Toby Walsh, Scientia Professor of Artificial Intelligence at the University of New South Wales in Sydney, Australia, is a world leader in the field of artificial intelligence. He is the driving force behind an open letter calling for a ban on offensive autonomous weapons that has been signed by more than 20,000 people and discussed at the United Nations. His new book, 2062: The World That AI Made, is a call for thoughtful decision-making to start right now, he tells Michael Brooks
How did you get into studying artificial intelligence?
My story is really a proof of the wonders of education as a way of helping people to discover themselves. I’m not from an academic family – in fact, my father left home without any qualifications at all. I was simply a bookworm who loved science fiction, and I decided I wanted to study artificial intelligence. But I also wanted to go to Cambridge, and you couldn’t even study computing there at the time. So I did mathematics and theoretical physics, and then went to Edinburgh to join the world’s first department of artificial intelligence.
Why does your new book pick out 2062 as a pivotal year?
For a couple of reasons. One is that I surveyed 300 of my colleagues about when machines would be as smart as humans. 2062 is the median – the midpoint – of the dates they gave. But that was cemented as the title when I realised that it’s also the year when my daughter will be the same age as I am now. It’s not really a book about my future, it’s more a book about her future – about making decisions now that will determine all our children’s futures. People make the mistake of thinking the future is fixed. It’s not: it’s primarily the result of decisions we make about technologies. Perhaps if we had decided that TV should not be used for political purposes, for instance, we wouldn’t have a US president who’s a reality TV star.
Has campaigning against killer robots always been part of the plan?
Not really – I’m very much an accidental activist. As a scientist funded on the public purse, I do feel a strong responsibility to inform the public conversation and make sure people are making good decisions and aware of any risks. The campaigning started when some people asked me if I could I round up some signatures from colleagues. We’ve now got somewhere around 3,000 academics on board, and I’ve found myself talking four times at the UN, which is a really strange place for a scientist to find themselves. It’s very reassuring to see the seriousness with which these things get taken there, however.
Do you think the campaign will succeed?
I’m very optimistic that we will ban such weapons because they will ultimately be recognised as weapons of mass destruction. After all, one person can program a thousand killer robots. We’ve already worked out that all other weapons of mass destruction should be banned, and I think it applies to weaponised AI. My only concern is that we’ll have to see the horror of them being used before we have the courage and conviction to do that.
Do you worry that AI has become such a fashionable topic?
The truth is, AI is not going to live up to the hype. It’s hard to imagine that we can meet all of the expectations placed on it. But even if we were to stop making any progress at all in the science, we have already made a practical difference to people’s lives, businesses and healthcare. We’re starting to see AI reading mammograms, and diagnosing pulmonary disease more cheaply, far quicker and far more accurately than humans. And where would we be without SatNav? The navigation algorithm in your smartphone started life in the 1960s as an AI route planner for Shakey, the first fully autonomous robot.
What do you do to relax?
I come from a sailing family, and I try to sail when I can. I even sailed across the Atlantic when I was younger. Moving with just the wind in your sail, the same way people have travelled for centuries, is just a wonderful way to get your mind to switch off.
Images courtesy of Toby Walsh