- The telecom industry still has no clear answer to the problem of monetizing 5G beyond basic connectivity
- AI may finally unlock the real promise of standalone networks — but it also risks creating dangerous ecosystem lock-ins around companies like Nvidia
- The West’s telecom strategy increasingly looks less like a security doctrine and more like an industrial competition with Huawei and China
I recently sat down with telecom analyst Will Townsend, VP and principal analyst at Moor Insights & Strategy, for a long discussion about AI, telecom infrastructure, 5G monetization and the increasingly strange battle to define the architecture of the next digital economy.
Townsend has been covering telecom and infrastructure for roughly three decades, and what made the conversation particularly interesting was that it avoided the usual AI evangelism and telecom buzzword soup. Instead, it exposed something much more uncomfortable: the industry still does not really know how to make money from the future it spent the last decade building.
“The challenge,” Townsend told me early in the discussion, “is monetizing beyond access.”
That single sentence may define the modern telecom industry.
For years, carriers were promised that 5G would unlock entirely new economic models through network slicing, private wireless, edge computing and AI-native infrastructure. Hundreds of billions were spent upgrading cores, deploying spectrum and modernizing networks. Yet most operators still generate the overwhelming majority of their revenue the same way they always did: charging for connectivity.
Townsend argued that the problem is not necessarily that the 5G vision was wrong. It is that the ecosystem itself remains fragmented and operationally immature.
“Only about 30% of operators globally have transitioned to standalone,” he noted. “That remains the challenge.”
That matters because standalone architecture was supposed to unlock the “real” 5G future — programmable infrastructure capable of supporting industrial automation, distributed AI and entirely new classes of services. Instead, much of the industry remains stuck somewhere between PowerPoint aspiration and proof-of-concept purgatory.
Yet AI may finally provide the missing catalyst.
Townsend pointed to deployments at T-Mobile, including AI-driven live translation services and early experimentation around AI-RAN, as signs that operators are beginning to integrate intelligence directly into network infrastructure.
But he also warned against the fantasy economics currently infecting parts of the AI infrastructure conversation.
“Nvidia would like operators to deploy GPUs on every cell tower,” he said. “But when you think about the operational cost and capital expense of doing that, it’s just not realistic.”
That tension — between AI ambition and operational reality — now sits at the center of telecom strategy.
And increasingly, the fear inside the industry is not simply technological irrelevance, but ecosystem lock-in.
Nvidia is no longer behaving merely like a chip company. It is attempting to define blueprints, orchestration frameworks and operational models for the emerging AI economy itself. Telecom operators, hyperscalers and enterprises all fear dependency on a single dominant AI stack while simultaneously needing the capabilities Nvidia currently provides.
At one point in our conversation, I remarked that the industry risks “making a cake from ingredients that are still sitting in separate containers in the kitchen.”
Townsend agreed.
That may actually be the best metaphor for the current state of digital infrastructure. Every layer of the future AI economy is evolving independently — cloud, AI models, radio infrastructure, optical transport, orchestration software, energy systems and edge computing — but nobody yet fully controls the integrated operational fabric.
The conversation became particularly interesting when we turned to China and the long-running campaign against Huawei.
I asked Townsend directly:
“Have you ever met a service provider that thought Chinese telecom vendors were using their equipment to spy on them?”
Townsend paused briefly before answering carefully.
“You know, I haven’t,” he replied. “I spent time with Huawei several years ago in Shenzhen and Shanghai, and their assertion was always that they provide the hardware, but the mobile operators themselves manage the network and have visibility into those things.”
He then added the caveat that governments obviously possess intelligence unavailable to the broader public.
But the exchange exposed the awkward reality sitting underneath much of the geopolitical telecom debate. There has never been publicly demonstrated evidence showing Huawei infrastructure being used for espionage inside carrier networks. That does not mean governments are wrong to worry about strategic dependency, supply chains or national resilience. Nations are entirely entitled to pursue industrial policy around critical infrastructure.
But increasingly, the battle looks less like a pure security doctrine and more like a struggle for technological and economic supremacy.
And while Western governments remain intensely focused on restricting Chinese vendors, there is still remarkably little evidence of a coherent strategy for building the full-stack AI infrastructure economy domestically.
That absence may become one of the defining strategic weaknesses of the next decade.
The most intellectually provocative part of the discussion, however, centered on 6G.
Unlike previous wireless generations, Townsend suggested that 6G may not primarily be about bandwidth at all. Instead, it could emerge as a software-defined capability layer sitting atop infrastructure already deployed for 5G.
That is a profoundly important distinction.
6G isn’t supposed to be a bandwidth upgrade. It’s a massive capability upgrade.
If that proves true, telecom networks stop being passive transport systems and instead become distributed sensing, orchestration and inference fabrics underpinning the entire AI economy.
At that point, telecom stops being telecom.
It becomes the nervous system of civilization itself.
FNTV will publish the video of the interview with Will Townsend next week.