Feature
Proponents of AI are continually reassuring people that while the technology is sure to change jobs, it won’t eliminate them. AI, they say, will organize and handle tasks more quickly, increasing efficiency and productivity while allowing workers to spend less time on routine work and more time on strategy. According to these proponents, AI’s end result on employment will be less dramatic than many people believe.
Writing in The Wall Street Journal, The Wharton School’s Peter Cappelli and Valery Yakubovich identify several obstacles to the wholesale elimination of jobs by AI. For one thing, they argue, much of AI’s promise is theoretical. There’s a vast gulf between what the technology can achieve on paper compared to what it’s actually done. For another, most occupations involve work whose complexity goes beyond AI’s ability to summarize information and respond to prompts.
In addition, generative AI's results can simply be wrong at times. When it can’t find an answer, GenAI is prone to what has been dubbed “hallucinations,” in which the bot simply makes up facts to fill in gaps. Attorneys in New York, Colorado and Vancouver have learned this the hard way when they included legal citations made up by ChatGPT in court filings. “I did not comprehend that ChatGPT could fabricate cases,” one of them told the judge.
The Demand for AI Returns Heat Up
The fact that the technology can hallucinate or make up facts isn’t necessarily a result of carelessness. More likely, it resulted from developers’ efforts to strike a balance between the quality of information and the risks presented by the technology.
“To err isn’t just human,” Tsung-Hsien Wen, CTO at PolyAI, a London-based developer of intelligent digital assistants, told PYMNTS. “Just as you can’t guarantee 100% efficacy from the human brain, even the most sophisticated language models operating on near-perfect neural network designs will make a mistake on occasion.”
Developers believe one of the most effective ways to deal with hallucinations and other challenges is to make sure human beings are involved in the process. “Nowadays, in many cases … both human moderation and explainability of the AI tools are considered mandatory,” said Roman Eloshvili, founder and CEO of XData Group.
Still, the idea that analysis is only as good as its underlying data holds true. So while AI may not enable the mass elimination of jobs, workers remain nervous. They recognize that business leaders have come under pressure to demonstrate some kind of return on their AI investments.
According to one survey cited by The Wall Street Journal, more than two-thirds of executives feel pressure from investors to show how spending money on AI will somehow bolster the bottom line.
What Employers Say vs. What Employers Do
Bloomberg Intelligence recently reported that AI will allow Wall Street banks to eliminate up to 200,000 jobs in the next three to five years, with workers involved with data analysis, financial trend assessment and risk evaluation being the most vulnerable. Financial companies are looking at ways AI could “not only supplement but supplant” scores of workers, wrote The New York Times.
That’s not surprising. AI’s potential to improve efficiency, boost productivity and advance automation has a number of businesses thinking about how the technology can be used to cut workforce-related costs. “Wider adoption of a job reduction mindset could have ripple effects across the economy and exacerbate unemployment for white-collar workers, which has been highest in tech, law and media,” read The Wall Street Journal article.
Much of this thinking is rooted in the idea of finding efficiencies, but part of cost avoidance involves rechanneling work from new hires to existing employees. “Adding folks is not just adding a body, it’s adding the management,” one executive told the Wall Street Journal. “We really want to focus on keeping the staff that we have, upskilling them and then just giving them better tools to do their jobs.”
That would be fine if employers actually put their money where their mouth is. Reskilling workers because of AI is the lowest priority for CHROs and other HR leaders, according to research by The Conference Board. While most said experimenting with AI was their top priority, only 7% have implemented reskilling strategies for affected job roles. Many of those employer experiments involve new ways a company can put AI to work, with a focus on pilot programs and use cases related to HCM.
Conference
Oct
20
Gartner IT Symposium/Xpo Orlando 2025
Register
Conference
Oct
27
Gartner HR Symposium/Xpo Orlando 2025
Register
Almost 40% of workers worldwide could feel the effects of AI, according to the International Monetary Fund. Unlike past instances of technology and automation, AI can impact work that goes beyond routine chores and requires advanced skills. Because of that, the IMF said, a greater proportion of jobs (60%) may be affected by AI. About half of those may result from AI enhancing productivity. For the other half, AI’s ability to take on work now performed by humans could lower the demand for labor, which in turn will pressure wages and hiring.
Experts may say AI’s impact on the workforce will be manageable, but that doesn’t mean executives will stop looking for ways to save money. Two-hundred thousand people on Wall Street may see that firsthand.
Related Article: Companies Say Generative AI Won't Come for Jobs. Employees Aren't So Sure
