We interviewed people who graduated from college in the past year about how AI entered their education, their work, and their hopes for the future. The interviewer was AI, too.
Their feelings are not a simple measure of optimism or fear. They change with one question:
Does AI help people do more—or make people matter less?
AI arrived mid-degree
They entered college before generative AI was part of everyday life. By graduation, it was helping them learn, code, search for work, and make decisions. Adoption happened quickly—and not always deliberately.
Commencement
By graduation, AI had moved from classroom policy to a symbol of the future waiting outside campus. Speakers framed it as both a threat to the job market and a tool graduates would need to embrace. The crowd did not always agree.
Where AI belongs
The share who felt positive or cautiously open to AI changed sharply by domain.
Science & medicine
79%
Office & knowledge work
78%
Art, music & film
14%
Source: coded responses from 14 AI-moderated interviews per domain.
The future they want
Their clearest hopes are concrete: faster medical discovery, better tools for learning, and less time spent on repetitive work. Optimism grows when AI expands human possibility instead of narrowing it.
Entering the workforce
AI is not a distant labor-market forecast for this group. It is already changing which fields feel viable, how applications get written, and whether an individual voice can still be heard inside a system optimized for volume.
The human core
The boundary appears wherever identity and empathy matter most. Creative work should carry lived experience. Care, teaching, and understanding require something beyond a plausible response: the reality of being human.
The physical footprint
Their environmental concern is not abstract. It is water, electricity bills, noise, and communities absorbing the cost of infrastructure they did not choose. The demand is not simply to stop building, but to make progress accountable to the people who bear it.
Privacy and control
The closer AI moves to everyday life, the more carefully they want its boundaries drawn. Personal data should not become an invisible price of participation, and useful autonomy should never become independence from human control.
Talking to AI
The interviewer became part of the finding. Some people felt freer speaking to a system that could not judge them. Others missed the active listening, memory, and empathy of a person. Both can be true.
The condition for trust
Across research, medicine, education, and public deployment, trust was conditional. They were willing to rely on powerful systems when someone remained responsible for checking the work—and accountable for what the tool does in the world.
Philosophy graduate
AI arrived mid-degree
01 / 29
What this asks of builders
The future they want is not one without AI.
It is one where people remain visible inside it.
Progress means more room to think, create, care, and choose. It also means making human oversight visible—and listening as closely to what this generation wants to protect as to what it wants technology to make possible.
Persona spoke with people who graduated from college in the past year about how AI entered their education, their careers, and their expectations for the future.
Every conversation was conducted by an AI interviewer. The voices on this page are shared anonymously, and the editorial synthesis was reviewed by people.
Keep listening