The seemingly unstoppable march of artificial intelligence (AI) continues to garner attention from business leaders, policymakers, and media commentators. A recent example is the breathless excitement around OpenAI’s ChatGPT—a “chatbot” capable of quickly completing many tasks previously reserved for humans. Since its release last fall, ChatGPT has been used to compose letters and entire essays, draw up wills, manage customer inquiries and produce training manuals for employees. There’s evidence that this so-called “generative” form of AI can even do coding and basic computer programming.

Not surprisingly, the latest developments in AI are sparking fears that vast numbers of jobs may be at risk as machines substitute for human labour across a growing array of occupations. In fact, a group of scientists, technology experts and academics recently launched a petition calling for a six-month halt to work on AI. That’s unlikely, to say the least. As countries and companies scramble to adapt to the digital economy, it’s unrealistic to expect everyone will agree to down tools.

Economists at U.S. investment bank Goldman Sachs estimate that AI could disrupt up to 300 million jobs worldwide within the next few years. Among the sectors most likely to be affected are finance, publishing, law, education, customer services, market research and other industries where large numbers of employees assemble, analyze and disseminate information.

However, consistent with past waves of technological innovation, the spread of AI will also spur employment growth in some industries and alter many current jobs. The same Goldman Sachs study finds that AI will increase labour productivity and boost global economic output by 5 to 7 per cent.

Advances in information and other technologies are hardly new. Automation of production and business processes has been underway for centuries, albeit the pace has accelerated in the last two decades. History furnishes little support for the view that the invention and take-up of new technologies leads to sharp and sustained declines in the total number of jobs in the economy.

Just consider today’s labour market. We live in a time of rapid technological change, yet almost everyone who wants a job is employed. In Canada and the United States, job vacancy rates hit all-time highs in 2022 while unemployment rates plumbed new depths. Human resources consulting firm ManpowerGroup says 77 per cent of the 39,000 firms it recently surveyed worldwide are experiencing difficulties in filling jobs, signalling a widespread talent shortage. The information technology sector apparently faces the biggest challenges in hiring and keeping staff, even as we read about large-scale layoffs by big technology companies.

To be sure, continued automation will affect many existing jobs, even as it eliminates some and stimulates employment growth in other parts of the economy. New research indicates that AI will impact up to two-thirds of all occupations. Most often, this will involve certain work tasks being automated, rather than jobs just disappearing. In some cases, automation will complement worker skills and knowledge, rather than substituting for them. As U.S. economic blogger Noah Smith puts it, automating tasks isn’t the same thing as automating jobs. At this point, we probably shouldn’t expect AI to trigger an employment apocalypse.