Jump ahead to:
Can AI replace human jobs?
Around the world, ChatGPT and its generative AI-powered counterparts are sparking curiosity about the future of work. Optimists predict that these artificial intelligence tools will manage menial parts of our jobs, like data entry and bookkeeping, and leave workers with more time to tackle more sophisticated projects. Others are less bullish: Last month, hundreds of well-known business leaders—including tech trailblazers like Bill Gates and Open AI CEO Sam Altman—signed an open letter warning that AI could not only take over our jobs, but also destroy humanity.
One thing is for certain: We are in the thralls of a quickly shifting economy. The World Economic Forum predicted there will be 97 million jobs created and 85 million axed worldwide between 2020 and 2025 due to automation and a new division of labour between humans and machines.
We asked two experts what are the top five jobs that ChatGPT and AI could replace—or at least drastically change—over the next five years.
Human resource management
Some large corporations have already trimmed their human resources department: IBM, for example, announced in May that its own AI, WatsonX, will soon replace 8,000 jobs at the company, especially in its HR department. IBM CEO Arvind Krishna said the AI transition will happen gradually over the next few years, with machines handling tasks like providing employment verification letters or moving employees between departments.
Sima Sajjadiani, an assistant professor at UBC’s Sauder School of Business who specializes in human resources, expects these types of cuts to become common. “If we can outsource repetitive HR jobs like resume screening, form filling and digital onboarding, departments can function with fewer people,” she says. But, she says, companies are nowhere near the point of outsourcing critical tasks, like hiring decisions and team building, to robots, as these responsibilities require the “human” aspect of HR. Likewise, leading a company-wide offsite that fosters IRL relationships cannot be done by AI.
Generative AI is allowing the automation of an ever-increasing number of varied customer-service tasks. Canadian banks, for example, rely on automation to deliver financial statements and offer investment advice through robo-advisors. Likewise, many businesses use automation to support customers solve problems online through functions like recognizing customer needs, translating their requests, and providing simple solutions.
But customer service reps who help clients through emotional situations like locating lost money, navigating a new service or troubleshooting important offerings like mortgage issues or locked bank accounts, will still have value in the future, says Sajjadiani.
“We all have experience trying to get a machine to fix our problem and being very frustrated with it,” she says. “When that happens, you just want to get on the phone with a human being who can be sympathetic to you and help you through your problem. That feeling won’t change anytime soon.”
Don’t worry: A robot is not likely to conduct your next physical exam. Instead, the AI revolution in medicine will happen behind the scenes in diagnostics. “Physicians who focus on diagnoses should worry, because machines will soon do what they do, and do it better,” says Avi Goldfarb, chair in artificial intelligence and health care at the Rotman School of Business.
For example, AI technologies are learning to detect brain changes indicative of Alzheimer’s disease, predict osteoporosis based on bone profile, and analyze heart rate to detect cardiac problems. Doctors in such roles, continues Goldfarb, should expect a shift in their job description—but that doesn’t mean it’s an entirely bad thing. Spending less time on diagnosing diseases could allow them to focus on other parts of their job, like prescribing treatments.
Coders and programmers are already feeling the squeeze of AI: Crunchbase reported in June that around 5 per cent of May’s tech job cuts in the U.S. were directly related to artificial intelligence. Goldfarb says that, like with customer service reps, programmers can now outsource several of their more mundane tasks—like debugging and copy and pasting simple code—to machines. So, the need for coders who add little to no value beyond those tasks decreases, he says.
“Programmers who can think through problems and come up with ways to use computers effectively to solve them will be indispensable,” he says. “But if all you are doing is line-by-line coding, you might be in trouble.”
ChatGPT might increasingly replace the tasks of content writers, including data analysis, research, report writing—and even BuzzFeed quizzes. But while ChatGPT can write poetry in the style of Shakespeare and Drake, it isn’t known for its factual accuracy just yet: In May, when a New York lawyer used the tool to search for case precedents, it led him to present false information in court.
That’s why when it comes to original research or real-time journalism (*ahem*), humans will have the upper hand. “We will always need reporters,” says Goldfarb, “but AI is making the value of someone who simply writes recycled content decrease.”