Galloway to young people: AI doesn't take your job, the person who knows how to use it does

By Zendoric · September 5, 2026.

Scott Galloway, marketing professor at New York University (NYU Stern) and one of the most followed tech commentators in the English-speaking world, spoke before thousands of young people at the Auditorio Nacional in Mexico City, as part of the México Siglo XXI 2026 program organized by Fundación Telmex-Telcel. His central message, according to local coverage: artificial intelligence will not be what directly destroys young people's jobs; the real risk is ending up on the wrong side of those who do learn to use it.

Galloway urged the scholarship students to strengthen critical thinking, analytical ability, creativity and their own judgment, and warned that technical mastery of AI tools, on its own, is not enough. He also questioned the concentration of wealth in Big Tech and put forward entrepreneurship as an alternative, though with a caveat: build something meaningful beyond pure profitability. He closed with two ideas less tied to AI but recurring in his discourse: social media erodes real-life interaction and breeds dependence on external approval, and visible success online almost always hides years of prior failure.

None of this is new in Galloway's repertoire; he has spent years repeating variations of this script at universities and conferences. But the line that sums up his appearance at the Auditorio Nacional —AI does not replace jobs, it replaces those who do not know how to use it— coincides, almost word for word, with a thesis we have been making at Zendoric when analyzing AI's impact sector by sector: it is not that technology eliminates work, it is that it redefines which work matters and who stays inside the circle that knows how to operate it. In tech, banking, law or education, the pattern repeats: the function does not disappear, the base of those who perform it without leaning on AI narrows.

That is also where the argument is weak. Telling a young person to prepare and learn to use AI is sound individual advice, but it does not solve the structural problem: access to good training, to the best tools and to the time to experiment with them is not evenly distributed. If the advantage of "knowing how to use AI" concentrates among those who already had more educational resources, the net effect is more inequality, not less, however meritocratic the message sounds. Galloway senses this when he criticizes the concentration of wealth in Big Tech, but he does not fully connect the two ideas: the same market power he denounces is what decides, today, who has cheap access to the best models.

This is where the part we at Zendoric consider most relevant in the medium term comes in: the faster the cost of quality AI tools falls —and the open frontier, with models already competing head to head with closed ones, is moving in that direction—, the smaller the advantage of "knowing how to use it" will be compared with the advantage of knowing what to ask it and what to do with the answer. In other words: Galloway's advice about critical thinking and independent judgment will, over time, be more relevant than his advice about technical mastery of the tool, because the latter is getting cheaper and more democratized faster than human judgment can be formed.

The talk's other strong point, less quoted but more honest, is the warning about social media and isolation. It has no direct bearing on generative AI, but it does relate to the same underlying problem: technologies that promise connection and end up replacing real bonds with intermittent validation. It is a useful reminder at a moment when the public conversation about AI focuses on employment and forgets that the mental health of a hyperconnected generation is also a matter of technology policy.

Taken as a whole, Galloway's appearance contributes no new data and reveals nothing we did not already know, but it neatly sums up a consensus taking hold among serious analysts: the transition will be hard for those who do not adapt, and adapting is not just learning to use a chatbot, it is retaining the ability to decide what to do with what the machine produces. That distinction —between operating the tool and exercising judgment over it— will probably be the line separating this decade's winners from its losers, far more than access to the technology itself.