AI in practice

IBM aims to replace up to 7800 jobs with AI over the next 5 years

Matthias Bastian
IBM could replace up to 7800 jobs with AI in the next five years

Midjourney prompted by THE DECODER

IBM CEO Arvind Krishna has announced a hiring freeze for positions that could be replaced by artificial intelligence (AI) in the coming years.

IBM CEO Arvind Krishna told Bloomberg News that his company is instituting a hiring freeze for positions that could be replaced by artificial intelligence in the coming years.

That mainly affects back-office positions like human resources, where about 30 percent of the roughly 26,000 positions could be replaced by AI and automation within five years. That amounts to about 7,800 positions. Krishna says he can "easily see" that number. Some of that reduction would come from natural attrition and positions that are not re-staffed.

Evaluating people remains a people job

Simple tasks such as issuing work certificates or transferring employees could be fully automated. Other HR functions, such as assessing workforce structure and productivity, are unlikely to be replaced over the next decade, Krishna said.

IBM currently employs about 260,000 people. It continues to hire for software development and customer-facing roles. In the first quarter, the company added 7,000 employees despite announcing plans to cut about 5,000 jobs. Through productivity and efficiency gains, IBM expects to save $2 billion annually by the end of 2024.

GPT-4 spurs labor market fantasies

GPT-4 in particular is starting to make managers think big. The large language model shows how it can be used as a new interface that goes beyond text generation to make individual workers more productive, or to automate complex processes with tools like Auto-GPT. As a result, it has people worried - or excited - about their professional futures when it comes to working with data, text, and code.

OpenAI, the company behind GPT-4, recently published a study on the potential impact of artificial intelligence on the job market. According to the study, for about 80 percent of U.S. workers, large language models will affect at least ten percent of their jobs. For 19 percent of workers, language models are expected to impact at least 50 percent of their tasks.

Sources: