Only 23 percent of workers’ wages could be cost-effectively replaced by AI, the study said.
Artificial intelligence (AI) may not be coming to your jobs just yet. A new study from the Massachusetts Institute of Technology (MIT) finds that AI may still be expensive for companies to replace human workers. The research looked at the practical aspects of replacing human labor with AI in jobs in the United States where computer vision – a type of AI that derives information from images and video – was employed. They found that most of the time, it would be cheaper for firms to continue using human workers for certain tasks than to use AI.
“Our study on automation with AI computer vision systems provides some encouraging news about job substitution, showing that Many tasks will not be economically attractive to automate for years or decades.” Laboratory (CSAIL) said CNBC Make it.
According to the study, only 23 percent of labor costs can be cost-effectively replaced by artificial intelligence. In some cases, human workers are still more cost-effective because AI-assisted visual recognition technology is expensive to install and operate, the researchers said. He also predicted that computer vision operations would still take decades to become financially viable for companies, even with a 20 percent reduction in costs per year.
“We find that only 23 percent of the compensation of workers exposed to AI computer vision would be cost-effective for firms to automate because of the large costs of AI systems,” the study states. From”.
The MIT researchers noted that previous predictions and studies about the timeline and extent of automation were particularly vague “because they did not directly consider the technical feasibility or economic viability of AI systems, but rather the Rather than using measures of similarity as tasks and AI capabilities indicate exposure.”
Also Read | A study of AI used in a military conflict simulation – it chose war every time.
However, he also said that AI could disrupt the job market — now and in the future. “Even in the short term, there will be workers whose jobs will be lost or whose responsibilities will change because of AI,” said Neal Thompson. CNBC Make it.
According to Mr. Thompson, in the future, AI systems may cost less to develop, potentially reducing how quickly companies choose to use them. However, it will likely take years for these costs to come down to a level where these AI systems can be widely deployed by companies.