What really matters for material living standards in the long run is productivity, the economic product of labor and capital. It’s easy to see why people think AI will one day improve productivity, but it’s equally clear that we’re not there yet.
Loading
Reserve Bank recent articles breaking news provides compelling evidence to suggest that many Australian businesses are at the stage of considering how to leverage AI, rather than executing detailed plans.
This article, by economists Joel Fernando, Kate McLoughlin and Ravi Ratnayake, is based on insights from a survey of more than 100 medium-sized to large companies that the RBA spoke to in its liaison programme.
They found that while companies expect their investments in AI to be transformative in the future, many companies are still in the early or experimental stages of AI adoption today.
Nearly 40% of companies said they had “minimally” deployed AI tools, usually in the form of “digital assistants” such as Microsoft Copilot or ChatGPT. These are AI tools (often of questionable fact) that can summarize emails or assist with research, but they don’t seem like they’ll make a big difference in productivity.
The RBA study found that Australian companies are adopting AI in “gradual” ways. This often included staff using ChatGPT.credit: Reuters
While some companies surveyed said they were using AI more frequently, these were generally larger companies, and economists said most companies had “a shallow history of AI adoption.”
RBA economists said: “Overall, many of the companies surveyed noted that the adoption of AI tools to date has been relatively gradual, with adoption often being employee-led rather than employer-led.”
But what about concerns about an AI-driven “jobpocalypse”?
The RBA’s data does not reflect the entire labor market, and of course the impact of AI will vary by occupation. Still, research and other evidence suggests that AI is having a fairly modest impact on staffing so far. Companies surveyed said that “most” workers displaced by technology (including AI) moved on to other jobs with the same employer, but companies expected an even more “disruptive” impact from AI in the future.
Roughly half of the companies surveyed expect AI to “slightly” reduce their total workforce over the next three years, due to factors such as natural attrition, fewer new hires, redundancies, or a combination of all three.
Westpac’s chief economist has been asked more about AI than his views on interest rates.credit: Lenny Nowitager
Many workers, especially in industries like my industry and the media, are rightly concerned about the risks of AI to their jobs, and the findings of this study do not alleviate those concerns. But it suggests that dramatic changes are still a long way off.
And for a company, it’s one thing to say you plan to reduce costs thanks to new technology; it’s another to do so without creating gaps in your operations. The Commonwealth Bank infamously announced earlier this year that it would cut 45 jobs as their roles were replaced by AI-powered chatbots, which later backfired.
From a broader economic perspective, major technological changes in the past have created new jobs and businesses, while also eliminating some. It’s very possible that this will happen with AI-related changes as well, but no one knows for sure and it will likely be debated for years to come.
And this is another big theme in the RBA research. Businesses are highly uncertain about potential use cases for AI, availability of sufficient skilled staff, potential costs, how it will be regulated, and more.
Could AI reverse the long recession and usher in a new era of rapidly increasing productivity? Perhaps, but likely not soon.
This uncertainty is important because if AI is truly to transform the economy, companies will need to make major changes to the way they operate.
Those who believe AI is the next big thing compare it to some of the biggest technological changes in modern history, such as the invention of the steam engine, electrification, the rise of the automobile and the internet.
Perhaps they’re right – but economists also know that these changes didn’t happen immediately.
Loading
For example, the transition to electricity required not only infrastructure such as power lines, but also a thorough renovation of factories.
Or, in a more recent example, it took many years for the productivity gains of computers to become apparent. Nobel Prize-winning economist Robert Solow famously said this in 1987. “The computer age is everywhere except in productivity statistics.”
To significantly improve productivity with AI, companies need to rethink all processes and staffing to see how AI can transform their business. It can be time consuming and expensive, requiring significant retraining of staff and possibly hiring more people.
Although it’s still early days, it’s clear that there’s a lot of interest in AI in the business world.
The fact that bosses at large companies are constantly communicating their AI plans to the market shows that many companies are actually planning how to get the most out of the technology.
Westpac chief economist Lucy Ellis says the companies she speaks to are more than “experimental”, and in fact she is asked more often about AI than her views on interest rates. Ellis believes the productivity benefits of AI will emerge faster than it took computers and the internet to show up in productivity statistics, but it will still take time. “We’re talking about two or three years, not three weeks,” she says.
Could AI reverse the long recession and usher in a new era of rapidly increasing productivity? Perhaps, but likely not soon.
One thing we’re more certain about is that ChatGPT won’t skyrocket your productivity by summarizing your emails. The changes will be much more profound than that.
