Sam Altman is glad to have been wrong. The CEO of OpenAI admitted a few days ago that the labor apocalypse associated with artificial intelligence is progressing with much less force than he himself had predicted. He expected a faster destruction of entry-level administrative and professional positions, the so-called white collars. Why? Artificial intelligence is changing work. It is automating functions. It is reducing the need for some tasks. It is forcing many companies to review processes, teams, and profiles. But something important has been discovered: that the great gain in productivity appears when AI works with people, under human judgment, within organizations capable of redesigning their processes well. The reason is simple. Artificial intelligence and people make different mistakes. They are, in mathematical terms, orthogonal.An AI can summarize a hundred-page report in seconds, sort data, compare documents, write a first draft of a text, or detect things that a saturated or tired person overlooks. It has formidable resistance to repetitive tasks. It maintains concentration. It processes enormous volumes. It executes without getting tired. The human being brings a different kind of value. It detects context. It perceives nuance. It suspects when an impeccable answer sounds absurd. It understands that a technically correct solution can be inappropriate, offensive, risky, or useless. It applies common sense, experience, responsibility, and intuition.This is where we find complementarity. AI is very good where people get tired. People are decisive where AI loses touch with reality. When both work together, the result surpasses either of them separately in many professional tasks. Professional work is not just a sum of tasks or operations. A lawyer does not just draft clauses; a doctor does not just apply protocols; a journalist does not just write sentences; a manager does not just process information. Professionals mix technique, trust, judgment, relationship, memory, and decision-making. This combination has proven to be more resilient than previously thought.The company that only sees AI as a payroll-cutting machine will miss out on an essential part of the opportunity. The company that integrates it as a lever to free up time, improve analysis, accelerate tasks, and elevate the judgment of its teams will have a much greater advantage. Altman's mistake says a lot about human work. It is more complex, contextual, and difficult to replace than many had imagined. Expressed as an algorithm, it would be: AI + Human > AI.
