing.” The contradiction was on full show this week when Scott Farquhar, co-boss of software giant Atlassian and chair of the Tech Council of Australia, addressed the National Press Club in Canberra. He laid out a vision where AI could help solve the housing crisis by taking on the burden of building approvals and play a meaningful role in improving Australia’s weak productivity by generating economic benefits across the economy, thanks to gains in business, health, finance and government. The billionaire admitted fewer workers will be needed in this new AI era but new, better-paying jobs will be needed to develop and support the software and build the huge, energy and water-hungry data centres essential to driving AI. As it happens, Atlassian that very morning had just laid off 150 workers, notifying them by a pre-recorded video. “If you make call centres twice as productive, then you need half as many people,” Mr Farquhar explained. The challenge for companies is to harness AI and the technology better, not just to its workers, but to their customers. If the messaging is that AI will improve service levels, then they need to deliver. Show them how it will make their lives better.