The government says more people need to use AI. Here’s why that’s wrong | UniSC | University of the Sunshine Coast, Queensland, Australia

But why exactly do people need to trust this technology? And why exactly do more people need to use it?

AI systems are trained on incomprehensibly large data sets using advanced mathematics most people don’t understand. They produce results we have no way of verifying. Even flagship, state-of-the-art systems produce output riddled with errors.

ChatGPT appears to be growing less accurate over time. Even at its best it can’t tell you what letters are in the word “strawberry”. Meanwhile, Google’s Gemini chatbot has recommended putting glue on pizza, among other comical failures.

Given all this, public distrust of AI seems entirely reasonable. The case for using more of it seems quite weak – and also potentially dangerous.

AI risks

Much has been made of the “existential threat” of AI, and how it will lead to job losses. The harms AI presents range from the overt – such as autonomous vehicles that hit pedestrians – to the more subtle, such as AI recruitment systems that demonstrate bias against women or AI legal system tools with a bias against people of colour.

Other harms include fraud from deepfakes of coworkers and of loved ones.

Never mind that the federal government’s own recent reporting showed humans are more effective, efficient and productive than AI.

But if all you have is a hammer, everything looks like a nail.

Technology adoption still falls into this familiar trope. AI is not always the best tool for the job. But when faced with an exciting new technology, we often use it without considering if we should.

Instead of encouraging more people to use AI, we should all learn what is a good, and not good, use of AI.

Is it the technology we need to trust – or the government?

Just what does the Australian government get from more people using AI?

One of the largest risks is the leaking of private data. These tools are collecting our private information, our intellectual property and our thoughts on a scale we have never before seen.

Much of this data, in the case of ChatGPT, Google Gemini, Otter.ai and other AI models, is not processed onshore in Australia.

These companies preach transparency, privacy and security. But it is often hard to uncover if your data is used for training their newer models, how they secure it, or what other organisations or governments have access to that data.

Recently, federal Minister for Government Services, Bill Shorten, presented the government’s proposed Trust Exchange program, which raised concerns about the collection of even more data about Australian citizens. In his speech to the National Press Club, Shorten openly noted the support from large technology companies, including Google.

If data about Australians was to be collated across different technology platforms, including AI, we could see widespread mass surveillance.

But even more worryingly, we have observed the power of technology to influence politics and behaviour.

Automation bias is the terminology we use for the tendency for users to believe the technology is “smarter” then they are. Too much trust in AI poses even more risk to Australians – by encouraging more use of technology without adequate education, we could be subjecting our population to a comprehensive system of automated surveillance and control.

And although you might be able to escape this system, it would undermine social trust and cohesion and influence people without them knowing.

These factors are even more reason to regulate the use of AI, as the Australian government is now looking to do. But doing so does not have to be accompanied by a forceful encouragement to also use it.

Let’s dial down the blind hype

The topic of AI regulation is important.

The International Organisation for Standardisation has established a standard on the use and management of AI systems. Its implementation in Australia would lead to better, more well-reasoned and regulated use of AI.

This standard and others are the foundation of the government’s proposed Voluntary AI Safety standard.

What was problematic in this week’s announcement from the federal government was not the call for greater regulation, but the blind hyping of AI use.

Let’s focus on protecting Australians – not on mandating their need to use, and trust, AI.

Dr Erica Mealy, Lecturer in Computer Science, University of the Sunshine Coast.

This article is republished from The Conversation under a Creative Commons license. Read the original article.