Teaching Programming in the Age of Generative AI

Programming has been considered the "essence of informatics"[5] since the beginning of computing as a discipline. But programming in the fifties was very different from what we know today, and one of the goals (or dreams) throughout the history of programming language technology, has been "automatic programming" [9]-the ability to automatically generate computer code starting from a high(er)-level description of the specification of that code. What this meant changed over the years, from punching paper tape [3], to compiling high-level programming languages [14], to program synthesis [6]. Today, however, the availability of machine learning artefacts that produce high-level code from natural language specifications has completely changed the traditional meaning. To the extent that some computer scientists have begun to question the received wisdom that the core of their discipline is deeply rooted in programming [19]. If programming and programming languages are no longer the essence of computer science, this changes the epistemology of the discipline itself. Moreover, if we are at the end of programming, we should also change the curriculum, where programming, algorithms and programming languages play a major role. Several recent papers reviewed the performance of code generators based on large language models on typical CS1 problems (e.g., from the many possible citations [ 2, 7, 13]) and how machine learning impacts K-12 teaching (e.g., [16, 18]). Starting from this data, I will argue for the role of programming in the curriculum, distinguishing between programming taught as part of a holistic curriculum (as in some non-technical high schools) or as a vocational tool. I will use Simondon's notion of (closed and open) technical object [17] as an interpretive lens, together with Calvino's reflections on the availability of writing machines capable of replacing the poet and the author [4].