The Wake-Up Call They Didn't Hear
Brad Smith, president of Microsoft, recently published a 3,000-word essay responding to a striking cultural moment: graduating students at commencement ceremonies across the country booing AI. Not booing a speaker, not booing a policy, booing a technology. He called it a "powerful wake-up call for the tech sector."
Then spent three thousand words explaining why they should get on board anyway.
It is worth pausing on the math here. The class of 2026 entered university the same semester ChatGPT launched. They have spent four years watching every major tech company announce layoffs and AI infrastructure spending in the same earnings call. Goldman Sachs estimated in April that roughly 16,000 U.S. jobs are being lost to AI per month. The CEO of ServiceNow warned that graduate unemployment could hit the mid-30% range within a few years. A Microsoft AI executive said earlier this year that white-collar work would be fully automated within 12 to 18 months. These students did not boo out of ignorance. They booed because they understood the arithmetic before the commencement speaker finished the sentence.
Smith's response was to ask them to embrace it.
He compared their rejection of AI to a Princeton beer jacket controversy, where graduating seniors replaced an AI-generated design with one celebrating "100 percent cotton" and "100 percent human." Which is the kind of analogy that sounds measured until you realize how quickly it turns economic displacement into a preference for natural fibers. He acknowledged the job market is difficult. He called it a "perfect storm" for the class of 2026 without engaging the numbers driving the storm. His closing message: adapt. His policy concessions: none.
I wrote about the first version of this backlash in The Googlebook and the Bubble Nobody Knows They're In, after the University of Central Florida crowd booed the claim that AI was the next Industrial Revolution. The point then was not that students were rejecting technology. It was that the people selling AI kept misreading what the boos meant. The graduates were not saying they would never use AI. They were saying they did not want their futures handed to them as a finished product by people who could afford to call uncertainty opportunity.
Meanwhile, Satya Nadella appeared on the Hard Fork podcast in San Francisco and said something more candid. He acknowledged that the perception problem is real, that AI use inside Microsoft has its own excesses, and that the economics only matter if the tools create actual value. In related comments, he has argued that AI agents will need real identities, tools, and oversight, the kind of language normally reserved for workers.
Which would land differently if the company saying it were not simultaneously treating AI agents as employees at its own developer conference.
Here is what strikes me about both responses. Smith and Nadella are not wrong that AI will reshape work. They are not wrong that some of the anxiety is about perception as much as reality. Where they miss is the gap between acknowledging a wake-up call and doing anything in response to it. A wake-up call means the alarm went off. It does not mean you got up.
The generation booing at commencement is not confused about what AI is. They are the most AI-literate entering workforce in history. What they are protesting is the structure of the deal being offered to them, which is: absorb the disruption, trust the prosperity will come, and by the way, here is a 3,000-word essay about why that is actually good for you.
Smith is right that it is a wake-up call. The question he does not answer is who is supposed to wake up.