5GLs

5GLs, or fifth-generation languages, describe a class of programming languages intended to let programmers specify what a program should accomplish, rather than how to implement it. Such languages aim to abstract away procedural steps by relying on artificial intelligence to interpret requirements, search the solution space, and generate code or configurations. They are typically declarative or goal-oriented and often utilize constraint solving, logic programming, or machine-learning techniques, along with natural language interfaces or graphical specification tools.

The concept emerged in the late 20th century as part of a generational taxonomy of programming languages. It is largely retrospective and aspirational, with few concrete languages achieving broad adoption. 5GL research has been associated with natural language programming, expert systems, automated planning, and program synthesis efforts that attempt to translate user intent into working software.

Common characteristics include a focus on problem specification rather than algorithmic steps, high-level or domain-specific modeling, and reliance on AI components to perform search, optimization, or translation. 5GL-like systems often provide interactive or batch interfaces where the user supplies a description of the goal, constraints, or examples, and the system generates or adapts a solution.

In practice, no widely used 5GL language exists today. The term is frequently treated as historical or theoretical, while contemporary trends in AI-assisted programming, end-user programming, and program synthesis share the same aspirations. Critics point to issues of expressiveness, reliability, debugging, and performance as barriers to general adoption.