Opinion: AI is only a tool if we can choose not to use it

I recently read a Dartmouth Alumni Magazine article about the developer of a forestry app, excited by what AI could do for soil science and ecosystems. It was a familiar way of pushing corporate propaganda through a personal story: AI as the tool past innovators dreamed of — one that, in thoughtful hands, would solve humanity's problems.

That is not how most people are experiencing artificial intelligence. And we need to stop pretending otherwise.

In Fort Worth, Texas, teacher Ale Checka has spoken publicly about the state takeover of Fort Worth Independent School District, describing the scripted curriculum imposed on teachers as "AI slop" — material shaped without the care, expertise, or local knowledge children deserve. Her objection isn't that technology exists in the classroom. It's that educators are made to deliver material they didn't choose or help create.

For Latino students, that's especially dangerous. Schools serving low-income students and students of color have long faced unequal access to devices, internet and experienced teachers. Bringing generative AI into that landscape doesn't automatically close a gap — without serious investment, it widens one. A well-resourced student learns to use AI critically; one in an under-resourced school may just get canned lessons and pressure to accept an algorithm's answer instead of learning to reason independently. That isn't equal opportunity. It's a new layer of sorting.

The risks extend beyond the classroom. Data centers powering the AI boom are expanding across Texas and New Mexico, where water is a condition of daily life, not an abstract issue. The question isn't whether a company can technically build a data center — it's whether communities consented to the water and energy demands, and whether those costs were disclosed honestly.

We keep reaching for the calculator or hammer metaphor: a neutral instrument that becomes good or bad depending on who holds it. But AI is not a hammer. A hammer doesn't require enormous data, water, electricity and capital merely to exist, or to reorganize schools, workplaces and immigration enforcement around itself, telling every worker: use me, work faster or fall behind. AI does.

That's why the nuclear-age comparison matters — not because a chatbot is literally a bomb, but because it asks the right questions about power. The atom was split by brilliant people, an extraordinary achievement. It also killed hundreds of thousands in Hiroshima and Nagasaki and helped inaugurate decades of proxy wars, coups and dictatorships across Latin America. For people in Argentina, Brazil, Guatemala and El Salvador, American technological supremacy was never abstract — it had names, faces, missing family members, graves and the people who paid the price were rarely the people who made the decisions. That's the story we risk repeating with AI: progress sold as inevitable while its costs land on people with the least power to refuse it.

For Latino families, the most immediate danger is that AI is closing off economic mobility just as many of us have finally reached it. For a first-generation college graduate, an entry-level job is the culmination of a parent's years in a kitchen, on a construction crew, in a field, cleaning someone else's home so their child could study. It's where someone learns a profession, builds a résumé and starts imagining a home, health insurance, a future.

Now that rung is being removed. Stanford researchers found that workers ages 22 to 25 in the most AI-exposed occupations saw a 13% relative decline in employment after generative AI went mainstream. Our community understands the stakes of entry-level opportunity better than most, because we know what it took to get there. We're told this is "efficiency," that productivity gains benefit everyone. But productivity has risen for generations while stable housing, health care and retirement have grown harder to reach — the gains flow to owners, not workers. When a company eliminates the jobs first-generation graduates need to begin a career, it's telling families their sacrifices were expendable, closing a door and calling it innovation, so a billionaire can become a trillionaire.

The cruelty doesn't stop with workers AI can replace. It extends to workers AI still needs. Farmworkers, kitchen and elder-care workers, house cleaners, janitors and meat-processing workers are increasingly caught in digital surveillance and algorithmic management — cameras, location tracking, automated scheduling, demands to move faster. These systems don't liberate anyone from hard labor; they intensify it, turning a person into a stream of measurable output. Often these are the workers with the least protection — undocumented, part of mixed-status families, living with daily fear that a raid or checkpoint could tear their family apart. AI makes that surveillance cheaper, broader and faster.

The consequences aren't theoretical. In July, Houston construction worker Lorenzo Salgado Araujo was fatally shot by an ICE agent while driving his crew to a job. In April, Alejandro Cabrera Clemente, a 49-year-old Mexican citizen, died in ICE custody in Louisiana after being found unresponsive in his cell. These are two of 52 deaths tied to immigration enforcement under this administration — workers, fathers, brothers, asked to carry the country's labor while denied its security. Meanwhile, the federal government's attempt to restrict birthright citizenship put immigrants' children at the center of an assault on the Fourteenth Amendment; a divided Supreme Court preserved it, but the message was unmistakable — even children born here can be treated as conditional. Ask Romeo, a U.S.-born four-year-old with Stage 4 kidney cancer, removed to Honduras with his family, without his medication or medical records. This is what it looks like when technology, labor exploitation and immigration enforcement converge — and then AI companies call it progress.

Underneath the economic and surveillance harms is something quieter and just as dangerous: AI is erasing reality. A few years ago, Alicia Keys sang a wrong note during a Super Bowl halftime show. Millions heard it live. Try to find that footage now — you can't. Only the corrected, copyrighted version the NFL controls survives; what actually happened has quietly disappeared. It's a small example of something larger: AI systems consume the books and knowledge they train on and hand back a smoothed-over, corporate-controlled version, where the "official" version is whatever the platform chose to keep. We lose the ability to say, with confidence, this is what was true.

That erasure extends to everything. Life used to have carve-outs — a scenic overlook on the highway where you're allowed to pull off and look at the horizon. School, worship, unstructured time with people you love. AI doesn't leave room for those anymore. Working or job-hunting, the road never lets up — no scenic overlook, only the demand to keep driving.

No member of the working class needs another promise of progress that extracts from us while the benefits flow elsewhere. We deserve the right to say no — to the elimination of entry-level opportunity, to workplace surveillance, to AI-generated schooling imposed without teachers and parents, to data centers draining our water and to immigration systems that use technology to make cruelty more efficient. A full human life includes the right to refuse a technology, a system or a demand. If we lose that right, we lose more than jobs. We lose the future our parents sacrificed to make possible.