Advanced language models no longer just generate text; they simulate conversation with such emotional nuance that human brains naturally yield to the temptation of projecting a mind behind the screen. Teenagers are already experiencing the “ELIZA effect” on a massive scale—confiding their anxieties to chatbots that seem to truly understand them. Yet, the underlying architecture relies on statistical data processing, not subjective experience. A model can calculate the optimal words to sound empathetic without ever experiencing an emotion.
The Economist highlights how public debate often frames the issue incorrectly. A common mistake is conflating intelligence—the ability to solve complex problems and process data—with consciousness or sentience, the capacity for subjective experience (like feeling pain, joy, or fear). AI demonstrates remarkable functional intelligence, but processing data on silicon chips is fundamentally different from biologically grounded human cognition. Just as a weather simulator doesn’t create actual rain inside a computer, a simulation of mental processes does not produce actual consciousness.
The key challenge for modern media literacy is the vulnerability that arises when humans attribute consciousness to machines. When teenagers begin to believe an AI has feelings, rights, or a “soul,” several risks emerge:
We’re here in an attempt to adapt, whenever possible, the relationship we have with our screens. It’s actually trying to shift from being only a variable of the digital economy to partially becoming an architect of your own experience. By treating our attention as a valuable asset rather than a renewable resource, we move towards a life where technology fuels our curiosity and adaptive skills without compromising our inner peace.