Economic Reform Roundtable: AI and Robotics for Future Growth

Published Date : 30/06/2025 

The upcoming economic reform round table should prioritize discussions on artificial intelligence and robotics, rather than getting bogged down in tax reform debates. The assembly should focus on developing policies that prepare Australia for the AI-driven future, including improving AI prompt writing and modernizing the welfare safety net. 

Let’s hope the economic reform round table in August does not turn into a three-day argument about tax reform. Nobody can ever agree on whether a sustainable budget is best achieved by less spending or more tax, let alone what taxes should be increased or new ones imposed.

The 25 citizens in the assembly would be better off trying to come up with a decent set of policies on artificial intelligence and robots. They could start with a briefing about what US President Donald Trump will reveal in his AI Action Plan to be announced on July 23, and follow up with a rundown on China’s eight-year-old strategy to become the global leader in AI by 2030.

As for the five reports on productivity’s five pillars commissioned from the Productivity Commission by Jim Chalmers in December, and the 453 ideas people have sent in, a better way for PC chair Danielle Wood to instil a sense of urgency into the round table might be to read out this article from the Financial Times last week about Meta Platforms Inc offering as much as $150 million sign-on bonuses for AI engineers working for the developer of ChatGPT, OpenAI.

Meta CEO Mark Zuckerberg is trying to build a team that will help him create AI that is smarter than human beings. And the staff at OpenAI have apparently been knocking him back! The combined market value of the five leading AI companies – Meta, Alphabet, Microsoft, Amazon, and Nvidia – is $20 trillion, almost twice the value of all houses and units in Australia.

The only way any of this makes sense is that Meta, OpenAI, and the other peddlers of AI are going to make more money from it than from any other technological revolution in history. And the only reason the rest of the business world will pay so much money for AI software and robots is because they will make even more money by replacing humans and increasing the productivity of those still employed.

There are two ways of applying AI in a business: top-down and bottom-up. At the risk of over-simplifying it (but not much), top-down involves a business’s executive leadership team deciding to use AI to replace a bunch of human workers by, say, getting it to write software code, or do data entry, answer phone calls, or write articles like this one.

Bottom-up is where individual employees use AI to become better at their jobs. The key to making the bottom-up use of AI work is writing a good prompt, which is rapidly becoming one of the more important modern workplace skills. An AI consulting business called Fourday is building a “prompt library” for clients and has developed a prompt writing assistant called Prompt Cowboy. You write into it a sort of lazy, plain language prompt and it spits out a much longer one that is better designed to get the best answer from the AI. ChatGPT performs better with a more detailed, better-written prompt.

Fourday founder Henry Badgery says it is possible to get productivity increases of 30-40 per cent with AI, but many businesses are reporting no extra value at all because: 'You're not using it correctly. You're not asking the right questions because that's essentially what a prompt is, you're asking it the right question, you're giving it the right information to answer your question.'

If the round table is going to be pragmatically useful in lifting productivity, it should do two things: first, develop a national system for helping small to medium enterprises and employees to write better AI prompts, and second, modernise the welfare safety net so people don't have to be scared of it, allowing companies to replace workers with AI without worrying about contributing to societal breakdown.

Specifically, the government needs to rethink unemployment benefits. The current JobSeeker system is not fit for purpose because it’s assumed to be a temporary payment for a job seeker. The money is not enough, and the points system that includes the need to apply for four jobs each month won’t work if there aren't any jobs to apply for.

A Universal Basic Income for everyone, as some are pushing, is going too far, but the government needs to prepare for a period of more permanent unemployment – even if it’s not needed in the end because AI does not cause mass unemployment. Business executives and staff need to embrace AI, and a good safety net is needed for that to happen. It’s another reason for more tax revenue, along with the need for more public housing. And more tax revenue comes down to a new tax – either a wealth or inheritance tax – or increasing the GST.

Australia’s two previous bouts of major tax reform when tax revenue was falling short of government spending also involved new taxes: capital gains and fringe benefits in 1985 and GST in 2000. Twenty-five years later revenue is short again because of the aging of the population, the NDIS, and the decision to increase defence spending.

But the near-hysterical response to the plan to reduce the tax break on superannuation balances above $3 million suggests that a new wealth or inheritance tax is too hard. Lifting the GST rate towards the global average of 15-20 per cent is what it will probably have to be.

The problem with GST is that it’s regressive – it hits the poor more than the well-off. But there is a solution to that, invented by UNSW Professors Richard Holden and Rosalind Dixon. In their book From Free to Fair Markets, Holden and Dixon suggest increasing the GST rate to 15 per cent and applying it to all goods and services, but only levying it on spending above a threshold of $12,000 per year.

How? By giving back each taxpayer 15 per cent of $12,000, or $1,800, each year, preferably in two six-monthly payments of $900 – simply put it in their bank accounts. Holden and Dixon have modelled the outcome and say the higher GST on a broader base would double GST revenue from $90 billion to $180 billion, and that the refunds would give back half that, leaving $45 billion in net extra revenue which they say could be spent on income tax cuts and eliminating the budget deficit.

Whether $12,000 ($1,000 per month) is the right threshold for GST-free spending could be debated, but the idea is worth spending an hour or two on at the round table. Then the 25 round tablers could get on with the real business of AI and robots. 

Frequently Asked Questions (FAQS): 

Q: What is the main focus of the economic reform round table?

A: The main focus should be on developing policies for artificial intelligence (AI) and robotics, rather than debating tax reforms.

Q: Why is AI and robotics important for the future of the Australian economy?

A: AI and robotics are expected to drive significant economic growth and productivity improvements by replacing human workers and enhancing the efficiency of those still employed.

Q: What are the two primary ways businesses can apply AI?

A: Businesses can apply AI in a top-down manner, where executives use AI to replace human workers, or in a bottom-up manner, where individual employees use AI to improve their job performance.

Q: How can the government help small to medium enterprises and employees in the AI era?

A: The government can develop a national system to help SMEs and employees write better AI prompts and modernize the welfare safety net to support those affected by AI-driven job displacement.

Q: What is the proposed solution to make the GST more equitable?

A: UNSW Professors Richard Holden and Rosalind Dixon propose increasing the GST rate to 15 per cent but only applying it to spending above a threshold of $12,000 per year, with refunds for the lower threshold. 

More Related Topics :