In the late 1990s, the Dot-Com Bubble jacked up the demand for deaf, dumb, and blind kid coders—the Java Wizards, as it were (JavaScript next, then Python). The US higher education system duly obliged, and stamped out droves of kid coders. Thus began a rapid decay in quality of the #ComputerScience curriculum at a typical college.
A quarter century later, an average #CS graduate working in #IT can no longer be distinguished from a Data Science graduate or a Coding Camp alum—the lot of them have become “code first, think never” Code Warriors. This coder class will be the first to be eradicated by the LLMs or their more-potent spawns.
If CS is to coexist with AI code generators, resist being swallowed head-first by DS, and survive as a proper field of science (as opposed to a mere vocation), an average undergraduate CS curriculum must reinstate the teaching of CS’s #mathematical foundations: Computability Theory, Type Theory, Abstract Algebra, Category Theory, Formal Languages, Complexity Theory, Algorithm Analysis, PDEs, etc.
Clearly, four years of undergraduate schooling is insufficient to turn an average college student into a practising theoretical computer scientist; after all, that is the province of graduate education—many more years thereof. But anyone who was granted a BSCS from an accredited US university must have been taught the basic reading skills, the fundamental mathematical skills, the essential lifelong learning habits, and the rough mental map of the further studies.
https://youtu.be/aOUqRZkR8dE?si=H56COiq-lHUlau2t
Germany, Stuttgart