Conceptually Speaking
Conceptually Speaking
Rod Naquin Talks Large Language Models & Dialogic Computing
For this episode I'm joined by friend of the show Rod Naquin, a Louisiana based education leader and doctoral student whose research and writing explores the intersection between dialogue, learning, and large language models. Drawing on thinkers and theories from his research, Rod invites educators to stop viewing artificial intelligence as a completionist tool or sentient machine and instead regard it as a new form of dialogic computing. His articulation of LLMs challenges common perceptions of AI as merely a productivity tool, instead proposing a more interactive, discourse-driven approach to using language models in educational settings.
Rod offers concrete examples of how educators can apply this approach, emphasizing AI's potential as an analytical partner rather than an omniscient source. He advocates for a nuanced approach that leverages AI's capabilities while preserving essential human elements in the learning process. This episode provides valuable insights for educators, researchers, and anyone interested in the future of AI in education.
Key concepts explored in this episode:
- Dialogic Computing: Reframing AI interactions as collaborative dialogues rather than simple input-output exchanges
- Paradigm Shifts: Examining parallels between historical communication changes and current AI-driven transformations
- Post-Literacy: Considering the emergence of an era where AI-mediated text takes on speech-like qualities
- Dynamic Equilibrium: Balancing AI assistance with human comprehension and critical thinking
- Affordances and Constraints: Understanding the capabilities and limitations of AI in educational contexts
In addition to his research on dialogue, Rod has deep expertise in high-quality instructional materials and hosts The Science of Dialogue podcast. A husband and father of twins, he resides in Bayou Gauche. You can find Rod on Twitter/X.com as @rodjnaquin and read his writings at rodjnaquin.substack.com.