February 26, 2025. This podcast is a recap of the SLPs Talk Tech meeting featuring a presentation by Dr. Darin Woolpert on using ChatGPT for language sample analysis (LSA). His study compared ChatGPT's accuracy to traditional LSA methods, revealing strengths in basic tasks like word counting but weaknesses in more complex analyses. The discussion emphasized the importance of human oversight to correct AI errors and biases. We also explore AI's potential in early language screening and discuss the ethical considerations and need for normative data. Future discussions may address voice AI, audio-based learning, and synthetic speech.

Send us a text