
Concerns: Professor Toby Walsh explores the existential risks posed by our accelerating efforts to realise an AI-powered future in God AI: Boom or Doom? What to Expect When Machines Outsmart Us. Photo: Supplied.
“Success in creating AI would be the biggest event in human history. Unfortunately, it might also be the last.” Physicist Stephen Hawking’s 2014 quote delivers the ultimate catch-22: while advanced artificial intelligence could fix our biggest global problems, it can also outgrow and overwrite us.
We share a perceptible dread when it comes to artificial general intelligence. Is anthropogenic collapse inevitable? One of the world’s leading AI researchers, Toby Walsh, explores the existential risks posed by our accelerating efforts to realise an AI-powered future in God AI: Boom or Doom? What to Expect When Machines Outsmart Us.
The professor of artificial intelligence at the University of NSW and chief scientist at its AI institute, UNSW.ai, admits God AI is not a book he wanted to write or wanted us to read. However, he notes it is the most important book he’s written.
“For the last 50 years, I’ve had a dream,” Professor Walsh writes. “It was a simple dream: I wanted to build artificial intelligence. I thought this was a good dream, since AI would help make the world a better place. And, in many ways, AI already has made the world a better place.”
He then goes on to outline AI’s benefits, including early disease detection in healthcare, and an AI tutor that provides personalised support in maths, science and humanities. Hard not to admire an AI like this!
“This is the sort of AI-powered future I had dreamt about for 50 years, ever since I was a young boy reading too much science fiction. It’s a future in which AI solves many of humanity’s problems. And it’s these AI dreams that have driven my passion for our AI future ever since. But recently something had changed. There is now increasing fear about our AI-powered future.”
While it is impossible not to be impressed by this kind of AI, fears also exist. AI is creating incredible tools such as new medicines, but it is also changing our jobs and relationships so fast that it is disrupting millions of lives. Advanced AI could also destroy humanity.
Professor Walsh looks at early forecasts. Alan Turing and Irving Good were two brilliant British mathematicians and cryptanalysts, who played vital roles in breaking secret German codes during World War II and shaping early computer science and AI.
In 1950, Turing wrote what is considered the first scientific paper on AI. A year later, he made a prediction on a BBC broadcast that continues to haunt the field today. Once machines could learn from experience like humans, Turing said, they “would not take long to outstrip our feeble powers … At some stage therefore we should have to expect the machines to take control”.
Good also made a prediction about how machines could take control, calling it the “intelligence explosion”. AI is being hailed as the ultimate economic game-changer of our era.
The question is not only how we got to AI doom, but what can be done about it. Professor Walsh presents facts, figures and visionaries with clear ideas about what the future should look like.
In his capacity as a laureate fellow, he raises concerns about losing control, the need for caution, how essential it is to learn how to play it safe and to manage the dangers.
Deft and insightful, God AI provides a level-headed overview of the ongoing AI controversy. Walsh outlines a future where change is the only guarantee. The idea that AI will reflect our own flaws and virtues is equal parts comforting and alarming.
“To understand what has changed — and what has made me, a long-time AI researcher, step back and question what I have been working towards for my whole life, you need to appreciate three things: a recurring pattern, a troubling paradox and a prior prediction.”
Over the past few years, he started to observe something odd. Many people he knew, who had also dedicated their lives to advancing AI, were becoming afraid.
He also points out that our problems aren’t only AI. It’s also human intelligence. “How good is our capacity to recognise and respond to existential risks before they become existential realities? Is humanity wise enough to respond in time?” These critical inquiries call for substantive responses. We face pressing concerns that must be addressed fully.
Apprehension towards artificial general intelligence cuts to our core, striking at the very roots of our consciousness. Crucial topics necessitate candid dialogue and meticulous scrutiny, before being executed with rigour.
God AI: Boom or Doom? What to Expect When Machines Outsmart Us, by Toby Walsh, La Trobe University Press, $36.99


