Something unusual happened at American commencement ceremonies this season: when speakers praised artificial intelligence, the graduates booed. Former Google chief executive Eric Schmidt was among those jeered for painting AI's impact in rosy terms to an audience about to compete with it for entry-level jobs. This week Microsoft President Brad Smith broke ranks to say the booing deserves a hearing. In a blog post, he called the pushback "a powerful wake-up call for the tech sector," writing that to executives pursuing a future where computers replace jobs, "the next generation has offered a compelling response: 'Not so fast.'"
The graduates have reason to be touchy. Anthropic's Dario Amodei warned last year that AI could wipe out half of all entry-level jobs. Microsoft's own AI chief, Mustafa Suleyman, predicted in February that computer work "will be fully automated by an AI within the next 12 to 18 months." Those forecasts were aimed at investors and policymakers, but the people who heard them most clearly were the ones applying for the jobs in question. Smith acknowledged that today's graduates face genuine automation of entry-level tasks, alongside hiring freezes at the very companies spending billions on AI infrastructure. His counterargument is that tech leaders habitually overestimate how fast new technology transforms society and "underestimate the capabilities of people."
New research published the same week suggests the discontent runs deeper than job anxiety. Workday's Human Connection Workplace Index, a survey of more than 2,150 workers, found AI delivering exactly what its sellers promise: 86 percent of employees report higher productivity, 62 percent report lower stress. But the same survey found a third of workers rarely have meaningful conversations with colleagues, and Gen Z employees report feeling disconnected at twelve times the rate of Gen X. One in five Gen Z respondents has taken time off work because of loneliness. Workday's chief impact officer, Carrie Varoquiers, put the tension plainly: as more questions, ideas, and even conflicts get routed through AI, "we risk losing the everyday human interactions that build trust, resilience, and a sense of connection."
That detail complicates the usual framing of the backlash. The youngest workers are not refusing the technology; they use it heavily and rate it well. What they are reporting is the cost of a workplace where the technology works as designed, and where the colleagues, mentors, and casual conversations that used to come with a first job have thinned out around it.
It is hard to miss the timing of the industry's softer tone. As CNBC notes, leaders who once warned freely of displacement, from Sam Altman to Palantir's Alex Karp, have pivoted in recent weeks to emphasising productivity and augmentation, just as OpenAI and Anthropic prepare for public listings and data centre projects face local opposition. Smith's post fits the gentler refrain: "people can use AI to make themselves better." The cynical reading is that the message changed because the audience now includes retail investors and town councils. The more generous reading is that the industry is discovering its future workforce and customer base was never sold on the story it had been telling. Either way, the booing worked. They heard it.