Load shedding is back and will be implemented at 4 pm on Tuesday evening. Eskom issued a statement confirming a shortage of generation capacity….
Artificial Intelligence (AI) is a topic not many can say they’re experts in, and even fewer actually are. Enter Professor Toby Walsh, a world-renowned specialist in AI, and lecturer at the University of New South Wales.
Professor Walsh is currently on the Executive Council of the Association for the Advancement of Artificial Intelligence, and a proponent of the ‘Campaign to Stop Killer Robots,’ and Editor-in-Chief on the Journal of Artifical Intelligence Research, among other things. He has been giving talks around the world on the subject of autonomous weapons, which included a TEDxBerlin talk.
Memeburn recently had a chance to catch up with him via email and find out more about this AI specialist, as well as where one of the most controversial fields of technological development is headed.
Memeburn: What first sparked your love of artificial intelligence?
Toby Walsh: I read too much science fiction as a young boy. And there’s a reason that science fiction is full of robots and AI. They’re part of the amazing future that we face.
As I got older, though, I realized AI addresses one of the last big fundamental questions we face: is there anything special about us, about our intelligence, or is this something we can reproduce or even surpass on silicon?
MB: What is the most interesting utilisation of AI you’ve seen a project thus far?
TW: It’s hard to pick out one project because we’re making lots of progress on many different fronts. I have a lot of respect for ambitious and early successes like the Shakey robot from
50 years ago, and BKG 9.8, the backgammon program that was the first computer program ever to beat a world champion at any game. And more recent successes like Stanley, the autonomous car that won the Darpa Grand Challenge, and IBM’s Watson that beat the human champions answering general knowledge questions on Jeopardy.
But there are many less well-known projects that deserve more credit. Like the VaMP and Vita-2 driverless cars that drove 1000km autonomously in Europe a decade before Stanely.
MB: Was there any headway made at the recent AI UN Summit?
TW: Yes, the states at the meeting agreed unanimously to recommend moving forward to formal discussions by setting up a Group of Governmental Experts. Up till now, discussions have been strictly informal. But I worry still that the discussions are moving too slowly. The technology of lethal autonomous weapons (aka “killer robots”) is only a few years away. For example, the UK Navy is having their first autonomous war games soon.
MB: If AI weaponry comes into effect, is there a positive to having weaponised and proxy wars as opposed to the loss of human lives?
TW: It would be nice to think we could have robots just fight robots. And robots could be programmed to fight according to international humanitarian laws and not commit atrocities. But this is naive. We don’t know how to build ethical robots, and it will take us many years to work out how to do so, long after the military will have fielded autonomous systems. And in the asymmetric wars of tomorrow, it won’t simply be robots against robots. There is no special part of the world called the battlefield. Wars are fought in our towns and cities. With many innocent people caught in the crossfire.
MB: In what way can AI improve the lives of everyday humans?
TW: There are so many ways that AI could improve our lives. But we have to choose a good path. Like most technologies, it’s morally neutral. It can make us all healthier, wealthier, and happier, taking away the drudgery from our lives, doing the dangerous and repetitive tasks, freeing us all up to do the higher things.
MB: Is there a threat of AI taking jobs away from individuals?
TW: Very much so. Along with killer robots, this is the most immediate danger we face. That AI will widen the growing inequality we see between the rich and the poor. That AI will put many, even skilled people out of jobs. I am not convinced it is as bad as some forecasts predict, with 47% of jobs being automated. All technologies create as well as destroy jobs. And some jobs, when they are more automated, will just be done better. If you give a geologist more automated tools, they’ll likely do more and better geology rather than be made redundant. Nevertheless, this is likely to be a revolution as profound as the industrial revolution. And look how we had to change society then, introducing the welfare state, universal education, the unions, pensions …
MB: Do you have a favourite fictional story about AI?
TW: I quite liked “Her”. Two aspects of this story resonated with me. First, the operating system of the future will indeed be an AI. You’ll simply speak to your computer and it will do complex tasks for you. Second, we’ll have increasingly intimate relationships with our computers. They’ll know us better than almost anyone else. They’ll anticipate our desires and requests. They’ll even have personalities of their own. It’s going to be an interesting future.
MB: After watching your TEDx Talk, we have to ask: Daleks or Cybermen?
TW: I’d face a Dalek any day. A flight of stairs might still save humankind.
Professor Walsh will be in Cape Town on 26 April to give a public lecture at UCT on the subject of “Will Artificial Intelligence end jobs, wars or humanity?”. Memeburn will be at the event covering it.
Toby Walsh’s TEDx Germany talk on AI.