- The next industrial revolution will emerge from machines, networks, automation, machine learning and human expertise — not chatbots
- Newmont abandoned Wi-Fi and deployed private 4G/5G because mission-critical autonomy demanded reliability and predictability
- Silicon Valley is building chatbots. Industry is building civilization
The next industrial revolution will not emerge from hyperscalers’ chatbot theater. It will emerge from the convergence of machines, networks, automation, machine learning and human expertise into systems capable of making the physical world programmable.
That may sound like a provocative statement at a moment when the technology industry appears consumed by large language models, copilots, AI assistants and ever-larger clusters of GPUs. Yet after spending several days at one of the largest open-pit mines on Earth, I am increasingly convinced that the most important AI story is unfolding far away from Silicon Valley and far beyond the walls of monolithic hyperscale data centers.
It is unfolding in places like the Australian outback.
The Boddington gold mine, owned by Newmont, stretches for kilometers across Western Australia. The scale is almost absurd. Giant haulage trucks move hundreds of tons of ore at a time. Drill rigs operate continuously. Control rooms oversee fleets of machinery spread across a landscape that resembles the surface of another planet. It is an environment defined by dust, heat, distance, danger and relentless operational demands.
What struck me most, however, was not the technology itself. It was the philosophy behind it.
Standing beside one of the mine's enormous autonomous haul trucks, I asked Boddington General Manager Chris Dark about artificial intelligence. Given the current industry obsession with AI, I expected a discussion about large language models, generative systems and digital assistants.
Instead, he looked at me and said: "It's stupid. AI doesn't matter until you've done everything else."
That answer reveals a great deal about where industrial digitalization is actually heading.
From the perspective of consumer technology, AI increasingly appears to be the story. From the perspective of industries that produce energy, food, raw materials, transportation and infrastructure, AI is often just one component of a much larger transformation. Before organizations can benefit from sophisticated intelligence, they need reliable systems, integrated operations, robust connectivity, disciplined processes and deterministic automation.
In other words, they need foundations.
Boddington learned this lesson the hard way.
Why Wi-Fi failed
Like many organizations embarking on industrial digitalization, Newmont initially attempted to use Wi-Fi as the connectivity platform for its automation ambitions. It did not work. In an office environment, Wi-Fi is usually good enough. In a giant open-pit mine filled with constantly moving vehicles, changing terrain and mission-critical operations, good enough is not good enough.
The mine ultimately replaced Wi-Fi with a private mobile network built on 4G and 5G technologies from Ericsson. That decision changed everything.
Private cellular connectivity provided the reliability, coverage, mobility and predictability necessary to support autonomous trucks, connected machinery, remote operations and the countless machine-to-machine interactions required to run a modern mine.
As Ericsson's Åsa Tamsons recently told me, "The biggest shift is that connectivity is no longer a utility. It becomes part of how operations run."
Boddington demonstrates exactly what she means. Connectivity ceased to be an IT service and became part of the production system itself. The network was no longer supporting the mine. In a very real sense, it had become part of the mine.
This is a pattern that is beginning to repeat across heavy industry.
Connectivity is key
Whether the environment is a port, a railway, a factory, a utility, an airport or a mine, the same realization keeps emerging: digital transformation is not primarily an AI project. Nor is it primarily a cloud project.
It is an operational transformation project.
The companies making the most progress are not starting with generative AI. They are starting by making the physical world observable, connected, automated and controllable.
That distinction matters because much of today's AI conversation is dominated by probabilistic systems designed to generate answers. Industrial systems operate under very different constraints. They must produce repeatable outcomes. They must meet safety requirements. They must operate predictably. They must work every time.
A hallucinating chatbot may be amusing. A hallucinating 200-ton autonomous Caterpillar truck is an entirely different proposition.
This is why deterministic machine learning has found a natural home in industrial environments while probabilistic large language models remain a complementary capability (an add-on) rather than the main event. Machine learning can optimize maintenance schedules, identify operational anomalies, improve resource utilization and help operators make better decisions, all without sacrificing the certainty of determinism.
Across Europe, Asia and Australia, many industrial organizations appear to understand this instinctively. Rather than replacing operational systems with LLM-based AI, they are embedding intelligence into operational systems. The result is often less glamorous than a chatbot demo, but it is vastly more consequential.
Boddington is therefore much more than a mining story.
It is a glimpse of what the next phase of industrialization looks like. A world in which machines, networks, automation, machine learning and human expertise operate as a single integrated system. A world in which connectivity becomes infrastructure, infrastructure becomes intelligence, and intelligence becomes operational capability. And a world in which America’s biggest differentiator, leadership in LLMs, looks increasingly out of sync.
For decades, communications networks were treated as something separate from industrial operations. Today, that distinction is collapsing. Connectivity is becoming part of the machine, part of the workflow and part of the means of production itself.
That may ultimately prove to be one of the defining technology stories of our time.
Silicon Valley is teaching machines to talk. Industry is teaching them to work.
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.