personal_asset
Choosing a Major Can Win Today and Still Lose by Graduation
Instead of predicting one hot job years ahead, combine foundations, current tools, real contexts, and multiple exits.
"With AI arriving, is computer science still worth studying?" The question is tempting because it appears to need only one major name as an answer.
A degree usually takes years. Today's job titles, tool popularity, and salary signals may change before graduation. A choice based only on what is hottest now uses today's price to carry a result that arrives years later.
A job is not one task
A saved essay discusses finance, mathematics, and computer science during a period of technological change. It argues that one generation's successful route should not automatically be assigned to the next. The essay also predicts that some occupations will disappear; I do not have enough evidence to repeat that conclusion. Moving the analysis from a major name to changing tasks, however, is useful.
Anthropic's 2026 Economic Index studies task-level AI use and success on work of different complexity using real Claude interactions. It can show that tasks are affected in different ways. It cannot prove that a complete occupation will disappear. Jobs contain many tasks, and a major is not identical to one job title.
Separate foundations, tools, and context
I would examine a major in three layers. The first is a durable foundation—mathematics, writing, experimentation, statistics, and understanding real problems. The second is a faster-changing tool—a programming language, model, or platform. The third is the ability to enter a real context: understand demand, judge evidence, collaborate across roles, and take responsibility for results.
Betting only on tools creates dependence on the current tool cycle. Discussing foundations alone can leave four years without contact with a real problem. A stronger combination lets foundations support continued learning, tools produce something early, and context test knowledge against reality.
Preserve more than one exit
This changes how I interpret the question "Does this major have a future?" Instead of guessing which job will be hottest in four years, I care whether a choice leaves multiple exits: graduate study, entry into several industries, movement into adjacent roles, or the ability to complete a small project independently. More exits reduce the cost of one prediction being wrong.
A major cannot remove uncertainty. The practical goal is not to predict the future perfectly, but to avoid putting every chip on today's loudest job title.
Sources: The Real Problem Is a Major Transformation, Not Whether One Major Can Still Be Studied; Anthropic Economic Index: New Building Blocks for AI Use.