Why most leaders are still getting AI wrong

All business leaders of the future are going to be living in an AI world. Already, generative AI (GenAI) is accelerating at a greater pace than many businesses can absorb. Billions are flowing into model development, the landscape for the technology itself is being captured by a handful of incumbents, and the policy environment – both globally and domestically – is fragmented. Yet inside most organisations, the real business value is yet to be felt. Are we in the midst of a hype bubble, or at the frontier of a huge global transformation?

Both, according to London Business School Professor Michael Jacobides and John Thornhill, a technology columnist for The Financial Times. The real AI divide won’t be between technologies, they assert – it will be between the organisations that can turn those technologies into new business models, and those that cannot.

Ahead of the launch of a two-day masterclass on GenAI for global leadership teams, developed by LBS and the Financial Times, Michael and John dig beyond the headlines to explore what’s really happening in startups, boardrooms and policy circles. Here are their key takeaways:

1. AI is both over-hyped and under-estimated.

“There is a huge amount of hype, and the problem with hype is that it does mobilise action, but at the same time, it sows confusion,” Michael asserts. Right now, for most businesses, that hype is running ahead of any real productivity gains. “We see senior executives and board members who think they need to do something about AI without really having concrete plans,” he observes.

But equally, the technology is advancing at breakneck speeds. In the long run, Michael and John both argue, we may be vastly underestimating what AI will do. According to Amara’s law (conceived by Roy Charles Amara, an American researcher, scientist and futurist) we tend to overestimate technology in the short term, and underestimate it in the long term – which is the likely case with AI too.

2. This is the first major technology wave funded before we knew what to do with it.

Unlike cloud or SAAS, GenAI research has been bankrolled by individuals or firms who made their fortunes in the previous tech cycle. Their focus was to see what could be achieved, without a plan for why they wanted to achieve it or what the economic benefits might look like. “It’s the first technology that has been funded for no other reason than people’s ambition,” Michael states.

3. The frontier for productivity is not AI writing your emails.

Writing emails better or reducing a few FTEs is not where the true value in these technologies lies. LBS research in this area shows that the leaders who are most excited by AI – and who are seeing results – are those who use it to invent new business models and value propositions. That’s what really turns the dial, Michael notes.

“All the debate tends to get conflated with chatbots,” John agrees, “and actually there are other sub-branches of AI developing very fast in different ways.”

4. Incumbents can win – but only a minority will.

A few incumbents in the physical world (such as Ping An and Majid Al Futtaim) are already taking their digital offering, coupling it with AI, and combining the two to create new products. They are able to do this because they already have three key ingredients – the digital infrastructure, the data, and a culture of experimentation. But they are in the minority.

5. Culture is the key determinant of winners and losers.

“Culture eats strategy for breakfast,” as the famous saying goes (attributed – perhaps erroneously – to Peter Drucker). “You can come up with a strategy for how to apply AI,” observes John, “but that might be obsolete within a year or two, because the models will have developed in such a way that your initial hunch is proven wrong.” Only cultures that can experiment, iterate quickly, reallocate resources and kill things fast – without triggering broad organisational fear – will win.

Far from being driven by the board, the true innovation often happens at the periphery of organisations, John notes. Leaders will have to develop ways to enable this innovation, while maintaining control of the downsides too.

6. “AI-native” insurgents now operate with drastically less capital and headcount.

European GenAI startups are scaling revenue faster, and on far leaner resources than previous tech waves. This makes the competitive dynamics brutally asymmetric: big firms are attacked from below by ultra-lean AI natives and from above by Big Tech platforms. “Either incumbents eventually figure it out and crush these startups,” argues John, “or the big model companies themselves will render many of them obsolete.”

7. Geopolitics will play a big part.

“A lot of people in Europe are very concerned about being wholly reliant on US technology,” John notes. “Are we becoming a vassal state of the US when it comes to technology?”

Meanwhile, we’re seeing extraordinarily rapid development of open-weight models in China that will likely fuel a further escalation of competitive geopolitical tensions. Europe (including the UK), then, faces a strategic question when it comes to AI which we’ve yet to answer – do we want sovereignty, or convenience?

The message from both Michael and John is clear: AI isn’t all about technology. And that technology is not a silver bullet. The winners of the GenAI era will not simply be the earliest adopters, or those shouting the loudest.

There is an AI bubble, argues Michael – and the circular investment loops between companies such as Nvidia and OpenAI will almost certainly trigger a correction. But that does not diminish the underlying shift. The advantage will go to organisations that can make sense of the technology in a practical way and turn it into new forms of value – not just marginal productivity tweaks. This requires a culture that allows them to rapidly learn, experiment and scale what works. The technology is real. The question is whether leaders are willing – or able to change at the pace it now demands.

Watch the conversation here: