Waymo Experiments with Gemini AI Assistant for Enhanced Robotaxi Passenger Experience
Waymo is in the process of developing an in-car AI assistant, powered by Google’s Gemini technology, for its autonomous vehicles. This information was discovered within the Waymo mobile app’s code by researcher Jane Manchun Wong. The unreleased feature, revealed in a 1,200-line system prompt internally titled “Waymo Ride Assistant Meta-Prompt”, signifies a significant leap beyond basic chatbots, aiming to create a comprehensive passenger experience.
The Gemini-powered assistant is engineered to answer rider queries, manage in-cabin functions such as climate control, and provide reassurance when necessary. However, there are notable limitations—the system is unable to control volume, alter routes, adjust seats, or operate windows. When queried about these unavailable features, the bot is programmed to respond with “aspirational phrases” like “It’s not something I can do yet.” Intriguingly, the assistant maintains a clear distinction between itself as Gemini the AI bot and the autonomous driving technology known as the Waymo Driver.
Although the feature has not yet been released in public builds, the detailed system prompts reveal Waymo’s meticulous approach to AI integration. This includes guidance on handling questions about competitors such as Tesla and Cruise. This development places Waymo in direct competition with Tesla, which is integrating xAI’s Grok into its vehicles. However, the two assistants serve different purposes—Gemini is focused on pragmatic, ride-specific functions, while Grok is positioned as a conversational companion for longer interactions.
Source: TechCrunch