The father of the internet says its biggest lessons are about to become AI's 3 biggest challenges

As AI races toward an agent-powered future, a legendary systems engineer says it reminds him of another technological revolution he knows quite well: the birth of the internet.

Vinton Cerf, the now 83-year-old who helped invent the networking protocols that underpin the modern web, has important guidance for the AI boom.

During a panel this week with computer scientists and Databricks cofounder Matei Zaharia at the Open Frontiers conference, Cerf highlighted several principles that could help determine whether AI reaches its full potential.

Cerf said the internet only became ubiquitous because no single company owned it, and everyone could use it. “In the case of internet, it only worked because it was going to be distributed to begin with,” said Cerf. “And so we left the rules very open.

We just said if you can find somebody to connect to and you follow the rules of the protocols, it should work.” Having common protocols meant that a university network in California, a government research lab, and a commercial internet service provider could all connect using the same technical language.

Cerf said AI is approaching the same inflection point at which the number of AI agents would demand “interoperability and standardization.” While we might speak to an AI agent in our native tongues, those vernaculars won’t be ideal for our agents to work together effectively.

Cerf said natural language may not be enough for agents to work together reliably or to create interfaces that enable developers to innovate without disrupting the entire system. “I don’t think English is going to be the best choice,” said Cerf. “There is ambiguity, and I think precision for inter-agent interaction is going to be very, very important.” Human languages often rely on context, inference, and multiple meanings for the same words.

Cerf said AI agents will need communication methods that minimize vagueness so each system can reliably interpret requests and commitments. “An agent really needs to be sure the other agent understands what it is that they just agreed to do together,” Cerf added.

Like the internet, Cerf said, transformative technologies aren’t stand-alone products but must become foundations that enable others to build. “A lot of the successes come from enabling technologies, whether it’s a platform or some other fundamental element that others can build on,” Cerf said.

Google, Amazon, Netflix, and millions of smaller developers built services on top of the same underlying internet infrastructure.

Cerf said the same principle applies to AI. “So if you really wanted to look for impact, think about things that enable other people to do things that they want to do,” Cerf added.