Towards better health conversations: Research insights on a “wayfinding” AI agent based on Gemini
Google Research
Google researchers designed and tested a Wayfinding AI agent based on Gemini that asks clarifying questions to help users find personalized health information, rather than immediately answering questions like traditional chatbots. In a randomized study with 130 participants, the Wayfinding AI had conversations averaging 4.96 turns (compared to 3.29 for baseline Gemini) and users rated it significantly higher across helpfulness, relevance, goal understanding, and personalization dimensions. The conversational approach that proactively seeks context before providing answers appears more effective for guiding users through complex health information needs.
Why it matters
Generative AI
