AI’s Role in the Workforce: Replacement or Augmentation?

Will AI replace us, or is there no place for AI in the workforce? According to a panel of experts gathered by the Innovation and Technology Caucus of the Texas Legislature (IT Caucus) near the Capitol on Tuesday, neither is the case: Humans augmented with AI will always be better than humans or AI alone.

Daniel Culbertson, a senior economist at Indeed, cited data from Indeed’s AI at Work report released in September, which found that most job postings on the platform aren't “highly exposed” to GenAI.

“Ultimately, when we took those results and paired it back with our job listing data, we found that it was only about 20 percent of the job postings on our site in the U.S. that have a high exposure to generative AI,” said Culbertson. “And it was about 33 percent that had a very low exposure and medium exposure took up the rest of the middle. I think that highlights one of the important takeaways when we're talking about this new generation of AI technologies, that it's more likely to be an augment to jobs in the labor market instead of completely replacing them.”

University of Texas at Austin professors Ken Fleischmann and Sherri Greenberg, chairs of the university’s ethical AI research team, Good Systems, echoed the same sentiments, further arguing for AI as augmentation.

“I think that [the] positions that ... will appear are AI as the assistant,” said Greenberg. “You may actually have a robot that’s helping you, not replacing you. It’s a human in the loop, but AI being of assistance to you.”

Both UT Austin professors also referenced Good Systems’ work with ethical AI applications in skilled trade work, specifically welding programs, in partnership with the city of Austin and Austin Community College. By implementing AI systems and simulated tools, which Good Systems refers to as smart hand tools, Good Systems hopes to improve job safety, streamline training and increase usability for more physical trades.

Hrishika Vuppala, senior partner at global management consulting firm McKinsey and Company, predicted that AI adoption will trigger a mass occupational shift rather than major job losses.

“We've published a lot on this, and we don't project job gains or job losses. What we do project are occupational shifts,” said Vuppala. “In fact, our data says that by 2030, we should expect north of 12 million occupational transitions. This is such an exciting time for us. It calls upon us as individuals, as organizations, as policymakers to say, ‘What can we do to make this more seamless, to make this easier and to make this exciting?’”

Culbertson agreed and expressed a need for fears surrounding AI to be alleviated and active participation in an adaptive economy.

“That comes from governments, employers and workers as well to be welcome and open to that change,” said Culbertson. “I think some of the changes we're going to see might be a little more subtle and a little more nuanced than what many people are expecting. I think part of the reason for that is it's not necessarily going to be wholesale changes across our economy. It's going to be ... jobs changing slightly, and there's going to be a need for upskilling. There's going to be a need for people to be willing to do that themselves, but also to make job changes as necessary.”

Upskilling aside, Culbertson places the overall onus of adaptation on employers, which was a consensus among the panelists.

“I think who needs to step up the most in the future would be employers to provide those opportunities,” said Culbertson. “Employees presenting themselves as a place that workers can come continue to upskill, continue to develop a career, is a great way to attract workers, and that's also a great selling point for public-private partnerships. These public agencies could be going to these employers and saying, ‘Look, this is something that's going to help you attract workers. It can hold on to the workers that you have,’ which is incredibly important.”

Vuppala pushed Culbertson’s point even further, questioning the current standard for credentialing in roles that are likely to be impacted by AI and recommending that companies collaborate with academic partners to create and restructure vocational pipelines.

“A lot of learning actually happens on the job,” Vuppala argued. “Therefore, one has to go back and question the kind of potential that we're asking of people entering into the workforce. Is that necessary? Or can we do something more creative? For instance, I believe the state of Pennsylvania has removed the four-year degree requirement for almost 90 percent of its jobs, and it's working. Something to consider in the same way is, what are some interesting ways in which we can think about workforce credentialing through digital means so that it's much easier for employers to identify the people in the market that they can take as opposed to using more traditional means?”

The panel of experts all agreed that employers looking to attract talent and retain workers would benefit from incorporating AI into their training and hiring considerations as soon as possible. Likewise, Greenberg encouraged both employers and education institutions to integrate technical skills in AI and data ethics into their curriculums.

“I'm fond of saying, ‘Your education starts the day you're born and it ends the day we die,’” said Greenberg. “I think that it's incumbent upon our institutions of education from elementary school on … that education and curriculum should be part of how we engage our residents, how we engage people in the state, local, federal and within our employers.”

During another panel held during the event on AI’s role in local and state government, Texas CIO and Executive Director of the Department of Information Resources (DIR) Amanda Crawford and other industry experts echoed many of the same sentiments.