My 19-year-old daughter is at college, but we video call a few times a week. Lately, these conversations drift toward AI and how my field of computer science is to blame for AI’s environmental impacts. “So, more data centers, then?” she’ll say, when I mention a new course or program in AI.
She’s not wrong to worry. And she’s in good company.
In Botetourt County, Virginia, not far from my home, Google has been preparing to break ground on a data center campus spanning 300 acres — one site for a company that consumed nearly 11 billion gallons of water across data centers and offices in 2025, according to its environmental reporting. At the same time, a drought has gripped our region, with municipalities asking residents to voluntarily conserve water.
This isn’t a Southwest Virginia problem, though Virginia is something of a bellwether. Virginia lawmakers passed the nation’s first tax on data center energy consumption this summer, and by some estimates as many as half of the data centers expected to open nationwide in 2026 are facing delays or cancellation due to permitting fights, power shortages and community pushback echoing the same unease my daughter voices during our calls.
None of this is new. A century ago, the rapid spread of electricity provoked real alarm — and real harm — as an unregulated grid killed dozens of people per year before safety codes caught up. We got a cleaner, safer grid because people insisted on it, not because the inventors designed for it. The interstate highway system is a harder story: built in haste, it tore communities apart, disproportionately slashing through Black neighborhoods, in ways our cities still haven’t recovered from. We didn’t organize in time. We didn’t insist. We just built.
Data centers deserve the reckoning electrification got and the interstate system never did: not retreat, but insistence on innovation that goes beyond the data center itself. Not “stop,” but “do this better, and not just for us but with us.”
This innovation starts with the energy that fuels a protest sign or a viral post. But even after that fire burns down to embers, the challenges behind data centers will still be there, glowing, waiting to be tended. AI is threading itself through nearly everything we do now, whether we notice or not. We still need data centers. The opportunity to build them responsibly is enormous, and there’s a seat at that table for innovators with every kind of expertise.
Innovations in economics are needed to build true cost analyses for the water and power a data center consumes, instead of hiding the cost in local utility bills.
Innovations in environmental science should measure the actual impacts of these facilities, not the PR version nor the panic-inducing one, and mitigate based on real numbers.
Innovations in materials science are needed to invent cooling systems and chips that generate less heat, so there’s less water and power spent chasing it away.
Innovations in computer science, the field that birthed AI in the first place, are needed to shrink it. Smaller, more efficient models can often achieve equivalent results without requiring a warehouse of GPUs to run.
Innovations in regional planning are needed to design data centers that give back as much as they draw, for instance, by using waste heat to warm nearby buildings.
Innovations in organizational leadership are necessary to avoid phenomena like token-maxxing and instead incentivize efficient, intentional use of AI that elevates and amplifies human cognition.
Innovations in education are needed to teach, learn and assess in an AI-saturated world lest we risk raising a generation that can prompt but can’t think.
And innovations in app design and usability are needed to build products that use less compute per scroll — including the apps a 19-year-old might doomscroll while texting her mother about data centers.
The people best positioned to lead these innovations are those just now choosing what to study, the rising generation, standing at the start rather than the middle of a career. They are also, not coincidentally, unlikely to be reading an op-ed. But boundaries between generations have flattened. Via social media, family dinners or a phone call on a Sunday night, most of us have a thread connecting us to this age group, whether we’re a parent, a teacher, a mentor or a boss. We can channel their energy rather than dismiss it, make room for them at tables that are only starting to exist, and back them as the leaders this next era of innovation needs.
The next time my daughter says, “So, more data centers, then?” I want a better answer. I want to tell her about the economist pricing what a data center truly costs a community, the chemist inventing a cooling system that needs less water, or the student — maybe my daughter herself — deciding these are opportunities worth her expertise and energy.
Christine Julien is a professor and head of the Department of Computer Science at Virginia Tech.
