Will AI make work obsolete?

Tech leaders and economists debate whether artificial intelligence will lead to mass job displacement or new career opportunities.

Recent surveys show that over half of U.S. workers are concerned about artificial intelligence’s impact on their jobs, and nearly 60% of young people say they fear how the technology could shape their employment prospects. For the first time in generations, the assumption that a college degree guarantees stability feels less certain as tools such as ChatGPT and Claude write code, diagnose diseases, and even generate art.

To debate whether AI will ultimately make work obsolete, The Hopkins Forum, a live debate series from Johns Hopkins University and Open to Debate, brought four experts to the Johns Hopkins University Bloomberg Center to examine the stakes. Explore their arguments on both sides of one of today’s defining economic and ethical questions.

AI will reshape the labor market, not eliminate it

Chris Hughes, co-founder of Facebook and the Economic Security Project, offered a historical parallel to the advancement of AI. When Barclays Bank installed the world’s first ATM in 1967, bank tellers feared for their jobs, even covering the ATM keypad with honey to make it less productive. But by 1970, there were roughly 300,000 bank tellers in the United States, and four decades later—after ATMs had spread nationwide—there were about 600,000.

ATMs made branches cheaper to operate, allowing banks to open more locations. For Hughes, this pattern repeats across industries. Labor markets adapt. Today, he noted, 60% of people employed in the United States work in jobs that didn’t exist in 1940.

“I don’t think we can even conceive of the jobs that will be created in the future,” Hughes said. “Let’s imagine AI enables us to speak with animals or understand their own language. Would you need a mediator to help if a pack of gorillas was in conflict with another?”

But even the framing of the question “Will AI make work obsolete?” may obscure what is really at stake, according to Rumman Chowdhury, CEO of Humane Intelligence, a nonprofit focused on AI safety. AI itself does not make work disappear. It is human choices that determine which tasks are automated.

“In 2018, I coined a phrase: moral outsourcing,” she said. “This was pre-Chat GPT, pre-Gen AI. What I noticed was this very dangerous language of anthropomorphizing AI in our sentences. Even in the premise of today ‘Will AI make work obsolete?’…AI does not ‘do’ anything. It has no wants, needs, desires. … Now, one thing we should think about is what the decisions are that these CEOs are making.”

Chowdhury argued that the real concern is how CEOs choose to use this new tool and that their focus should be on workforce development. If companies automate entry-level roles without planning for talent pipelines, then senior developers, programmers, or designers may cease to exist. With careful planning and investment in training, AI can complement human work rather than eliminate it.

AI’s unique capabilities put the labor market at risk

Unlike previous industrial revolutions, today’s technological changes are fundamentally different from past waves of automation that society has adapted to, according to Simon Johnson, Nobel Prize-winning economist and Massachusetts Institute of Technology professor.

“This is not the age of the textile machine or the automobile or the bank teller machine,” Johnson said. “This is the age of artificial intelligence, and artificial intelligence is designed deliberately to replicate and replace human intelligence.”

Entrepreneur and former presidential candidate Andrew Yang echoed that concern with his own experience running his company, Noble Mobile. The company had job postings for junior engineers until recently, when Yang’s chief technology officer said that AI tools are so good that the postings could be removed. And that’s not only happening with smaller companies, Yang said.

“I recently sat with the CEO of a publicly traded tech company who told me this: ‘We are going to fire 15% of our workers. Two years from now, we’re going to fire another 20% of workers. Two years after that, we’re going to fire another 20% of workers,” Yang said. “His last statement was, ‘I have no idea what college students are going to do in five years.’”

For decades, Americans viewed a college degree as the surest path to a stable and high-paying job. But now the wage gap between U.S. college graduates and those without college experience is shrinking. The disruption, Yang argued, will not be limited to a handful of sectors.

“It is not just coders or designers or the millions of customer service personnel,” he said. “It’s anything with the words research or analyst in it, finance, law, consulting.” Even fields once considered the best for job security, such as computer coding, are now seeing graduates struggle to land jobs.

Preparing the labor market for AI disruption

Despite disagreements, the panelists converged on one point: The transition is inevitable, and policy choices matter. The outcome of this new era will depend not only on technological capability but also on how governments, companies, and workers respond.

“In my mind, instead of trying to debate whether this is happening,” Yang said, “it’s to accept that this is real and then move forward to solutions.”