In one of the photos from Julia Freeland Fisher’s years on the College Prep varsity volleyball team, the whole squad is gathered on the gym floor in their jerseys, trophy in hand. It’s the kind of picture that looks like it’s about winning, and it is, a little. But what Julia remembers most about those four years isn’t the matches. It’s the older players. The juniors and seniors who already knew things she didn’t: how high school worked, what came next, what was possible. Closing that gap between grade levels is what she now recognizes, with the precision of someone who has spent two decades studying it, as genuine mentorship. Someone a few steps ahead, deciding to bring you along.

That instinct, to look at a relationship and ask what it’s actually doing and what it’s making possible, turned out to be a career. 

Julia is the director of education research at the Clayton Christensen Institute working at the intersection of technology and human connection studying something most education reformers overlook—not what students learn, but who they know. Today, as a researcher and writer focused on how AI is reshaping education and relationships, she finds herself asking a question that feels increasingly urgent: what do we lose when we let algorithms stand in for people?

Julia describes today’s AI tools as a “giant self-help machine.” They are powerful, convenient, and quietly working against some of the habits that matter most. Her concern isn’t that students are using AI, it’s that the easiest version of using AI lets them avoid something essential such as asking another person for help. Across multiple disciplines, the evidence is consistent — relationships are foundational to healthy development, access to opportunity, and long-term resilience. Mentors, sponsors, and near-peers are the mechanism by which young people learn what’s possible and find their way into those possibilities. When a chatbot becomes the path of least resistance, those relationships are easy to skip. The transaction may be cleaner, but something real isn’t being built that creates a paradox. Students can access more help in the short term, but fewer opportunities in the long term.

What society tends to miss, Julia argues, is that efficiency and connection aren’t interchangeable, and treating them as if they are carries real costs. When students turn to chatbots instead of people, they’re not just choosing a faster answer; they’re quietly eroding the social networks that underpin long-term development and opportunity. Her response to that problem is practical rather than nostalgic. Working with platforms like Roadtrip Nation, she’s developing what she calls “pro-social AI.” These are tools designed to strengthen ties with mentors and advisors rather than stand in for them, and to show young people that even the most accomplished professionals got where they are through help they rarely talk about publicly. The goal is to shift how students think about asking for help in the first place. It is not a sign of weakness, but rather a marker of progress, and a skill worth deliberately building.

Looking back at her time at College Prep, Julia sees seeds planted for the work she does now. Playing varsity volleyball all four years at high school gave her a sense of  belonging and, just as importantly, access to older students further along in their journey. Closing the gap between grade levels was what she now recognizes as genuine mentorship. She also names an influencial teacher, Mr. John Faggi. “He treated me like a writer before I was a writer.” That kind of belief which was extended before it’s been earned, by someone who has decided to see you a particular way is irreducibly human, and not a feature any model can replicate.

When asked for a concrete recommendation for College Prep, her response was striking. Alumni are one of education’s most underutilized assets, and schools should be building toward them deliberately. As AI takes over more entry-level work the mentorship embedded in those roles disappears with them. Her vision is specific. Every student should have dozens of substantive career conversations with alumni before graduation, not as a networking exercise but as something closer to infrastructure. This would be a way of ensuring that every student has adults in their corner who can show them what a path forward looks like.

“Human experience, with all its texture,” she says, “is still outside what AI can model. Schools should be building toward that edge.”

She closes with advice that’s simple on the surface and harder than it looks in practice: work hard and ask for help. Not as competing instincts, but as a pair that belongs together. In a moment when technology makes it easy to look self-sufficient that combination takes intention. Julia’s work is, in many ways, an argument for why it’s still worth making.