Sergey Brin says his commute involves talking to an unreleased Gemini AI model about building data centers

Sergey Brin, co-founder of Google, has been utilizing the company's AI technology during his daily commutes. After stepping down as president of Alphabet in 2019, Brin has re-engaged with Google's AI initiatives, particularly focusing on the Gemini project. At a recent panel at Stanford University, he shared insights into how he sharpens his mind on the road. When asked about his favorite podcasts, Brin revealed, "I often talk to Gemini Live in the car." He elaborated on his conversations with the AI, mentioning a specific query: "I want to develop a data center; I need to know how many hundreds of megawatts of this kind of power, what it will cost, and similar details." Gemini Live, a voice-enabled AI chatbot developed by Google, was first showcased in May 2024 at Google I/O and became publicly accessible in September 2024. However, Brin advised against using the current version, suggesting that a significantly improved iteration is on the horizon. He stated, "The publicly available version right now is not the good version. Give me a few weeks to actually ship what I have access to." Brin indicated that the current model of Gemini Live relies on an outdated framework, while recent enhancements in AI, particularly the launch of Gemini 3 in November, have set new standards in the industry. This latest model is touted for its advanced visual capabilities and is Google's most accurate to date. It was also integrated into Google's search features from day one, allowing users to access it easily by selecting "AI Mode." The impact of Gemini 3 has caught the attention of other AI developers, with Meta's team reportedly gaining access to Gemini 3 Pro. During the panel, Brin humorously acknowledged the self-promotional aspect of his comments but also expressed his fondness for podcasts, including the popular 'All In' show, where he has been a guest multiple times.