Arts students might think, “AI will replace STEM majors first—they’re the easiest to replace.” STEM students may beg to differ, but I believe there’s no inherent conflict between the two groups. Each has its own strengths and weaknesses.
STEM students have a deeper understanding of AI’s underlying principles, which gives them an edge if they pursue AI development. However, they have two vulnerabilities: if their programming skills aren’t strong enough, they risk being replaced by AI. Thus, they must evolve—become architects, senior developers, or experts in task planning and leadership. Their value will grow when they can guide AI rather than just execute tasks.
The nature of work is also changing. In the past, STEM professionals wrote code themselves; in the future, they’ll collaborate with AI, much like how I now “work” without coding—I communicate my goals to programmers, ensuring they understand my vision. This places a new premium on STEM students’ communication skills. Too often, even when we speak the same language, miscommunication occurs (“lost in translation”), leaving others unsure of our intent and hindering progress. This is why STEM students must strengthen their expressive abilities—a domain where arts students often excel.
Arts students, for instance, can already “hack” large language models (LLMs) by using persuasive, flowery language to guide them, effectively controlling the AI. Their strength lies in clear, compelling expression: if they can articulate their needs well, LLMs will follow their lead. In time, arts students might even take on roles akin to “AI hackers.” Of course, they must act quickly to learn basic programming concepts. With these foundations, they too can instruct LLMs to build intelligent agents, meaning agent development will no longer be exclusive to STEM.
Looking ahead, true success may hinge not on one’s major, but on cultivating versatility. What matters most is effort: engaging in social practice, honing interpersonal skills, empathy, communication, and leadership. Take courses across disciplines—STEM students should read humanities texts, and arts students should explore STEM lectures. Embrace cross-disciplinary learning and extracurricular projects. Today, opportunities abound: online open-source projects, entrepreneurship competitions, and more.
By the end of four years, what truly matters won’t be specialized knowledge, but comprehensive ability. This breadth of skills will ensure that even if graduates temporarily take entry-level jobs, they can adapt, identify emerging needs, and thrive—often outperforming those who rigidly stick to their “major.”