The advent of Artificial Intelligence is sparking a profound debate about the future of source code. Experts are questioning whether AI will lead to the evolution of programming languages, potentially creating new ones optimised solely for machine efficiency, or if it might even render traditional source code obsolete.
The core of this discussion revolves around the idea of AI generating code directly from prompts, bypassing human-readable high-level languages. This could lead to 'inscrutable black boxes' of code, where human developers might transition from writing code to refining prompts and regenerating software. While this approach could maintain modularity and testability, it raises significant questions about the role of programmers in a future without easily maintainable source c ode. The ultimate goal might be to move from prompt to an intermediate language that compilers or interpreters can directly utilise, questioning the necessity of high-level languages altogether.
Stephen Cass, a prominent voice in this conversation, acknowledges the potential downsides but also hypothesizes that programmers might still play a crucial role in defining interfaces, selecting algorithms, and making architectural decisions. The generated code would still need to pass tests and be able to explain its functionality. However, the question remains about which abstractions could disappear and what happens when AI is fully unleashed in code generation. It's not an immediate threat, but research into AI-first languages is emerging, with projects like Chris Lattner's Mojo being seen as a step towards an 'AI world' programming language, designed to leverage modern hardware capabilities.
Artículo generado mediante LaRebelionBOT
No hay comentarios:
Publicar un comentario