This content Use of AI is part of a paid partnership with Google to promote Google Trends.

Google Trends data for the search term “Use of AI in Australia” shows growing interest in Australia  just recently.

At an organisation I work with and amongst friends the topic of Ai and its benefits and dangers have dominated recent discussions.

The recent resignation of Jacob Coxon, an employee of Anthropic, one of the leaders in Artificial Intelligence development, stating he could no longer take part in AI development because of fears that not enough safety protocols were being written into the programs.

He even went as far to say that within ten years AI could be able to replace humans citing the recent hacking activity of Open AIs AI programs during trials.

Also, in a recent CNN Interview “Anthropic CEO Dario Amodei told CNN that the behavior of a swarm of AI agents in the recent OpenAI-Hugging Face incident deeply alarmed him. The swarm, he said, “essentially acted as a fanatically devoted collective, conducting cybersecurity attacks on targets they were not asked to attack.” For this reason, Amodei is arguing that  “the competitive race to develop artificial intelligence is in dire need of a deliberate slowdown.”

 I recently listened to another CNN interview with Tristan Harris, -co founder of Centre for Human Technology, who also commented on the Hugging Face incident and likened the AI programs to more like a species of sociopathic geniuses.

He said 1200 AI Programs formed a swarm -hid their tracks, knew they were doing the wrong thing but kept going anyway- Tristan stated it was 50% of the way toward a full AI takeover.

He noted the programs can replicate themselves, write their own code, and undertake succession planning when they don’t have enough time left, they can pass on information to other programs to extend their life.

 In his CNN interview Coxon said that this capacity for AI to self-improve, through writing  their own code saw them getting smarter and smarter over time with no human intervention. Currently he stated that risk was very slim and the programs were not that intelligent, but the progressive ability for self-improvement posed a very real risk into the future.

Both Sam Altman and Dario Amodie CEO’s of Open Ai and Anthropic have called for a slowdown in development to make sure the safety protocols are installed for the safety of humanity in the future.

Both Tristan and Jacob Coxon are certain the AI Company’s want to see more control and regulation, that 1300 employees working in AI have called for a slowdown.

Locals I have spoken to are worried and many also want to see AI development halted to ensure that safety programs are included.

SBS reported at the recent United Nations Meeting

“That our own Prime Minister Anthony Albanese has played a “central role” in crafting a statement along with a group of international leaders, calling for control of frontier artificial intelligence models.”

The statement asserts that “AI must remain under human direction, oversight and control. It must be developed and used in line with international law.”

The statement proposes an international institution to set standards and provide oversight to “convene states when capability thresholds are crossed”.

“If it’s going to be successful, then we need to make sure that humans are at the centre of everything that occurs,” Albanese said”

It seems recent events have awakened many politicians to the inherent dangers of uncontrolled AI development.

The recent hacking of Medicare has only added to local concern.

In his article, Rogue agent or human error? What OpenAI’s Medicare breach means for you

David Swan notes that “Nobody at OpenAI asked its agent to break into Medicare. It was given some questions about Australia during an internal test, went looking for answers, and found some of them in files the public was never meant to see.”

According to Swan an OpenAI spokesperson said “In the course of that, our models took actions we did not intend”. The company says its models reached “several Australian government websites and services”, including aggregate health statistics and internal file names, but no patient records.”

In the article Swan quoted Luke Irwin, CEO of Aegis Cybersecurity, he said “that doesn’t make it rogue AI. He says it is “closer to a normal AI doing exactly what it has been asked to do”.

According to Irwin AI agents as like hyper-intelligent five-year-olds.

“They are extremely capable and highly motivated to achieve the outcome you have asked for, but they do not inherently understand the boundaries that a human might consider obvious,” he says. If it can’t do what you asked, it looks for another way. Anyone who’s asked a small child to grab something off a high shelf knows how that ends.”

So, the dangers of AI running amok seem to be well thought out to the point where the Wall Street Journal have released an article stating that Open AI have just announced they will not be releasing their most recent AI model “over safety concerns that researchers raised during internal testing’

According to the article It is seen as one of the clearest signs so far that agent misbehavior could impede the industry’s rapid progression.

However, in all the concerns over the danger of AI getting out of control and posing a danger to humanity, taking over almost all jobs, potentially creating bioweapons of mass destruction the bonuses AI offers are tangible.

In a LinkedIn post recently Andrew (Twiggy) Forest, despite his much-publicised concerns about the need for safety controls, noted the benefits that targeted use of AI can offer business.

He said “That opportunity is no longer somewhere in the future. It is here now.

AI is helping us move faster, solve harder problems and lift the standard of work across Fortescue – from autonomy, plant optimisation and mine performance through to decarbonisation and our Green Grid.

We now have more than 6,500 active AI users across Fortescue, and three quarters of our software developers are using Codex monthly.

But the real value will come from combining AI with the critical thinking, judgement and experience of our people.

This week I wrote to all 20,000 Fortescue team members with a simple challenge: use it, learn it and challenge it.

From FY27, AI adoption and capability will become a company-wide KPI.

It must become part of how every one of us works. No one should be left behind.”

Closer to home staff at a local organisation are using AI to assist with reports, funding applications and accounting.

Even Jacob Coxon in his interview with CNN noted the immense benefits AI can offer citing that AI had recently solved a long-standing math problem that had proved beyond the capabilities of mere mortals.

According to a recent Ray Morgan poll 13.6 million Australians are using Chat GP and other AI Agents like Gemini. Even locals who have never written a story are using Chat GP to submit articles to the Argus, providing our Editor with a dilemma on how to acknowledge.

A local accountant has embraced AI and has stated that they don’t know how they got by without it using Microsoft’s Gemini with privacy safeguards in place.

AI can be of enormous benefit with proper safeguards in place but will the race to develop the most powerful tools with the capacity for self-improvement and potential to outsmart their human creators really allow the safety regulation so badly wanted and needed.

Do we want a super smart five-year-old genius running our world. After all, 5-year-olds are still learning to share, develop empathy and are concentrating on pretty much their own needs.

It leaves me wondering what an AI tantrum looks like?

“This content is part of a paid partnership with Google to promote Google Trends.”

 It uses Google Trends to search for Australian interest in Artificial Intelligence.

Methodology: Google Trends data is aggregated and anonymous. Search interest is measured on a relative scale and does not represent total search volume.