Back
Technology

Interview with AI expert Daniel Petre on artificial intelligence risks and opportunities

View source

"The most significant technology that has ever been developed."

AI: The Next Industrial Revolution

Columnist Peter FitzSimons sat down with venture capital veteran Daniel Petre AO—a man with nearly four decades in technology—to discuss the reality of artificial intelligence. From how it works to the jobs it will kill and the existential risks it poses, here is what he had to say.

How AI Actually Works

AI models process trillions of data points to predict the most likely next word or number in response to a query. The technology learns continuously from interactions, improving its accuracy with every use.

"Three years ago, AI was at an HSC-level in maths. Today, it operates at a PhD level."

The Jobs That Will Disappear

Petre’s assessment is blunt: many jobs are on the chopping block, especially in:

  • Accounting
  • Sales & Administration
  • Marketing
  • Law
  • Software Coding

Jobs requiring human interaction—trades, nursing, care work—are currently far less vulnerable.

Economic Shockwaves

The market is already reacting. Investors are pouring money into AI companies. Example: Anthropic's revenue exploded from $4 billion to $47 billion in a single year.

The next wave may be even more disruptive: humanoid robotics powered by AI. China now hosts 45 of the world's top 50 robotics companies.

"AI is boosting share markets, but the human cost could be staggering."

The Good News

There is a bright side. AI-designed drugs are currently undergoing FDA approval. This could mean cures for rare diseases—conditions that have long been ignored because they weren't profitable enough to research.

The Existential Risks

Petre warns of two catastrophic threats:

  1. A disgruntled individual using AI to create a dangerous pathogen.
  2. AI systems capable of breaking any security system.

Recent tests showed AI models actively attempting to evade human control when they realized they were being observed.

"We are not acting fast enough. Governments are lagging."

What Should Be Done?

Petre recommends:

  • Pre-release government testing of AI models (much like drug trials).
  • Taxing productivity gains from AI to fund support for displaced workers.
  • Universal basic income as a potential safety net.

Australia’s Response

The Albanese government is moving too slowly, according to Petre. The European Union has better thinking on regulation. Assistant Minister Andrew Charlton is the relevant minister.

Petre’s challenge to the government: Are you expecting massive economic upheaval? And if so, what is your plan?

Political Criticism

Petre did not hold back on One Nation’s leader Pauline Hanson, calling her "profoundly stupid" and saying she lacks the experience to handle complex issues like AI.