At the Delphi Economic Forum 2026, one intervention dramatically cut through the wider discussion on geopolitics, climate and instability. It was during a special Thinking the Unthinkable brainstorming debate with the title Shock Of The New. 

Alvin Wang Graylin is an expert in Artificial Intelligence from Stanford’s AI and Digital Economy lab. He laid out a stark reality. That the limiting factor in the AI revolution is no longer technology, and the speed of change is something we are unprepared for. 

It came as new report published by London’s City Hall suggests the exposure is vast. More than two million jobs will be transformed or eliminated. Hundreds of thousands of workers are already in roles where AI can perform a large share of daily tasks. See: https://www.bbc.co.uk/news/articles/c3r3reylj9ro    

At Delphi Graylin argued that “The technology is no longer the limiter… it is already good enough to do pretty much everything that we want it to do.”.

This is a critical shift. For years, debate around artificial intelligence has focused on what the technology might be capable of in the future. That question, he argued, is now outdated. Instead, the real constraint sits inside organisations and how to deploy the technology. 

 “The real issue is the structural aspect of companies; the culture, the leadership, the incentive systems. That is what is slowing it down.” 

In other words, the barrier is no longer technical, it is now human.  

This creates a very different risk profile. Once organisations begin to adapt, once leadership, workflows and incentives align, adoption will not be gradual. It will be rapid. Graylin argues “When companies start to figure out what they need to change internally… the adoption is going to be very rapid.” That acceleration has direct consequences for the workforce. 

Graylin added “About 65% of companies are going to use displacement as a way of making up for productivity gains.” “19% said they will no longer hire… and 45% said they will lay off workers.” This is no longer a distant scenario. 

What makes this wave of disruption fundamentally different? Previous industrial revolutions displaced manual labour but created new forms of work. This time, the disruption is moving up the value chain. 

 “The technology is actually exceeding human capabilities in most cognitive tasks already.” 

Human thinking, analysis, and decision-making is no longer a differentiator. “Our brains were really the only thing that we had over machines… that’s no longer the case.” The implication is not just economic. It is societal. 

Mayor of London Sadiq Khan has warned that AI could “usher in a new era of mass unemployment” if left unmanaged, particularly in cities like London that rely heavily on white-collar work. 

The proposed solution? Graylin suggests shifting towards more human, service-based roles. But the challenge is the speed at which it now needs to happen. “We need to move from cognitive labour to more service labour… things that are human to human.” 

The problem is whether that transition can happen fast enough. Graylin suggests that “We actually need to create time… slow down the adoption process so that we can make time for that transition.” 

This is where the tension becomes more acute. AI development is accelerating. Competitive pressures push companies to adopt quickly. But society moves far more slowly. 

The result is a widening gap between capability and preparedness. Perhaps the most concerning insight was not about technology, but about leadership response. Graylin suggests that “There’s a lot of sticking heads in the sand right now.”  

Despite the scale of change, many decision-makers are still defaulting to familiar narratives, that disruption will balance out over time. But, as Graylin points out, those assumptions were based on previous transitions, ones where humans retained a clear advantage. 

The discussion at Delphi made one thing clear: this is no longer a future scenario to debate. It is a present reality to manage. As Graylin made clear, the risk is not just technological disruption, but societal instability if that transition is mishandled. Right now, the systems meant to manage that transition are still catching up.