Search Is a Hammer, Generative Chat Is a Loom; Beware the Technological Attribution Error

This paper explores the paradigm shift from traditional search engine usage to the emergent interaction with generative chatbots, specifically in the context of educational settings. We investigate how generative chatbots, exemplified by ChatGPT, represent a fundamentally different tool for inquiry and learning, resembling a loom for weaving thoughts, rather than a hammer for nailing down facts. Through a case study in an upper-division philosophy course on metaphysics, we analyze whether dialogues with ChatGPT can enhance the quality of argumentation and critical thinking among students. Our findings reveal that the conventional “one-shot” query approach, typical of search engine and encyclopedic interactions, significantly underutilizes the potential of chatbots. We argue for the need to reframe our interaction strategies with these AI tools, emphasizing the art of prompting and iterative refinement. The paper sheds light on the importance of adapting to the dialogical nature of generative chatbots as part of AI literacy and its implications for educational practices and cognitive development. Failure to do so may reflect a technological attribution bias.