AI will only steal jobs from a few people

Researchers at the Massachusetts Institute of Technology (MIT) examined how much AI costs companies. The result: Because artificial intelligence is often too expensive, only a few jobs are at risk.

Since the continuous development of artificial intelligence, many people are asking themselves one question: Will AI take my job away? Pessimists warn that AI systems could drive up the number of unemployed. But so far that has not happened. Because it doesn't seem worth automating jobs.

A study by the Massachusetts Institute of Technology (MIT) recently clarified this. The researchers looked at the costs of artificial intelligence (AI) compared to human labor. They came to a surprising result. Despite the rapid progress in AI technology, human workers are still more cost-efficient in many areas. An AI revolution therefore seems rather unlikely.

Costs of AI: Cheaper for humans in many areas

The researchers focused primarily on the use of computer vision. This is a form of AI that can process visual data from the real world. They found that 23 percent of the jobs studied could actually be made more cost-efficient through the use of AI.

For the remaining 77 percent, however, using humans is still the better and more cost-effective choice. A key reason for the limited cost effectiveness of AI is the high costs of initial installation and operation. The researchers analyzed data from 1,000 different tasks in 800 professions and found that AI can currently only perform about three percent of these tasks cost-effectively.

But that could also change. The team predicts that this value could increase significantly by 2030. Thanks to the greater efficiency of AI applications, algorithms could cost-effectively take on 40 percent of the tasks currently being investigated.

AI as a tool to optimize processes

Especially in retail, transport and warehousing, AI with computer vision shows the potential to optimize processes and make them more efficient. In contrast, in craft sectors such as construction or mining, as well as in the real estate sector, AI applications are currently not very cost-effective.

The study shows that AI can be seen as a supporting tool rather than a complete replacement for human workers. It seems more likely that we will collaborate more with algorithms in the coming years and not be replaced by them.

Also interesting: