Many employees will need to "substantially change the work they do" as the artificial-intelligence revolution takes hold, McKinsey analysts say, but they shouldn't worry about being out of work entirely - at least in the medium term.
AI was previously thought to bring risk primarily for workers in low-skilled jobs who were doing tasks that could be easily automated. But the recent wave of interest in generative AI, fueled by the popularity of ChatGPT, has made it clear that higher-skilled knowledge workers in fields ranging from software engineering to marketing to sales could also see their daily tasks upended by technology.
In fact, businesses on the whole could recognize $2.6 trillion to $4.4 trillion in additional value by leveraging generative AI, a type of artificial intelligence that can produce content and that relies on foundational models, according to a new report from McKinsey & Co. that assessed potential impacts on 63 different business-use cases. Companies have the opportunity to become more productive by automating tasks currently done by human workers, resulting in increased revenue or profit.
At first glance, the concept may sound good to workers: A machine could instantly dig up customer data, get a coding project started or automatically tailor email marketing campaigns for different scenarios, reducing the amount of time that employees must spend on menial aspects of their jobs. The reality, though, will likely be more nuanced. Companies will have to choose whether to reassign workers to do higher-value tasks while AI handles the rote work - or keep their scope of operations, laying off workers and taking their profits to the bank.
Also read: Yes, AI is coming for your job. Here's how to prepare.
"What happens when you deploy these technologies is going to be a decision," said Michael Chui, a McKinsey partner who is one of the report's authors. Software companies should be able to dramatically accelerate the pace at which they develop software if they let machines write the first draft of programming, he said. But in a case where a company could boost productivity of its workers by 50% through AI, he doubts many would opt to lay off a third of their workers.
"If you ask companies if they need more or less software written, they want more," he said, and would likely press workers to "produce more and better software."
He even sees the possibility that some companies would move toward a four-day workweek, mirroring the way employers started requiring fewer working hours from employees after the Industrial Revolution: "You could simply give that time back," he said.
Still, Chui acknowledges that not all companies will choose to expand what they do once machines take on more tasks. "Other managers might say we're going to take productivity potential directly to the bottom line," he told MarketWatch.
Now read: Artificial intelligence is coming for seniors as AI's dark side targets older adults in scams
While some fear AI will put people out of work en masse, Chui says that previous technological revolutions would suggest a less dramatic medium-term picture.
"We don't have a lot of people who operate elevators anymore, but it's not the case that we're at 70% unemployment," he said. "We have always found more work for people to do."
That work is going to have to change, however. Teachers, for example, might swap out time spent grading assignments for hours spent providing tutoring, the authors noted in the report. Other workers may have to change occupations entirely.
"If worker transitions and other risks can be managed, generative AI could contribute substantively to economic growth and support a more sustainable, inclusive world," the authors wrote.
In a way, the generative-AI revolution may be somewhat equalizing, Chui noted. "People who have college degrees, graduate degrees, might have previously thought all this tech stuff affects frontline workers," but AI has proven able to handle the types of analyses that MBA students learn to do, for instance.
See also: How will AI affect stocks of semiconductor companies? Here are the 18 expected to grow their sales most quickly.
Could the growing relevance of AI make the working world more equitable for individuals without fancy degrees? Perhaps, Chui said.
"People have noted prior to this technology that it would be useful to build a more skills-based workforce," he said. A four-year degree "shouldn't be a barrier to people who have the appropriate skills" to take on a number of roles, including software-engineering positions and executive assistant jobs.
The key for employees will be showing that they can use AI to augment their own productivity.
"Workers will be more competitive workers if they know how to use these technologies," Chui said.