AI is changing everything. Universities must teach Americans how to keep up
A Harvard Business School study reveals AI's "jagged frontier," showing how misusing AI tools can actually hurt performance more than not using them.
A century ago, when universities wanted to mold students into highly "literate" graduates, they exposed them to the greatest works of literature, assigned analytical essays, and held classroom debates.
As the internet age dawned a few decades ago, colleges recognized the need to update their curricula and teaching methods — to ensure that students didn't merely master the great books, but also acquired the digital literacy skills to succeed in the 21st-century workforce.
Now, another seismic shift is underway. Artificial intelligence is reshaping how people learn and work at every stage of life. This time, universities will not only have to update what and how they teach, but also how long they teach it. In a world where AI evolves faster than any four-year degree can keep pace, universities have a unique opportunity and a responsibility to redesign their offerings and programs to ensure their students — past, present, and future — are AI literate.
CHRISTIAN UNIVERSITY AIMS TO PRODUCE 'IRREPLACEABLE' GRADUATES AS AI RESHAPES HIGHER EDUCATION
Americans are adopting AI faster than any other technology in history. According to Pew Research Center, roughly half of U.S. adults now use AI in their everyday lives, up from one-third in 2024. People of all ages are turning to these tools for everything from financial advice to travel planning and healthcare information.
But there is a big difference between simply using AI and using it effectively.
THERE IS NO AI IN ARMAGEDDON. ARTIFICIAL INTELLIGENCE IS A TOOL, NOT AN APOCALYPSE
A landmark experiment by researchers at Harvard Business School and Boston Consulting Group illustrates this distinction. The study found that consultants using leading AI models produced higher-quality work and completed tasks faster — but only when they recognized the technology's strengths and limitations. When they relied on AI for tasks outside its areas of competence, their performance was actually worse than that of a control group of consultants who weren't allowed to use AI at all. The researchers described this uneven boundary as AI's "jagged frontier" — it is superhumanly competent at certain tasks, while surprisingly incompetent at others. As a result, AI can boost productivity in areas where we have expertise, but we must be wary that it can mislead us in areas where we don’t.
As the models become more and more capable, that frontier will continue to shift. But it will always exist. And understanding where that frontier is — and how to leverage the technology's strengths without turning a blind eye to its weaknesses — is the essence of AI literacy.
THE BEST DEFENSE AGAINST HIGH-TECH AI SCAMS IS LOW-TECH SKEPTICISM, EXPERT WARNS
And this literacy is increasingly important at any age. Recent college graduates may use AI to build software or analyze data. A mid-career manager may use it to automate routine work, summarize reports, or test new business ideas. A retiree may use it to navigate healthcare information, manage medications, or compare Medicare options.
Each of these tasks requires more than knowing how to open an app or coding harness. People must also know which questions to ask, how much confidence to place in the answers, and when independent verification or expert advice is required — no matter whether they're filling out tax forms, comparing loan terms, or reviewing legal documents.
Mastering those skills will require continuous learning, especially as AI becomes more capable and more deeply embedded in our lives. That is precisely where colleges and universities have an important role to play.
Universities already have the basic infrastructure to meet that need. After all, many institutions maintain relationships with alumni through continuing education, professional certificates, executive education, and lifelong learning programs.
Programs aimed at lifelong AI literacy — and, ultimately, AI fluency — belong alongside those offerings. These might take the form of short courses, stackable certificates, alumni refresher programs, community workshops, or employer partnerships.
For example, MIT offers professional courses in AI prompt engineering. And my organization recently launched a free Applied AI microcredential for alumni. The program helps graduates across industries keep up with AI in real-world contexts — covering everything from common AI terminology to AI bias.
Beyond career-focused instruction, universities could offer practical AI education for every stage of life. A new parent could learn how to evaluate AI-generated health advice for their child, while a retiree could take a workshop on spotting AI-powered financial scams.
In the years ahead, AI literacy will become as essential as digital literacy is today. Universities have long helped people adapt to transformative changes in society. In the AI era, higher education should continue to embrace that responsibility by becoming lifelong partners in AI education.