Description
I'm the PI for a British Academy/Leverhulme SRG newly-funded project (co-applicants: Brian Ball and David Freeborn) titled ‘Reverse Engineering Prompts’. The project combines expertise from philosophy, particularly debates over what are, and what ought to be, apt question-answer pairs, with empirical and data science methods, to address the problem of how best to exploit the outputs of chatbots.
Education is increasingly being disrupted by AI-powered chatbots like ChatGPT. Though limited in various
ways, such chatbots are formidable tools to use alongside traditional pedagogical methods. To derive
maximum benefit from them, we must overcome the prompt engineering problem: how to formulate human
user input (i.e. prompts) to maximise the chances of getting meaningful output from chatbots. Our project aims
to lessen the severity of this problem by conducting case-based research on prompt engineering, focusing on
the subject of basic coding skills. The research consists of two parts. First, we seek to determine effective and
efficient ways we humans phrase queries to other humans. Second, we test the efficacy of those ways in the
context of human-chatbot interactions. The findings of this project will throw light on how to fruitfully exploit
chatbot technologies in pedagogical settings and are aimed at educators, researchers, developers,
policymakers, and the general public.