Economists in Stanford The university has found evidence stronger than artificial intelligence begins to eliminate certain jobs. But the story is not so simple: while younger workers are being replaced by some industries, the most experienced workers are emerging new opportunities.
Erik Brynjolfsson, a professor at the University of Stanford, Ruyu Chen, a research scientist, and Bharat Chandar, a postgraduate student, examined the ADP data, the largest family -owned provider in the United States, from the end of 2022, when Chatgpt made his debut, until mid -2015.
The researchers discovered several strong signs in the data, especially that the adoption of the AI generative coincided with a decrease in employment opportunities for younger workers in the sectors previously identified as particularly vulnerable to Automation fueled by IA (Consider the Customer Service and Software Development). In these industries, they found a 16 percent decrease in employment for workers from 22 to 25 years old.
The new study reveals a nuanced image of the impact of AI on work. Although advances in artificial intelligence have often been accompanied by nefarious predictions about the jobs that are eliminated, there has been no data to make a backup. Relative unemployment for young graduates, for example, began to fall around 2009, long before the current wave of AI. And the areas that may seem vulnerable to the AI, such as translation, have seen an increase in jobs in recent years.
“It’s always difficult to know (what happens) if you only look for a particular company or listen to anecdotes,” says Brynjolfsson. “So we wanted to look at it much more systematically.”
By combining payroll data, Stanford’s team found that the impact of AI has more to do with the experience and experience of the worker than the type of work they do. The most experienced employees in industries where the AI generates were isolated from labor displacement, with opportunities that remain flat or a little growing. The finding is backing up what some software developers told me about the impact of AI in their industry, that is, repetitive work, such as writing code to connect to an API, has been easier to automate. Stanford’s study also indicates that the AI eliminates jobs but does not lower salaries, at least so far.
Researchers considered potentially confused factors, such as Covid pandemic, remote increase and recent dismissals in the technology sector. They found that AI has an impact even when these factors are counted.
Brynjolfsson says the study offers a lesson on how to maximize the advantages of the AI throughout the economy. He has long suggested that the Government could change the tax system so that it does not reward companies that replace labor for automation. It also suggests that AI companies develop systems that prioritize the collaboration of human machines.
Brynjolfsson and another Stanford scientist, Andrew Haupt, argued in an article in June that AI companies should develop new reference points for the “centaur” AI that measure human-ii collaboration, to encourage more focus on increasing instead of automation. “I think there are still a lot of tasks where humans and machines can overcome (alas alone),” says Brynjonesson.
Some experts believe that more collaboration between humans and IA could be a feature of the future labor market. Matt Beane, an associate professor at the UC Santa Barbara who studies AI automation, says he hopes that Ai Boom believes a increased demand for labor, as the AI production is becoming more and more important. “We will automate as much as we can,” says Beane. “But that does not mean that there is no increasing mountain of work increasing for humans.”
Ai progresses rapidly, and Brynjolfsson warns that the impact on younger workers could be extended to those with more experience. “What we need to do is create a dashboard system to help us track in real time,” he says. “This is a very consistent technology.”
This is an edition of Will Knight’s Ai lab bulletin. Read the previous newsletters Here.
