From Transcription to Search: Pondering AI on the CCATP Podcast

On a recent episode of Chit Chat Across the Pond, Allison Sheridan and I kicked things off with a discussion of my efforts to measure AI transcription accuracy (see “Comparing Audio Transcription in Notes, Audio Hijack, and MacWhisper,” 28 February 2025).

Our conversation quickly evolved into exploring how AI chatbots are becoming effective search tools. While Perplexity pioneered AI-driven search and Arc Search’s Browse for Me offered an alternative approach, ChatGPT and Claude now incorporate live Web searches to provide synthesized, contextual responses. Drawing on UC Berkeley professor Allison Gopnik’s characterization of chatbots as a new kind of “cultural and social technology,” I contend that these tools represent a fundamental shift in how we extract and synthesize human knowledge—progressing beyond simple keyword matching to an interactive dialogue that helps refine our understanding of complex topics.

I’m working on a longer article in this space, but the podcast offers a sense of my current thinking about how AI-powered search tools could fundamentally transform the way we discover, understand, and utilize online information.

 

Read original article