ChatGPT, an advanced language model developed by OpenAI, was used in a graduate-level cybersecurity policy class at Johns Hopkins University’s School of Advanced International Studies. Over five days, students interacted with the AI, posing questions and receiving detailed responses. They found ChatGPT to be a useful educational tool, providing comprehensive and accurate answers to complex queries.
Despite its strengths, ChatGPT had limitations. It sometimes produced incorrect or misleading responses, lacked contextual understanding, and was unable to engage in a back-and-forth conversation. It also exhibited biases present in its training data. For instance, ChatGPT consistently referred to the US as the world’s leading cyber power, reflecting a US-centric perspective.
The experiment demonstrated the potential of AI in education, but also highlighted the need for human oversight. AI tools can supplement human educators, but they cannot replace them. They can provide quick answers, but are unable to provide the nuanced understanding and contextual interpretation that human educators offer. The use of AI in education is a promising development, but it must be approached with caution and an understanding of its limitations.
Go to source article: https://alperovitch.sais.jhu.edu/five-days-in-class-with-chatgpt/