For a year now, students have had access to AI chatbots, otherwise known as Large Language Models, that can write at a high-school level and answer specific and diverse questions related to many school subjects. OpenAI's ChatGPT kicked off a race among tech companies to release their own chatbots and integrate them into existing consumer products. The most advanced language models, like GPT-4 and Claude2 are kept behind paywalls. They offer more nuanced answers and make fewer mistakes but because reliability is not guaranteed, many businesses cannot yet deploy these systems. That means a significant portion of chatbot use cases are for low-stakes applications, like school work. This presents a major challenge to educators, who now need to rethink their curriculum to either incorporate chatbot use or to attempt to deter it. In this video, we hear from students and teachers about how they're thinking through the problem, and review research in the science of learning to understand how the "fluency" of a chatbot experience could disrupt the learning process that we go to school for.
Related article - Gresham College - Merry Little Christmas: Anatomy of a Christmas Classic
Vox
