AI: The direction is obvious. The winners are not.
At a recent Cambridge Wireless event, Peter Whale and I found ourselves in a familiar conversation. AI dominates headlines, panels and board agendas. The capability is impressive. The trajectory feels inevitable. And yet, as we spoke, it became clear there was still a more useful question to explore.
Not how fast AI is advancing.
Not how powerful the models are becoming.
But where value will actually settle?
Where is the value?
Peter has spent much of his career taking technology from idea to market, used by millions of people around the world. His perspective is grounded in having seen the process of innovation move from promise to commercial reality numerous times and the recognition that the widespread adoption of new technology is often a long journey. What strikes him about the current moment is the gap between how widely AI is being used and whether enduring value is yet being created.
“Amongst individuals, uptake has been massive,” he said. “Hundreds of millions of people are using AI daily for both trivial and more purposeful pursuits. But in the corporate environment, there’s a yawning gulf between the promise of what AI can deliver and the reality of what is or isn’t happening.”
For sure, there are trillions of dollars being spent by a small concentration of big tech companies on data centres to train the latest AI models. Advances are rapid and impressive. Many businesses are offering “AI enhanced” versions of existing products. In the corporate world, things are moving more slowly. Organisations are experimenting. Proof of concepts are widespread. But evidence of sustained economic advantage or societal benefit is currently thin on the ground.
“The capability is there,” Peter observed. “But the real transformation is still to come. This has also been true of previous waves of disruption – be that the printing press, the steam engine or the Internet.”
That gap is not evidence of failure. It is a reminder that technology develops much faster than businesses can adapt or new business models can develop.
Jagged frontier
Recent labour market data reflects this uneven impact. Despite widespread concern, white-collar employment in the US has risen by around three million since ChatGPT was launched, while blue-collar employment has remained broadly flat. Even in areas seen as highly exposed, such as coding, employment has increased (Source: The Economist, Jan 2026).
The reason is partly what economists describe as a “jagged frontier”. AI performs some tasks exceptionally well, yet struggles with others in unpredictable ways. That inconsistency forces established organisations to tread carefully, be risk-averse and to test where it works and where it does not.
Electricity was commercially viable in the 1880s. It took decades before productivity gains appeared at scale. Factories had to be redesigned. Workflows rethought. Management structures reshaped. AI may follow a similar path.
“We’re right in the middle of something transformational,” Peter said. “But when you’re in the middle of it, it’s hard to see what the impact will look like in ten years, let alone fifty years or more.”
Visibility trap
Most leadership teams can see the direction of travel. AI will impact every organisation and is both a threat to existing businesses if ignored and a potential massive business advantage if embraced. AI will be embedded everywhere and ever more deeply into systems that in turn form the fabric of societies and economies. AI will automate large aspects of decision making. AI will alter how work is organised and which human skills will be most needed. And which may not be needed at all.
Predicting the overall trajectory is not too difficult, although getting the timing right is notoriously difficult. However, predicting who will capture value from such disruptive change is anything but obvious. History offers many warnings – Kodak and digital photography. Nokia, an early leader in smartphones. Sony knew digital music was real. They did not lack awareness, rather they feared cannibalisation of their existing profitable businesses. They all struggled to reorganise themselves to respond to disruptive change. Yet none of them thought they were being complacent at the time.
“It’s difficult to disrupt your own business when it’s still working,” Peter said. “You have to think differently inside a large organisation. That’s hard.”
In stable markets, listening closely to customers is good strategy. During periods of change, it risks anchoring you to what already works and blindsides you from major disruptions to the operation of a market.
“If you ask customers what they want,” Peter reflected, “they’ll often describe an improved version of today, because customers – just like us – find it hard to envisage the impact of disruptive technology.” Henry Ford famously said, “If I had asked people what they wanted, they would have said faster horses.”
Incremental improvement feels rational. It also feels safe. This is what we might call the Visibility Trap. When a technology is visible everywhere, leaders assume advantage will accrue to the most visible players or the earliest adopters. Advantage, though, often forms quietly at the edges outside existing structures and incentive systems.
Innovation in the shadows
In his years as a technology scout, Peter watched behaviour at those edges. Younger users. Unexpected uses. Emerging patterns. Converging technologies.
“You can often see clearly enough the roadmap for what becomes technically possible,” he said. “What you can’t see is who’s going to capture value and how they win.”
The first iPhone was not the best phone on the market. In several respects it was inferior to competitors. But it redefined the category. Disruption rarely appears fully formed. It looks incomplete, economically ambiguous and occasionally inferior. The current AI landscape carries similar characteristics. Giants are investing at extraordinary scale. Infrastructure, energy, compute. It resembles an arms race. Yet heavy investment does not guarantee durable advantage.
“In the early days of the web, the business models weren’t obvious,” Peter noted. “Some hadn’t even been invented.”
Infrastructure providers invested massive sums of money to build the rails, notably the mobile network operators to realise the promise of “the mobile internet”. But disproportionate value accrued to platforms that sat above them. These platforms were delivered by new businesses that hadn’t existed a few years earlier. Google, Facebook, etc…
These new businesses were funded by business models that were new and unproven in the early years, focussed on monetising user data. The uncomfortable implication for the commercialisation of AI is this – many of the organisations that will capture significant value from AI may not yet exist or are fledgling startups that today appear unlikely stars.
The real risk
For boards and investors, the risk is not technological ignorance. It is misallocated capital and misplaced confidence. Continuing to fine-tune a model that may already be shifting beneath you. The danger is not blindness but quiet overconfidence.
Most leadership teams are asking how quickly AI will arrive. Or how to extract short-term ROI from pilots. Those are understandable questions. But they may not be the most strategic ones.
The more consequential question is where value will settle as behaviour shifts and new business models emerge and establish. That requires bravery and lateral thinking under uncertainty, looking beyond existing customers and markets to as yet small “signals” in the noise of the marketplace. Predicting the future is always challenging, but it does require resisting the comfort of incrementalism. It requires understanding that what is commonly accepted truth about markets today does not determine tomorrow’s winners.
“The direction and pace of what is possible is becoming clearer,” Peter said. “What’s far less clear is who will be the household names of the future who will capture the value.”
AI is here and it’s not going away. But advantage will not belong to those who simply see the future coming. It will belong to those willing to seize the opportunities to be bold innovators and to be prepared to change business models, before market realities eventually become the new obvious.
Chris Hopwood
Founder