When AI becomes a commodity, operational intelligence becomes the advantage

At Cisco’s recent AI Summit, technology investor Marc Andreessen offered an interesting perspective on what this shift means for the broader technology economy. He argued that AI could become the next major engine of productivity growth, potentially reversing the slower economic progress seen since the 1970s. At the same time, he describes open-source AI as an “asteroid strike” for proprietary models – dramatically reducing margins for those building the models themselves while accelerating innovation across the rest of the industry.

As Andreessen puts it, we may be entering a moment where software becomes significantly cheaper to build, which ultimately means we will see more software, more tools, and more SaaS platforms.

For enterprises, this shift has an important implication: as AI capabilities become widely accessible, the technology itself begins to commoditize. When everyone can access similar models and tools, competitive advantage moves elsewhere. Increasingly, it shifts to unique data and the ability to act effectively.

Nowhere is this more relevant than in the modern digital workplace. Organizations today rely heavily on collaboration platforms such as Microsoft Teams, alongside a wide range of communication and productivity tools that support everyday operations. These platforms generate vast amounts of operational data every day.

This data includes detailed call quality metrics, device performance information, provisioning activity, policy and configuration changes, service health signals, and patterns in how users interact with the platform. Individually, these signals may seem routine, but collectively they form a highly specific dataset describing how an organization’s collaboration environment actually operates.

The challenge is that much of this intelligence remains fragmented across multiple tools and administrative portals. IT teams often have visibility into individual components of the collaboration stack, but rarely a unified view of how the system behaves as a whole.

The result is a familiar operational pattern. Issues are often discovered only after users report them; troubleshooting can take longer than necessary; and valuable insights about service performance remain hidden across disconnected systems.

If AI is becoming widely available, the real opportunity lies in what organizations feed into it. Generic AI models can provide powerful capabilities, but they are most effective when paired with proprietary operational data – the signals generated inside an organization’s own systems and workflows.

Within large collaboration environments, this operational data reveals valuable patterns. It highlights the root causes behind recurring support tickets, uncovers early indicators of call quality degradation, identifies configuration inconsistencies across users or locations, and reveals trends in how collaboration services are being consumed across the business.

However, collecting data alone does not create value. The real advantage comes when organizations can correlate signals, extract meaningful insights, and act on them quickly.