Follow us on our channels:
Artificial intelligence (AI) is not just the next big thing; it is the current big thing. At this year’s IT Nation Connect ANZ event, Australian mathematician, broadcaster and keynote speaker Adam Spencer returned to the Business Focus podcast to share his take on how AI has shifted over the past year and where it is heading next.
From the release of GPT-5 to the possibility of artificial superintelligence, Adam offered a grounded, witty and thought-provoking perspective. Here is a deep dive into the key takeaways for Australian businesses and tech leaders.
Adam kicked off with a reminder of how fast the AI landscape is moving. When he first started keynoting about AI back in 2016, long before ChatGPT became a household name, many people could not see how neural networks would impact their work. Fast forward to March 2023, when GPT-4 arrived, and everything changed. The leap in benchmark performance turned heads around the globe and ignited mainstream excitement.
Since then, dozens of models have launched. While many expected GPT-5 to be a similar game changer, Adam says it is “somewhere between underwhelming and okay.” Instead of a giant leap forward, GPT-5 has raised the floor more than pushed up the ceiling, focusing on reliability and reducing hallucinations rather than blowing past previous capabilities.
To illustrate the speed of change, Adam shared a personal experiment: using GPT-5 to build a simple maths quiz app. Within five minutes, and with only a few text prompts, the AI generated 300 lines of working code. Ten years ago, that would have earned a first-year computer science student an A.
It is not just coding. AI tools like Soundraw now let anyone create music in minutes, while image and video generators are producing fully AI-made ads for major sporting events. Australian brands are already experimenting with campaigns that proudly highlight their AI origins.
One of the most fascinating and daunting topics Adam explored is the race to artificial general intelligence (AGI) and ultimately artificial superintelligence.
Some experts predict AGI could arrive by 2030 and superintelligence by 2040 or 2050, decades sooner than early estimates. Companies like Meta are investing heavily, reportedly offering tens of millions of dollars to attract top AI researchers.
Adam points out that being even six months ahead in this race could give a company or nation a near-unassailable advantage. Unlike pharmaceuticals, where a new drug does not destroy a competitor’s market overnight, achieving superintelligence first could be game changing.
AI’s influence is transforming workplaces and challenging traditional career paths. Adam highlighted several key areas:
For technical staff, AI is freeing them from repetitive tasks like basic troubleshooting, allowing them to focus on more complex and rewarding problems. Entry-level roles may evolve to become stepping stones into tech careers, as non-technical employees gain confidence and skills through AI-assisted work.
Education is both benefiting from and being challenged by AI. Teachers are using AI to prepare lesson plans faster, and students who use AI to complement their learning can achieve strong results. But some rely on ChatGPT to write essays at the last minute, often producing work that lacks depth and creativity.
Adam believes assessment methods must adapt. Instead of standardised exams, he sees a future where assessments are more personalised, similar to an oral exam where the teacher continues asking questions until they find the limits of a student’s knowledge.
However, an MIT study found that essays produced with AI tended to be more homogeneous, suggesting a risk of creativity filtering if students rely too heavily on these tools.
AI is only as good as the instructions it receives. Adam stressed that effective prompting is a skill in itself, and it is not necessarily technical.
His tip is to create AI personas. By giving the AI a detailed character brief, for example “Gary the Research Assistant” or “Jamila the Critical Business Analyst,” you can get more tailored and reliable responses.
He uses personas to:
This approach shows that AI is not just about automation. It is about creating the right context to get the most valuable output.
Perhaps the most futuristic trend Adam discussed is the creation of digital twins, AI-powered versions of real people.
Futurist Mark van Rijmenam, for example, has built a digital version of himself using five million words of his own content. Subscribers can interact with his virtual self, asking questions and receiving answers in multiple languages that mirror his own thinking.
Adam predicts we will soon see virtual CEOs, politicians and thought leaders. Employees could use these digital twins to workshop ideas before pitching them to the real decision-makers, making organisations more efficient.
All this progress depends on data. AI models are only as strong as the information they are trained on. As publicly available data becomes saturated, people are beginning to hoard unique, unmined datasets such as books, archives and content untouched by AI training. This “pre-AI data” may become the next big commodity, essential for training future generations of models.
For Australian organisations, Adam’s insights are both inspiring and cautionary:
AI’s rapid evolution is impossible to ignore. From GPT-4’s breakthrough to the early days of GPT-5, and from digital twins to the race toward superintelligence, the next decade promises even more disruption and opportunity.
As Adam puts it, we are entering a world where everyone from lawyers to filmmakers, educators to MSPs, will need to adapt. Whether you are leading a business or just curious about the future, one thing is clear: the AI revolution is well and truly underway in Australia, and the time to prepare is now.
Want to keep up with the latest in AI and business strategy? Subscribe to the Business Focus podcast for more conversations with industry experts like Adam Spencer.
“`