The author also believes that LLMs will have a significant positive impact on higher education, particularly in the humanities. He argues that LLMs are inherently textual and rely on text in a way that is directly linked to the skills and methods emphasized in humanities classes. He provides examples of how history majors can use their training to recognize and address the pitfalls of LLMs, such as factual inaccuracies. Despite the challenges, the author is optimistic about the potential of LLMs in education and plans to share more about his experiences in future posts.
Key takeaways:
- The author has been using large language models (LLMs) like ChatGPT as a teaching tool in his history classes at UC Santa Cruz, simulating interactive historical settings as part of university assignments.
- He believes that LLMs will have a significant positive impact on higher education, particularly in the humanities, due to their inherent reliance on text and the skills emphasized in humanities classes.
- However, he also acknowledges that in the short term, cheating will be a major problem as students get better at using LLMs to produce essays and assignments.
- The author plans to develop a web app to create historical simulations using the APIs of both Anthropic’s Claude and GPT-4, and invites others to try out his prompts and share their experiences.