To Espoo, then, and Nokia's HQ, where I have come to meet the company's CEO, Justin Hotard.
We meet in the customer center, where my host, Nokia's VP of corporate communications, Maria Vaismaa, has brought cinnamon buns she baked at home especially for the occasion. I am consequently jet-lagged but in a good mood.
I arrive bearing a concern, however.
Nokia has enjoyed remarkable success over the past year selling the optical and IP networking equipment that connects the enormous AI data centers now being built around the world, as well as the systems within them. Its third-quarter results, released the day before our meeting, showed sales to AI and cloud customers surging 123% year over year, while comparable operating profit jumped 35% to €605 million. Optical Networks sales rose 18%, while IP Networks grew 43%.
But success can be dangerous, particularly when it comes so quickly.
Was Nokia becoming too comfortable with the extraordinary growth of its AI infrastructure business, I mused between bites of pastry (golden-brown, pillowy and fragrant with toasted cinnamon, balancing caramelized sugar with enough restraint to avoid inducing coma)?
Was it at risk of resting on its laurels — enjoying the financial rewards of the data-center gold rush without preparing adequately for what comes next? (I pondered this question as I slipped another bun into my bag for later.)
Because there will be a next.
My view is that the greatest economic transformation from AI won't necessarily take place inside those vast computing factories. It will happen beyond them, as intelligence moves into industrial networks, manufacturing plants, transportation systems and autonomous machines.
I arrived armed with three questions for Hotard: Where will the next wave of growth come from? How will telecom operators make money from it? And who will control the infrastructure on which this new digital economy depends?
I needn't have worried.
Beyond the AI factory
Hotard sees Nokia as a connectivity infrastructure company riding the next great technological wave, following digital voice, mobile communications and the internet.
"The wave that we're in is the AI wave," he says.
AI and cloud companies are currently the largest investors in connectivity infrastructure, but Hotard points to opportunities beyond the hyperscalers: carriers repurposing network facilities as edge inference centers, continued broadband fiber expansion and connecting AI applications to enterprise data and conventional computing infrastructure.
I raise my concern about the debt financing America's AI data-center expansion. Hotard acknowledges higher borrowing costs but rejects the suggestion that a slowdown is inevitable.
"I don't think there's going to be a slowdown in data center," he says.
Even if investment moderates, he argues, connectivity demand would continue growing.
But it is when we turn to the physical world that Hotard becomes particularly animated.
"The network can actually sense the physical world," he tells me. "We're proving that already. We're doing it today."
This is the significance of AI-RAN, Nokia's approach to integrating AI computing with radio access infrastructure.
The network becomes something more than a communications pipe. It becomes a sensor, capable of detecting activity in the physical world and providing intelligence to machines operating within it.
For me, this is the moment when AI stops being a collection of highly intelligent mushrooms, kept in the dark and fed whatever information we choose to give them. Connect those systems to billions, eventually trillions, of sensors and they acquire something fundamentally different: the ability to perceive and act upon the world around them. Agentic AI with agency, if you like.
Robotics, drones, autonomous vehicles and industrial automation all become potential beneficiaries.
"And I think that's actually where AI is going to be even more transformational than what we're seeing today," Hotard says.
Can carriers finally make money?
Telecom operators spent billions deploying 5G networks, frequently without establishing a convincing business case beyond greater capacity and improved performance.
Could AI-RAN be different?
Hotard argues that improvements in spectral efficiency and operating costs can justify the infrastructure investment. But that's only the beginning.
"If you just walk in there and say, all I'm going to do is something a little bit better," he says, "it's fundamentally incrementalism."
The more interesting opportunity comes from general-purpose computing, sensing and location intelligence.
Consider autonomous vehicles. Today, each vehicle depends heavily on its own sensors to understand its surroundings. But what if the network could warn vehicles approaching an accident?
"Wouldn't it be nice if the network could tell any vehicle, hey, there's an incident in front of you, you should stop?" Hotard asks.
He also points to the difficulties autonomous vehicles encounter in snowy conditions. Network-based sensing could supplement their onboard systems.
"All of a sudden maybe we can actually handle snow," he suggests.
These are services telecom operators could potentially sell, rather than simply providing more capacity at greater expense.
AI-RAN therefore offers the prospect of both reducing network operating costs and creating new revenue streams. The latter remains a commercial proposition rather than an established market, but it is considerably more compelling than asking consumers to pay more for faster mobile broadband.
Hotard expects manufacturing, transportation and logistics to be early beneficiaries. Healthcare could ultimately be even more transformational, with robotic medical services reaching remote communities, although safety requirements will slow adoption.
One network, every orbit
Autonomous machines will also require persistent connectivity.
Hotard offers a mundane example: his Microsoft Teams calls sometimes disconnect when he leaves Wi-Fi and reconnects to another network.
For a person, that's annoying. For an autonomous vehicle deciding whether to turn or stop, it is potentially disastrous.
"So we need persistent connectivity," he says.
That helps explain Nokia's ambition to integrate terrestrial, Wi-Fi and satellite networks into a communications fabric that can be orchestrated dynamically, wherever a device happens to be.
The objective is not simply better mobile coverage, but connectivity reliable enough to support intelligent machines operating in the physical world.
Which leads us to my third question.
Sovereignty is about competitiveness
Operational sovereignty has become an almost unavoidable subject in my conversations with telecom executives. I tell Hotard that it arises unsolicited in the overwhelming majority of my recent interviews.
Why has it suddenly become so important?
"Sovereignty is something that we didn't probably appreciate until we realized we didn't have it," he replies.
Communications networks carry the livelihoods of individuals, businesses and entire nations. Their resilience has become inseparable from national security.
Hotard discusses Nokia's satellite work and an exploratory initiative with Iceye in Ukraine, where resilient communications are hardly an abstract concern.
But his most interesting observation comes when I suggest that sovereignty has become too complicated for any country to achieve without trusted technology partners.
"There is no place in the world today where sovereignty exists in isolation," he says.
"Sovereignty is about partnership. And ultimately sovereignty is about competitiveness."
He then delivers the line that, to my mind, gets to the heart of the issue:
"Because if you're not competitive, you're by definition not sovereign."
Sovereignty isn't necessarily about building everything yourself or excluding foreign technology. It is about retaining the ability to operate, compete and make independent decisions within an interconnected technological ecosystem.
For Nokia, trust becomes an asset every bit as important as its engineering capabilities or product portfolio.
The next Nokia
As our conversation draws to a close, I ask Hotard what he wants Nokia to become, and how he would like his performance to be judged a few years from now.
"We are back to being a growth company," he says, recalling his message to employees following the latest earnings announcement.
He wants Nokia to be an innovative technology company, a trusted partner that delivers on its commitments and an organization whose employees are energized by the opportunity ahead.
"This is a company that's made many pivots over its history," he says. "We're at the cusp of one. We're making one now."
There are formidable challenges ahead. Nokia must turn its AI-RAN vision into commercial reality, help operators establish sustainable new business models and compete in an AI infrastructure market whose economics remain uncertain.
But Hotard is attempting something considerably more ambitious than capitalizing on the data-center boom: repositioning Nokia as a foundational infrastructure provider for the next industrial revolution.
After we'd finished the interview, I asked Hotard a more personal question.
Did he feel any sense of relief, I wondered, at how much Nokia had accomplished since he took the helm?
"I don't really think that way," he replied.
He acknowledged that he was proud of the progress Nokia had made in such a short time. But there was little suggestion that he considered the job done.
And perhaps that tells us more about Nokia's future than any earnings report.
The company stands at a historic turning point, moving beyond its traditional identity as a telecom equipment incumbent toward becoming an infrastructure powerhouse for the emerging AI economy.
It appears that the way Justin Hotard thinks — always looking toward the next technological transformation rather than celebrating the last — is precisely the kind of thinking Nokia needs at this moment in its history.
Read more about AI on Fierce Network:
Opinion: AI-RAN could give telecom a second chance to own the AI economy
Nokia campus sale nears as Hotard bets on industrial AI
Report: The FNTV 2026 Global AI Economy Index
Ciena's Gary Smith: AI is forcing a global rebuild of network infrastructure
Stephen M. Saunders MBE is a communications analyst and USPTO-registered inventor examining how digital infrastructure — 5G, cloud and AI — is reshaping industry, power and society, as well as underpinning the emerging, ubiquitous global digital economy. As anchor of FNTV and a longtime industry insider, he focuses less on growth narratives and more on execution, risk and how hyperscale technology is distorting markets, governance and society at scale.
Opinion pieces from industry experts, analysts or our editorial staff do not represent the opinions of Fierce Network.
