Vinton Cerf, one of the computer scientists widely credited with helping build the foundations of the modern internet, is set to retire from Google next week after more than 20 years at the company. Cerf, 83, has served since 2005 as a vice president and chief internet evangelist. His retirement was announced during the Open Frontier conference hosted by the Laude Institute, where UC Berkeley professor Dave Patterson recognized his long tenure and influence. Google had not publicly commented by publication time.
Cerf and fellow researcher Robert Kahn are credited with developing TCP/IP, the networking protocols that enabled different computer networks to communicate and became a core part of the internet's infrastructure. Work that began in the 1970s helped shape the global network in use today and later earned Cerf major honors including the Turing Award and the Presidential Medal of Freedom. During his years at Google, he was associated not only with technical leadership but also with promoting the openness and broad accessibility of the internet.
At the same conference, Cerf also spoke about artificial intelligence and argued that the rise of autonomous AI agents could push the technology industry toward more formal standards. He said systems from different developers interacting with one another would require interoperability and precision, and he expressed skepticism that natural language alone would be reliable enough for those exchanges. Cerf warned that ambiguity in ordinary language could create misunderstandings between software agents, making standardized protocols increasingly important.
That view emerged alongside a broader discussion about the direction of AI development. Some participants pointed to the concentration of advanced AI models within a small number of well-funded labs as a contrast with the decentralized design principles that helped the internet endure. Cerf argued that agent-based AI could eventually reverse some of that trend by creating pressure for shared standards. Supporters of that outlook see a continuation of his long-standing emphasis on open systems, while the discussion also underscored uncertainty over who might shape the rules governing the next phase of AI infrastructure.