While AI is not yet a substitute for human employees, advances in the technology have companies and economists wondering what impact AI will have on the workforce in the future.

. . US labor productivity growth has returned to its pre-COVID-19 anemic trend with stark consequences for the economy. Working Smarter is a four-part series that explores how the issue and the push to adopt new technology to try to fix it affects businesses and workers. AI offers welcome potential boost to sluggish US productivity growth Employees resisting return to office calls as companies reimagine workspaces Private equity looks at speedy AI adoption to boost value, maximize efficiency .

Generative AI — a type of AI that can produce content, such as text, images, code or audio — has become a focal point in 2023 with the emergence of OpenAI LLC's ChatGPT and other large-language models. Companies across industries, including mobile payment company Block Inc., private equity firm The Carlyle Group Inc. and cosmetics giant The Estée Lauder Cos. Inc., have become early adopters of ChatGPT Enterprise. Others are building models internally, such as consulting giant Ernst & Young Global Ltd. launching its own AI platform in September following a $1.4 billion investment. Though companies are looking to maintain their competitive edge and improve worker productivity through the adoption of AI, workers have been left hanging in the balance as the threat of eventually being supplanted by the technology looms large.

Those fears are expected to materialize earliest for certain white-collar occupations, particularly those with repeatable, structured workflows, said Yong Suk Lee, assistant professor of technology, economy and global affairs at the University of Notre Dame. While historical precedent suggests that advances in technology will open new types of occupations, AI adoption in the workplace also gives rise to a fresh set of concerns about inequality.

"If a big chunk of the population [is] not as comfortable with these new technologies, there is going to be some imbalance in new jobs being created and jobs being displaced," Lee said.

An inflection point

The next few years will be crucial in determining the impact of AI on white-collar workers and their jobs.

A June 2023 survey of over 700 mid-level and senior IT professionals by 451 Research shed light on the race to apply AI tools across US companies. When asked about their organization's investment plans, 64.6% said they had already invested in specific use cases where generative AI had been leveraged, while 33.3% said they had not invested but planned to in the future. A mere 2.1% said they had no plans to invest in the technology.

"We are at an inflection point, where the next few months and even years will be critical in determining what work is going to look like in terms of AI playing a role in workforce productivity," said Conner Forrest, a workforce productivity and collaboration analyst with S&P Global Market Intelligence. "We know that trust is high and AI technologies are being deployed in traditionally back-office departments such as human resources."

In many cases, adopting AI tools into existing workflows does not necessarily involve a worker's active decision to do so, as AI is being incorporated into software already in use. Patterns of AI adoption can be problematic to pin down for other reasons, too.

"Anyone who uses Google tools for search or Amazon for procurement is engaging with AI as part of their normal work, even if it just seems like an incremental improvement to their existing workflows," said Ben Armstrong, executive director at MIT's Work of the Future initiative.

Productivity boost

One of the foremost justifications for AI usage in the workplace is its anticipated positive impact on employee productivity.

"The idea is that AI removes the manual tasks, the repeatable tasks that are associated with workplace friction, and optimizes roles to be the truest version of themselves," said Market Intelligence's Forrest.

Even as some workflows could already be executed entirely without human workers, the vast majority will continue to be optimized rather than overtaken by the implementation of AI tools, said Ramayya Krishnan, faculty director of Carnegie Mellon University's Block Center for Technology and Society.

"At this stage, I expect augmentation selectively in some tasks versus others, depending on the skill of the worker," Krishnan said. "As we gain more experience, this may result in workflows being reimagined, but we are not there yet."

451 Research survey results suggest that excitement about the potential benefits of AI in the workplace outranks fears of job displacement, at least for mid-level and senior decision-makers. When asked about the top benefits of automation, almost 54% of those surveyed named productivity and 46% selected time savings. When asked about top concerns, only 17% named the risk of job displacement. Concerns about inaccurate output creating more work were far more prominent.

Work of the future

Those skilled at process improvement — including the ability to look at a workflow and identify ways to improve it — are likely to become increasingly vital to organizations adopting generative AI technologies, Armstrong said.

"Given the importance of editing and process improvement, I think there remains extraordinary value in mastery and expertise," Armstrong said.

While high-level positions may be safer, entry-level positions in professions such as copyediting or coding could vanish as the technology improves. Within the broader generative AI market, text generators are expected to bring in $1.1 billion in revenue in 2023, according to 451 Research estimates. Code generation is much smaller, with projected 2023 revenues of $231 million, but the segment is expected to experience the strongest growth. 451 Research forecasts for 2023–2028 suggest a compound annual growth rate of almost 73% for code generators.

"I think [code generation] is one of the areas we might see quite a lot of productivity growth," said Alex Johnston, a data, AI and analytics analyst at Market Intelligence.

With entry-level positions potentially more at risk than senior ones, that could exacerbate inequality, said Notre Dame's Lee.

"While we will see a productivity increase happening at the upper end of the same occupation, it might look like ChatGPT can also complement lower-skilled workers. But at some point, it could potentially just replace them," Lee said. "As the technology improves, it's probably going to shift [toward the] higher end, but very highly productive workers will maintain their jobs and perhaps benefit from it."

451 Research is part of S&P Global Market Intelligence.