Why Study CS? Thoughts on LLM-assisted software engineering (kmicinski.com)

Recent discussions surrounding the impact of large language models (LLMs) on software engineering highlight a significant shift in how programming is approached. Anthropic's CEO previously predicted that an AI assistant would soon generate 90% of code, a claim that is proving to be an optimistic forecast. The emergence of Claude Code has demonstrated that much of what is perceived as productive coding can now be automated, shifting the paradigm from manual coding to using AI as a powerful tool for building and iterating code more efficiently. This evolution prompts crucial questions for computer science students: as AI systems increasingly excel at generating code, what is the value of a traditional computer science education? While AI can handle many routine tasks, it also underscores the importance of deep understanding and critical thinking in software engineering. Students must not only leverage AI’s capabilities but also develop robust mental models and problem-solving skills that go beyond mere code generation to ensure their relevance and adaptability in a rapidly changing field. Emphasizing the balance between generation and verification, educators and students alike must recognize that true mastery lies in understanding complex ideas and frameworks that enable innovation in a world dominated by AI.