Economist Daron Acemoglu, Nobel laureate in Economics and professor at MIT, warns that unemployment in the United States could triple in the next 10 years if the development of artificial intelligence continues on its current trajectory.
The automation thus obtained would eliminate more jobs than the new activities generated by AI would create.
Daron Acemoglu draws attention to the unprecedented speed at which AI is currently transforming the labor market. The current pace of automation is much faster than in previous technological revolutions.
"What we are experiencing now with AI is unprecedented," said Acemoglu, quoted by CBS News.
Comparing the current transformation with the first stage of the Industrial Revolution, the latter took place over approximately 80 years. "Today it happens in one or two years and occurs in many different sectors at the same time," emphasized the Nobel laureate.
New jobs would not compensate for the losses
Obviously, the development of AI will also create jobs, including for individuals involved in training and developing artificial intelligence systems.
However, the issue lies in the difference between the number of these positions and the number of jobs threatened by automation.
"The number of people hired in these positions is small compared to the number that will be replaced," explained Acemoglu.
This is, in fact, the purpose of automation: reducing the human work needed to perform certain activities.
Although the unemployment rate in the US is currently relatively stable, some data already indicate effects on certain categories of employees, the source added.
Research conducted at Stanford shows that hiring has decreased among young people for jobs most exposed to AI, such as those in software development. Data from the US Census Bureau also indicate decreases in hiring and salaries for college graduates in fields exposed to artificial intelligence.
"It's not about saying NO to technology"
Economist Acemoglu does not advocate for stopping the development of AI, but for changing the way it is used.
"Absolutely not," replied the economist when asked if the solution would be to reject technology. In his opinion, companies could design AI systems to make people more productive, rather than turning them into easily dispensable employees.
Another solution could be changing the tax system so that companies have greater incentives to hire people and not just to automate activities.
However, the economist warns that time for such decisions is limited. The future of the labor market is not predetermined, says Acemoglu, and the policies adopted now can influence how AI will transform the economy.
"Because the future is so uncertain, because it depends so much on our choices and actions, it is all the more important to have this discussion," said the Nobel laureate.