Who’s Minding the Machines?
Given all its promise, artificial intelligence (AI) might be considered a great leap forward for humanity. Driverless cars have logged millions of kilometres with few accidents. Police rovers go where humans fear to tread to remove dangerous objects. Medical robots operate with extreme precision to perform life-saving surgery.
But not so fast, says Ian Kerr, whose job it is to question what all this means. The uOttawa law professor and Canada Research Chair in Ethics, Law and Technology warns that AI’s incredible potential is not an unmitigated good.
Artificial intelligence creates systems that can function without human intervention or oversight. For better and worse, we are letting robots out of their cages and intentionally giving them more freedom.
“New techniques in AI allow machines to operate beyond their initial programming, displaying emergent behavior – in some cases without us knowing exactly what the machine will do or why,” Kerr says. “Even though these machines will often make good decisions, we won’t always understand those decisions. And that can be worrisome.”
Most worrying is the advent of AI-guided weapons. These won’t be killer robots running amok and attacking their creators, he says. But if machines are allowed to operate without sufficient human oversight, such weapons could make unintended targeting or kill decisions.
This concern has led Kerr to spearhead a movement involving hundreds of Canadian AI researchers who have sent an open letter to the Canadian government calling for an international ban on AI-controlled weapons. These top AI experts believe that preventing autonomous weapons from being built or used should be an urgent global priority.
Kerr says that even the more benign aspects of AI’s use need careful consideration. For example, take the technology’s medical applications: recent research has shown that AI tools can outperform the most experienced doctors in spotting and diagnosing cancerous tumours, as well as other diseases.
But does this mean that doctors should delegate the job of detecting cancer to these machines? Kerr would argue there are dangers in doing so. Wisdom dictates that human oversight and participation in such crucial decision-making are paramount.
As an ethicist, legal expert and guardian of the public interest, Kerr seeks to answer such questions and propose appropriate legal mechanisms for minding the machines in light of the rapid and accelerating pace of technological change. At uOttawa, he is ideally situated for this task, given that the university is a centre of learning in a city that is a nexus of high-tech research, government and the law.
“It is not an exaggeration to say the uOttawa Faculty of Law is known globally in technology law – and not just AI law and policy, but also when it comes to robotics, blockchain, big data, the Internet and biotechnology,” Kerr says.
His mission is also to bring students into the debate and to get them thinking differently about law, a traditionally a backward-looking field that examines precedent to determine how rulings should be made. Kerr believes it’s time to change the paradigm so that the law looks forward to anticipate situations – and solutions.
It sounds like an opportunity to make sure robots are used to advance human ends, not undermine them
In 2015, the University of Ottawa launched a $400 million fundraising campaign. Defy the Conventional: The Campaign for uOttawa is raising funds to support priorities in every faculty. The campaign will help uOttawa recruit and retain top talent and enrich the student experience. Donations will also support innovative capital projects.