Several weeks ago, we interviewed physics teacher Michael Carrigan to ask a deceptively simple question: how is artificial intelligence changing education? Here’s what he had to say about the AI classroom of the future.
After interviewing Mr. Carrigan, we decided to ask Claude.ai, a large language model built by AI safety and research company Anthropic, to write a 400-word blog post summarizing our interview. We gave Claude.ai an unedited transcript of our entire 15-minute interview with Mr. Carrigan, and it gave us the following blog post in about five seconds.
As artificial intelligence advances at a blistering pace, teachers like Mike Carrigan are considering how to adapt their classrooms and teaching methods. In a recent interview, Carrigan reflected on the emergence of powerful AI systems like ChatGPT and their potential impact on education.
Though these technologies are still new, their capabilities have captured the imagination of both students and teachers. Carrigan first learned about large language models like ChatGPT from media coverage, and he immediately recognized their potential benefits and risks in the classroom. Students could use AI to cheat on homework or generate entire essays and projects with little effort. This will likely render many traditional take-home assignments obsolete. However, Carrigan also sees opportunities for AI to enhance learning. As conversational agents, chatbots could tutor students one-on-one and provide customized explanations and practice. Teachers might also use AI tools to develop innovative lessons, assignments, and assessments more efficiently.
To adapt, Carrigan expects a continued shift toward flipped classrooms, where students learn concepts at home and practice applying skills in the classroom. This allows teachers to supervise students as they do the core learning work, rather than just assessing the end product. Teachers will also have to help students learn how to properly interact with AI, framing prompts and questions in ways that yield meaningful responses. Rather than banning AI, educators can teach students to use it as a constructive tool.
The future remains uncertain, but AI appears poised to radically reshape education. With thoughtful integration guided by teachers' wisdom, AI could make learning more engaging, personalized, and effective for students. But leaving it unchecked also risks undermining the integrity of learning assessments and student work. By acknowledging both the benefits and dangers, educators like Carrigan hope to create AI-inclusive classrooms that empower students while upholding the values of meaningful education.
[Editor’s Note] Teachers at Holderness School and other schools across the world are working to find new and inventive ways to use AI to improve teaching and learning. To be clear, the Holderness School Student Handbook stresses that AI should only be used by students at the direction of their teachers. Her is our current AI policy:
At Holderness School, we expect students to complete their own work. In certain situations and at the discretion of faculty or staff, students may be permitted to use A.I. tools. However, unless such permission has been granted, students should not submit work created by an A.I. system, including chatbots such as ChatGPT or Bard. The School realizes that the A.I. revolution will bring big changes to schools and businesses in the near future, and we are evaluating our practices and policies accordingly as a result. At this time, we remain committed to teaching our students how to write effectively and think critically using proven educational methods. A.I. might evolve into a helpful tool we can utilize effectively in teaching the craft of writing, but it should only be used under the guidance of a teacher while this technology is in its infancy and we assess how it evolves.