When IBM announced in 2023 that it would eliminate about 8,000 jobs as part of a sweeping automation effort, the move was seen as another sign of a future in which artificial intelligence would steadily replace human workers. But less than a year later, the company found itself in a very different situation. It had quietly hired thousands of people back.
The reversal surprised many observers. It also revealed something more complicated than the familiar storyline that pits humans against machines. IBM had automated large sections of routine administrative work, including 94 percent of its human resources tasks through an AI system called AskHR. The company saved an estimated $3.5 billion in productivity gains across more than 70 business divisions. But freeing up that much capacity created new demands elsewhere. Software engineers, sales teams and marketing specialists were suddenly needed to support the company’s shift toward higher value services that AI could not manage on its own.
The result was a workforce that grew rather than shrank.
The pattern reflects a broader shift across corporate America, where leaders are discovering that automation often reshapes work instead of eliminating it outright. Tasks disappear more quickly than jobs do. New responsibilities emerge just as quickly. In many cases, AI becomes less a tool for displacement and more a catalyst for redeployment.
Frank Palermo, the chief operating officer of NewRocket, said the assumption that AI wipes out entire job categories overlooks how organizations actually function. “The idea that AI is taking jobs misses the point entirely,” he said. “Automation is not replacing people, it is removing the repetitive work that keeps them from doing their best thinking. The organizations that succeed with AI are the ones that use it to restore purpose, not just productivity.”
IBM’s experience illustrates that point. After the introduction of AskHR, millions of employee interactions were routed through the automated system. In 2024 alone, it processed more than 11.5 million requests. Yet about 6 percent of issues still needed a human being, often because they involved a complex question or an emotionally sensitive situation. Those human interactions became more important, not less, as routine cases vanished. The workers handling them needed stronger problem solving skills and far better communication abilities than before.
Across industries, a similar dynamic is taking shape. Companies are automating predictable, repeatable functions. At the same time, they are discovering that new opportunities require roles that are more technical, more strategic and more interpersonal. Economists sometimes describe this as a skill shift rather than a job shift. The roles that remain often command higher pay and demand training that many workers do not yet have.
The World Economic Forum has estimated that more than 90 million jobs globally could disappear by 2030 due to automation. But it also projects the creation of entirely new categories of employment. IBM’s case suggests that the transition may not be a straight line. Eliminating positions frees up significant capital, which can then be invested in areas where human skills remain essential. Innovation, relationship management and long-horizon planning all fall into that category.
For workers, the transition raises practical questions about how to stay valuable in an economy where the nature of work shifts quickly. For corporate leaders, the challenge is to understand that automation alone does not guarantee efficiency. The companies that stumble are often those that replace workers without redesigning the broader system around them.
Palermo believes that the most successful organizations treat AI as a companion to human judgment rather than a substitute for it. “When people no longer spend their days buried in repetitive tasks, they rediscover ownership, creativity and pride in what they do,” he said. “That shift gives work meaning again, and that is the real transformation AI makes possible.”
Whether this pattern continues will depend on how employers approach the next wave of automation. IBM discovered that the real cost savings came not from cutting staff, but from using technology to give workers room to apply skills that machines cannot replicate. If more companies take that path, the future of work may be less about replacement and more about reconfiguration.
