In Javascript + Open AI students learn about emerging AI technologies by using them.

Javascrpt + Open AI was offered in the Spring and Fall terms of 2024 in response to the rapidly changing state of the art in artifical intelligence, machine learning, and large language models. OpenAI released GPT-4 less than a year before the course began, and GPT-4o released during the class. These technologies, and their rapid progress, raise critical questions regarding creativity and ethics, automation and employment, privacy and security, bias and fairness, and regulation and governance. I believe that students are best equipped to engage with these questions when they have a practical understanding of the technologies themselves. This course was designed to help students build that understanding by using these technologies in their own projects.

The coursework in the class was squarely focused on hands-on, practical, and technical skill building. From the first weeks, students imagined and realized projects that used OpenAI's large language models via their public API. These models, like GPT-4o and o1, are the same ones that power ChatGPT and represent the state of the art in natural language processing. Students also learned to design, build, and deploy full-stack web applications using Deno, HTML, CSS, and Javascript.

Terminal App

This was the first long project. Students designed and built a terminal application integrating GPT. Terminal apps run on the command line, presenting a text-based interface. Terminal apps are a natural fit for LLM-based projects, but most students had limited familiarity with this kind of app, presenting a unique user experience design challenge. This assignment was supported with technical workshops, work-in-progress reviews, and a final critique.

Web App

This was the second long project and the course final. This project challenged students to design, build, and deploy a full-stack web application powered by an OpenAI LLM. This challenging project pushed them to develop a compelling concept, craft an innovative user experience, and write code across multiple programming languages.