Fiber for Breakfast Week 25: Building AI-Ready Networks: The Next Evolution of Broadband

Fiber for Breakfast Week 25: Building AI-Ready Networks: The Next Evolution of Broadband 

For decades, broadband providers have built networks around people. People streaming video. People working from home. People browsing the internet, joining video calls, and connecting devices.  

On this week’s Fiber for Breakfast, Gary sat down with Robin Olds, Senior Sales Business Development Manager in the Broadband Program Office at Cisco to discuss what it means to build networks that are ready for the AI era. The conversation wasn’t about faster broadband. It was about how AI is changing the role of the network itself and creating demands that broadband networks weren’t originally designed to support.  

One of the most interesting parts of the discussion centered on how AI alters traffic patterns. Historically, networks were built around what engineers call north-south traffic—people accessing applications, content, and services from centralized locations. AI introduces something different. 

As Olds explained, “AI introduces more east-west traffic” as data moves between GPU clusters, cloud environments, edge locations, and data centers. That shift may sound technical, but its implications are significant. AI workloads don’t simply consume bandwidth. They create constant communication across the network, generating new demands for capacity, visibility, and automation. The challenge becomes even more apparent as AI agents begin to emerge.  

One observation from Olds stood out: “Humans click, but agents swarm.”  

A person might submit a search query, upload a file, or generate an image. An AI agent, by contrast, can operate continuously— gathering information, interacting with other systems, and completing tasks without waiting for a human to press the next button. Instead of occasional bursts of activity, networks may increasingly support workloads that are always running in the background.  

For broadband providers, the question isn’t simply whether more bandwidth will be needed. The question is how networks will be managed when demand becomes more persistent, more distributed, and increasingly automated.  

That reality is driving a broader shift in network design. Rather than focusing solely on adding capacity, providers are increasingly looking at observability, automation, and operational intelligence. Understanding where traffic is flowing, identifying congestion before customers notice it, and predicting potential service disruptions are becoming just as important as expanding network reach.  

Olds pointed to examples where AI is already helping operators identify issues before customers notice them, improving both network performance and customer experience without necessarily requiring large-scale infrastructure overhauls.  

The conversation also explored a question many providers are beginning to ask: if AI drives more traffic, where does the business opportunity come from? 

According to Olds, the answer may not be bandwidth alone. As businesses adopt AI applications, AI agents, and real-time collaboration tools, demand is growing for secure connectivity, low-latency performance, managed services, and specialized service-level agreements. Providers may also find opportunities around edge computing, private networking, and AI-enabled services that help customers deploy and manage AI workloads. 

That represents a meaningful shift from a market where broadband has often been viewed as a commodity. Providers that can deliver outcomes, performance, and reliability —not just speed —may have new ways to differentiate themselves. 

The broader takeaway is that AI may represent the next major inflection point for the broadband industry.  

The last two decades were about connecting people and communities. The next phase may be about supporting an economy increasingly powered by AI, automation, and machine-to-machine interaction.  

As AI becomes embedded in businesses, schools, healthcare systems, farms, and homes, the network is no longer just carrying traffic. It’s becoming part of the infrastructure that enables how work gets done.  

The providers that succeed won’t necessarily be the ones with the biggest networks. They’ll be the ones with networks that are adaptable, observable, and prepared for the demands AI will place on them.  

Click here to watch the full interview.