Illustration by Hope Mcconnell
It’s the US National Governors Association Meeting in July, 2017. Icy blue light washes the wall behind Elon Musk as he sits on a stage, garbed sleekly in a suit. Leaning forward with hands held before him, he assertively looks at a member of the questioning panel.
“You said that artificial intelligence is the fundamental existential risk facing civilisation?” his questioner asks.
“In my opinion, it is the biggest risk that we face as a civilisation.” Musk fidgets in his chair as he says this, aware of the weight of this statement.
He goes on to recount how an artificial intelligence (AI) programme recently beat the world’s top human player at the challenging game ‘Go’. “People thought a computer would never beat the best player or that it was 20 years away, and last year AlphaGo, which was done by DeepMind, [a] Google subsidiary, absolutely crushed the world’s best player,” recalls Musk.
“Now that it can crush the top player, it can play the top 50 simultaneously and crush them all.” He intently gazes at his questioner and no further explanation is demanded.
Aside from Musk, Stephen Hawking is one of many scientists who has voiced concerns about AI’s impact. With heaps of predictions that AI will become more intelligent than humans, calls for more thorough research have reached a roaring chorus.
AI and Us - What Does the Future Hold?
Some academics predict AI will soon surpass human intelligence and thus deem humanity to be doomed. Yikes, how’s that for cheery news? However, there are many experts who roll their eyes at this.
One of these challengers is Greg Cross, who works as the Chief Business Officer at Auckland-based artificial intelligence company Soul Machines. While talking with him, it became super clear that he doubts that it’s possible to predict our future with AI. “It’s nearly impossible to judge what society and humanity will look like when it occurs, because there will have been a world of change in all other aspects of society by that time,” Cross explains.
Others think it’s the way AI might be used that is the problem. Tom White, a Senior Lecturer in Media Design at Victoria University, reckons that, “The main existential threat to humans is not from software systems with suddenly emerging autonomy, but from existing institutions that use AI capabilities unethically.”
White points out how China is using AI to create a national surveillance system and also Facebook’s programming that allegedly induces social media addiction. Only a little bit shady, right? “The long term existential threat to humans from AI systems is an interesting philosophical debate. But focusing on this remote possibility is an irresponsible distraction in the face of more clear and present dangers to humanity enabled by current AI systems being used irresponsibly by people and organisations in the present and very near future,” he says.
So at the end of the day, there’s no point worrying about AI leading to the downfall of the human race when we’ve got more than enough on our plate right now.
The Job Factor
The most immediately concerning impact that AI will have on our society is probably the effect it could have on our job market. Ridiculous numbers of low-wage roles are being automated away. Self-service supermarket checkouts are just one pretty obvious example of automation, and it won’t be the only sector affected, according to Tom White. “Thanks to more powerful AI-based software systems, we will see similar shifts in the coming years across broad swaths of the labour market, such as taxi drivers and radiologists,” he predicts.
And it’s not just going to be blue collared jobs that are out of the game – for anyone in a white collar profession, this will have an impact on your industry, too. However, some optimists reckon new jobs will be created as others are lost, in effect replacing these roles.
Additionally, history has shown that new employment opportunities arise after technological revolutions. New Zealand Law Foundation Director in Emerging Technologies Colin Gavaghan points to the role of a blacksmith as an example. “When steam locomotives displaced horses, there was a lot less demand for blacksmiths. But new jobs were created for people who were good with metal and good with their hands,” he asserts.
But is it realistic, or just plain flakey to assume history will repeat itself? While new jobs may be created, Benjamin Liu, Senior Lecturer in Commercial Law at the University of Auckland, doesn’t believe the amount will be sufficient to match lost jobs, and upskilling may prove tough. “If a taxi driver loses his job to autonomous cars, it would probably take years for him to retrain for the new positions created by AI, such as data scientists and AI system developers.”
On the other hand, there is huge potential for AI to do all the mundane and repetitive tasks while people are left to focus on more important and fun work. Soul Machines’ Greg Cross says that: “AI will augment many tasks in the workforce that should be done by machines which will provide humans with the time to spend being creative… The net result of this will be an uplift in productivity in the economy.”
However Tim Dare, Head of Philosophy at the University of Auckland, says it may be hard to replace that ‘human touch’. “There may still be jobs for humans – computers are unlikely ever to be very good at providing emotional support, and sometimes that is what people are looking for from their lawyers and doctors,” he says.
“The main existential threat to humans
is not from software systems with
suddenly emerging autonomy, but
from existing institutions that use AI
The Ethics of Artificial Intelligence
One aspect of AI that has caused tons of controversy is the topic of ethics. Professor Dare, who teaches ethics at The University of Auckland, says it’s difficult to identify ethical truths as people can’t agree on what is ethical. “It’s not that there are none, but we can’t agree on what they are, and ethics requires us to cooperate,” he explains. So while it’s definitely possible to programme an AI with set behaviour that some people might call ethical, us humans are never going to all agree and have a consensus on what is ethical. Just look at religion, gun control or any other hugely divisive issue, and you get the idea.
Mercedes has caused a lot of controversy over their claim their cars will be programmed to prioritise saving the car’s passengers first before pedestrians or other road users, even if that means more people die over all. This obviously is really ethically problematic. NZ Law Foundation Director in Emerging Technologies Colin Gavaghan points out the many unanswered questions regarding AI programmed cars like this. “Should we allow companies to sell a product that will kill more people than necessary? Will this prioritise the lives of the rich – who can afford these sorts of cars – at the expense of more ‘disposable’ poor lives?” he asks.
Aside from the whole ethics thing, there are heaps of concerns about AI being biased or racist in some form, whether intentionally or not. AI will have a huge and wide ranging impact on our lives, from finances to employment to the criminal justice area, so it’s vital that AI makes decisions that are fair. “At first glance, it might be hard to see how an algorithm can be biased,” says Gavaghan. “But algorithms are only as good as the data that informs their decisions.”
AI is Here to Stay
According to the experts, AI is set to be a make or break factor for us. It will either help speed forward technological development that will help us with things from cancer research, to finance and education. Then again, it could also be one of the most silently destructive forces we have ever encountered. To help rein in any negative impacts, really thorough research and well-considered governmental intervention are going to be an absolute must. Victoria University media design lecturer Tom White warns, “Left unchecked, current advances in AI software will continue to exacerbate income inequality and concentration of capital as those with the means of production are able to do increasingly more with fewer labour resources.”
Could a universal base income be a possible solution to this? Colin Gavaghan says, “If AI is going to produce benefits overall, but at the cost of real harms to some, then it seems right that those gains should be shared around to compensate those who lose out.”
As far as work goes, it’s clear that it’s going to be drastically different in the future. Flexibility and some kind of IT-know-how is going to be key for us. Gavaghan believes, “People will need to become more adaptable in the workforce and we should explore as a society how we support people during times of change and reskilling.”
While people will have to compete with machines in the job market, this doesn’t mean all hope will be lost and we should throw up our hands in despair. “As long as we are willing to keep improving ourselves, we will still have a place in this world,” says Senior Lecturer Benjamin Liu.
This may seem daunting to us students, but Greg Cross from Soul Machines reckons it isn’t as hard as you may think. “Information has been democratised by the internet. Preparing for change in the world has never been easier, provided one is curious and has the initiative to learn about it,” he says. “There’s no need to wait for permission in the future.”