CC Biz Buzz is a column series that features insightful commentary from a faculty member of the Columbia College Robert W. Plaster School of Business.
By Justin Dijak
One of the most common questions I hear from students today is also one of the most economically interesting: “Why do I need to learn this if artificial intelligence can already do it?”
It is a reasonable question. Generative AI can summarize articles, write computer code, analyze spreadsheets and draft reports in seconds. If technology can perform tasks that once required considerable education and training, it is natural to wonder whether the value of that education is declining. From an economist’s perspective, however, the more useful question is not whether AI can complete a particular task, but how AI changes the value of human knowledge and skill.
Economist David Autor’s task-based framework provides a useful way to understand this distinction. Occupations are not single, indivisible activities; they are collections of tasks. Technologies can automate some tasks, complement others and leave many largely unchanged. Artificial intelligence extends this process into forms of cognitive work that were previously difficult to automate. The result is not necessarily the disappearance of occupations, but a change in their task composition and in the skills that workers use within them.
One important consequence of technological change is a shift in what is scarce. If AI can generate information, analysis and alternatives at very low cost, then producing more information may no longer be a primary constraint facing an organization. Attention and judgment become relatively more important. Someone must still determine which evidence matters, which assumptions are credible, which risks deserve attention and which recommendations should influence a decision.
Consider a financial analyst using AI to summarize earnings reports and generate forecasts. The cost of producing analysis may fall considerably, but the investment decision does not disappear. The analyst must still determine whether the assumptions underlying a forecast are reasonable, whether important risks have been overlooked and whether the conclusions deserve to influence the allocation of capital. As AI reduces the cost of generating information and alternatives, economic value shifts toward the human capacity to evaluate those alternatives and determine which deserve attention.
This has important implications for education. Gary Becker’s theory of human capital treats education as an investment in knowledge and skills that increase an individual’s productivity. Like any investment, however, its returns depend on the economic environment in which it is used. When technology changes production, it also changes the relative value of different forms of human capital and the education required to develop it.
The relevant economic relationship between AI and human expertise is not simply substitution but, depending on the task, could also be complementarity. AI may reduce the value of human labor devoted to some tasks while increasing the productivity of expertise applied to others. The technology becomes more productive when paired with someone who understands the underlying discipline.
Domain expertise may become particularly important as AI-generated content becomes easier to produce. When polished analysis can be generated cheaply, employers, clients and consumers may have greater difficulty distinguishing careful analysis from plausible but inaccurate output. The problem is one of information asymmetry: The quality of an answer may not be readily observable to the person receiving it. Knowledgeable professionals reduce that uncertainty by evaluating the analysis, explaining its limitations and accepting responsibility for how it is used. Knowing how to operate an AI tool is different from knowing whether its output can be trusted.
The educational implication is not that students should choose between technical preparation and broader intellectual development. They increasingly need both. Employer surveys continue to emphasize critical thinking, communication, problem solving and adaptability alongside growing demand for technical and AI-related skills. Disciplinary expertise provides the knowledge needed to evaluate AI; AI literacy allows students to use the technology productively; and broader capabilities such as judgment and adaptability help them apply both as technologies and workplaces change.
Artificial intelligence is changing the returns to education rather than eliminating its economic value. Students should not judge the value of learning by whether AI can perform a particular classroom task. They should ask whether their education is developing the expertise, judgment and adaptability required to use increasingly capable technologies productively.
AI is unlikely to be the last major technological transformation today’s students encounter. The most valuable educational investment may be one that prepares graduates not only for their first job, but for a lifetime of adapting to technologies that have not yet been invented.
Justin Dijak is an assistant professor in Columbia College’s Robert W. Plaster School of Business. He serves as program coordinator for the school’s Business Analytics undergraduate and graduate programs.
Take your next step
Are you or a family member interested in continuing your education at Columbia College? It’s not too late to enroll for in-seat or online classes for the fall session beginning Aug. 31.
