Artificial intelligence is transforming the world, something that will also affect employment and could, in some cases, lead to job losses.
An experiment developed by teams at MIT and Oak Ridge National Laboratory has created a simulation designed to analyze the extent to which a set of autonomous agents could replace jobs currently held by people in the United States.
The project, known as Project Iceberg, functions as a digital twin of the labor market. Its objective is not to accurately predict the future, but to observe what would happen if a workforce powered by artificial intelligence operated in parallel to the country’s more than 151 million real workers.
To do this, ORNL’s supercomputing infrastructure is used, capable of handling complex scenarios with thousands of economic and demographic variables.
According to data collected by CNBC, this theoretical workforce could cover 11.7% of the US labor marketand the potential economic impact could reach $1.2 trillion, especially in sectors such as finance, healthcare and professional services, where automation shows greater capacity to replicate human tasks.
The model analyzes 151 million employees, classifying their tasks, location and skills. In total, maps more than 32,000 professional skills, distributed among 923 occupations and 3,000 counties. Using this data, researchers compare the ability of artificial intelligence agents to perform the same jobs and thus estimate the degree of possible replacement.
Although this initiative cannot precisely anticipate how employment will evolve, it stands out for its usefulness as a preventive tool. Unlike other studies, Project Iceberg covers all 50 states and can break down the impact even at the census block level.
This report comes at a time of uncertainty, since some companies have had to reinstate employees after trying to replace them using artificial intelligence systems.