Manuscript received August 21, 2024; revised September 4, 2024; accepted October 28, 2024; published February 14, 2025
Abstract—With the growth of large-scale Artificial Intelligence (AI) models has come the growth of AI Application Programming Interfaces (AI APIs), where the model remains at the remote site, and is called remotely via an API. The goal of this work is to bring accessible coding for AI APIs into the classroom. This paper describes two key innovations. First, a new JavaScript coding site, “Ancient Brain”, designed for education with a focus on AI. And second, the building of an extensive set of open-source JavaScript AI API programs that can be run by students in the browser and can also be edited and modified by students in the browser (on Ancient Brain). A search of coding sites struggles to find any comparable collection of AI API JavaScript programs for students in existence anywhere. While AI APIs can be called in any programming language, this paper discusses the advantages of a JavaScript-based system. We discuss issues such as public versus hidden code, API key management, API response delays and CORS issues, with consideration of how to address these in an educational environment. Future work would be to build AI API exercises for secondary school students using this starter collection.
Keywords—Artificial Intelligence Application Programming Interfaces (AI APIs), coding sites, education, open source, code archives, JavaScript
Cite: Mark Humphrys, "Bringing AI APIs into the Classroom with a JavaScript Coding Site," International Journal of Information and Education Technology, vol. 15, no. 2, pp. 272-279, 2025.