But when COVID hit soon after, he realized he may have made a mistake.

“That didn’t work because there were no more cafés,” he told students at Stanford University’s School of Engineering centennial celebration last week, Business Insider reported.

The transition from president of Google parent company Alphabet to a fortysomething retiree ended up not being as smooth as he imagined, and soon after he said he was “spiraling” and “kind of not being sharp” as he stepped away from busy corporate life.

Therefore, when Google began allowing small numbers of employees back into the office, Brin tagged along and put his efforts into what would become Google’s AI model, Gemini. Despite being the world’s fourth-richest man, with a net worth of $247 billion, retirement wasn’t for him, he said.

“To be able to have that technical creative outlet, I think that’s very rewarding,” Brin said. “If I’d stayed retired, I think that would’ve been a big mistake.”

By 2023, Brin was back to work in a big way, visiting the company’s office three to four times a week, the Wall Street Journal reported, working with researchers and holding weekly discussions with Google employees about new AI research. He also reportedly had a hand in some personnel decisions, like hiring. 

Skip forward to 2025, and Brin’s plans for a peaceful retirement of quiet study are out the window. In February, he made waves for an internal memo in which, despite Google’s three-day in-office policy, he recommended Google employees go into the company’s Mountain View, Calif., offices at least every weekday, and that 60 hours a week was the “sweet spot” of productivity.

Brin’s newfound efforts at work may have been necessary as OpenAI’s release of ChatGPT in 2022 caught the tech giant off guard, after it had led the field of AI research with DeepMind and Google Brain for years.

To be sure, Google for its part has been rising in the AI race. Analysts raved last month about Gemini 3, the company’s latest update to its LLM, and Google’s stock is up about 8% since its release. Meanwhile, OpenAI earlier this month declared a “code red,” its highest alert level, to improve ChatGPT. 

Brin added in the talk at Stanford that Google has an advantage in the AI arms race precisely because of the foundation it laid over years through its neural network research, its custom AI chips, and its data center infrastructure.

“Very few have that scale,” he said.