When Amazon employees discovered they were being penalized by an algorithm for “inactivity,” and managers started letting ChatGPT write their performance reviews, something began to feel off. The numbers add up, the systems are efficient — but the relationship feels hollow. We’re no longer being led by people; we’re being managed by machines. And somewhere deep inside us, something resists that. Not out of nostalgia, but out of psychology.

Leadership and followership are ancient solutions to the coordination problems of group life. Our ancestors had to hunt together, migrate together, wage war together, and decide who would steer the group. To survive, we evolved a finely tuned psychological compass that helped us recognize who was worth following. But equally important: We never surrendered our autonomy lightly. In traditional societies, people entrusted decision-making power to a leader only temporarily — usually someone with proven skill at a critical moment. During the hunt, or when someone had stolen something. Leadership was situational, personal, and above all: revocable.

With the agricultural revolution and the rise of states and corporations, we began giving away large parts of that autonomy to systems, rules, and institutions: from who decides whether we go to war to what time we must be in the office. And now AI stands poised as the next step in this long evolution of delegated autonomy. That raises a simple question: how much more are we willing to give up?

AI systems are brilliant at what they do. They analyze millions of data points, forecast outcomes, and optimize decisions with a precision no human can match. But let’s be honest: the AI we work with today is the worst it will ever be. We are in the Nokia 3310 phase of

© Psychology Today