Jason Furman and Anthropic Chief Economist Peter McCrory say the evidence so far does not support the sweeping predictions that artificial intelligence would rapidly eliminate large numbers of jobs across the U.S. economy.
Furman pointed to the current unemployment rate and broader economic data, arguing that the dramatic forecasts of mass displacement have not materialized.
“They also told me that all the jobs would be gone by basically now,” Furman said. “I was rolling my eyes, and I’d like to think that they’re on the West Coast right now, talking to someone else, telling this exact same story about how they were chastened and wrong.”
He noted that unemployment remains historically low and said the broader economic data does not show anything close to the scale of disruption that some AI forecasts suggested.
“So why is this? The unemployment rate is 4.1 percent. It’s basically sustained basis, the lowest we’ve ever seen,” Furman said.
Furman acknowledged that there may be signs of stronger productivity growth, but said the numbers do not yet show an economic transformation matching the magnitude of the technology itself.
“Productivity growth-you can squint and argue it’s picked up a little bit. Total factor productivity hasn’t picked up really at all, arguably, but just nothing you can see in the economic data that incredibly obviously screams anything commensurate with the magnitude of this technology,” Furman said.
McCrory agreed that unemployment remains low, but said he sees more evidence of productivity improvements in sectors where AI adoption has been stronger.
“My sense from looking at the labor productivity numbers is that it’s a little bit more than squinting,” McCrory said.
He said productivity growth appears stronger in parts of the economy that have incorporated artificial intelligence tools.
“You can see that productivity growth has been relatively stronger in sectors of the economy where AI has been adopted,” McCrory said.
At the same time, he emphasized that the labor market remains strong.
“But exactly to your point, the unemployment rate is close to what the Fed would deem as maximum unemployment. The prime-age employment-to-population ratio is at multi-decade highs, and so what explains that?” McCrory said.
His answer is that AI currently appears to be augmenting human labor rather than replacing it wholesale.
“Here, across a lot of our research, what I see in our data is that it looks more like a skill-biased, labor-augmenting technology than a technology that outright automates and displaces jobs wholesale,” McCrory said.
He pointed to research based on the Department of Labor’s O*NET taxonomy. According to McCrory, about half of U.S. jobs include at least some tasks that resemble the types of work people are using Claude to perform.
“so for about half of all jobs across the U.S. economy, based on the Department of Labor’s O*NET taxonomy, we see a quarter of the tasks showing up as the sorts of things that people are using Claude for,” McCrory said.
However, he said no occupation in that taxonomy currently appears to have every task automated by Claude.
“But there’s no job in that taxonomy where every single task is being automated by Claude,” McCrory said.
He argued that the tasks AI still struggles to perform act as a barrier to removing humans entirely from many jobs.
“and those weak links, essential aspects of our jobs that are hard to automate, limit the extent to which you can fully remove the human in the loop,” McCrory said.
McCrory also said sophisticated AI output often depends on equally sophisticated human input.
“Moreover, when we look at what actually precedes very complex output from the model, so if I ask Claude to, or I see people using Claude to build very complex financial models systematically across tasks and across geographies, people are providing sophisticated complex inputs,” he said.
That means the quality of the final result can depend heavily on the expertise of the person using the system.
“So that complex output is in part reliant on the complexity of input provided by the humans,” McCrory said.
He added that AI systems tend to perform better when people remain actively involved.
“And we find evidence that when there’s humans in the loop, the models tend to be more successful,” McCrory said.
McCrory said AI can therefore allow workers to take on more difficult tasks instead of simply eliminating their jobs.
“People are able to tackle harder, and more complex problems, and the trade-off between complexity and success with using AI is lower. So there’s less of a trade-off,” he said.
His conclusion is that AI is changing which skills are more valuable, but is not yet producing the wholesale labor displacement many predicted.
“All of that points in the direction of changing and restructuring returns to different forms of expertise, but not wholesale displacement and disruption for for workers who are affected by AI.”
WATCH:
!function(r,u,m,b,l,e){r._Rumble=b,r[b]||(r[b]=function(){(r[b]._=r[b]._||[]).push(arguments);if(r[b]._.length==1){l=u.createElement(m),e=u.getElementsByTagName(m)[0],l.async=1,l.src="https://rumble.com/embedJS/u1vds3"+(arguments[1].video?'.'+arguments[1].video:'')+"/?url="+encodeURIComponent(location.href)+"&args="+encodeURIComponent(JSON.stringify([].slice.apply(arguments))),e.parentNode.insertBefore(l,e)}})}(window, document, "script", "Rumble");
Rumble("play", {"video":"v7e082o","div":"rumble_v7e082o"});