At the beginning of this month, the UK hosted the first-ever AI Safety Summit. Held at Bletchley Park – the base for second world war codebreakers – the summit brought together world leaders, technology executives, and computer scientists to discuss how nations and companies should work together to address the risks of AI. The summit made clear that the impact and potential threat of AI – particularly large language models like Open AI’s ChatGPT – must not be underestimated.
So, what does this mean for our consumer-facing sectors? What systems should businesses put in place? How to best embrace the opportunities presented by AI while simultaneously protecting against potential dangers? And what does this mean for the future of leadership and talent?
Earlier this week, we hosted an intimate roundtable dinner in partnership with Faculty AI to explore these questions. The evening brought together leaders from one of the most critical functions in the application of AI in business: finance.
Alongside John Gibson and Josh Muncke at Faculty AI, our group included Chief Financial Officers from leading supermarkets, major travel companies and global media groups, each committed to understanding how AI is going to drive change in their industry.
Our lively discussion was held under Chatham House rules, but here we have outlined the key themes which resonated most strongly throughout the evening.
1 – Generative AI represents a real, genuine, leap
AI is not new. The technology is at least as old as modern computing, and over the past decade it has become fundamentally embedded into our lives – from using Google Maps for our directions to receiving content recommendations from Netflix or Spotify. But it’s clear that the emergence of Generative AI like ChatGPT is a real and profound development. Large language models – which are trained on written language rather than numerical data – have propelled AI into new spheres, dramatically broadening the scope of what the technology can do. It is not, the group agreed, just hype.
“Large language models have propelled AI into new spheres, dramatically broadening the scope of what the technology can do.”
2 – Businesses need to brace themselves for the rapid pace of change
It is now accepted as fact that every business is a digital business. In just a few decades, we’ve seen wholesale digital transformation: filing cabinets have been replaced with databases, cheques with online bank transfers, and letters with emails. We can expect this level of mass adoption to take place with AI – but at a far greater pace.
Somewhat mind-blowingly, the language models being developed today are being trained on 100 million times more data than they were ten years ago. Given this rate of progression, it’s predicted that we’ll see full ‘AI transformation’ within one decade or less.
Indeed, what differentiates ‘AI transformation’ from ‘digital transformation’ is that the foundations have already been laid: most businesses have already gone through a data migration, and have the frameworks in place on which to integrate AI technology. In other words, the stage has already been set for an AI revolution.
3 – The role of the CFO: mitigating risk
Central to a CFO’s role is risk outlook, and AI presents a host of known and unknown risks. Much has been written about the potential dangers of AI – from bias to misinformation – and the most forward-thinking CFOs are making it a priority to understand the potential outcomes.
This is especially pressing today. In the years to come, we agreed, there will certainly be robust systems in place to help organisations use AI in an ethical, safe, and reputation-protecting way. Just as all businesses today have anti-virus software, in the future we’ll see anti-bias software. Similarly, we’ll see external auditing of AI being offered, in which management consultancies provide assurance that your technology is protected against risk. But for now, none of these frameworks are mature, and it’s down to leaders within businesses to take responsibility.
4 – The role of the CFO: establishing AI as a value-driver
It was fascinating to hear how different organisations are using AI to boost efficiency, drive growth and create new revenue channels. We spoke about IKEA, which has replaced more than 8,000 call centre staff with AI customer support agents. Rather than treating it as a saved cost, the company has re-trained the colleagues as remote interior design advisors, who now support a revenue stream worth €1.3bn. The CFO has a crucial role to play in this kind of decision-making, and in encouraging leaders to see AI as a central piece of growth strategy.
“The CFO has a crucial role to play in this kind of decision-making, and in encouraging leaders to see AI as a central piece of growth strategy.”
Moreover, private equity firms are increasingly considering the value of data assets as part of due diligence processes, and we can expect to see this expand to AI activity in the years ahead.
5 – The role of the CFO: identifying use-cases
With both of these core responsibilities in mind – mitigating risks and driving value – the CFO has a crucial part to play in deciding where and when AI is implemented. There are obvious considerations around risk: the dangers associated with using AI to produce marketing material are considerably less severe than using it to build an investor statement, for example. CFOs can also prove the value of AI by picking the right initial use cases. One CFO reflected that he wished that he had first used machine learning tools in areas that the executive committee were more engaged in. “It’s taken me years to prove the value of data in the business,” he said, “but I think if I’d used it to drive sales, I would have won hearts and minds faster.”
While the recent wave of innovation in generative AI presents exciting new opportunities, CFOs should not be distracted from the technology that has existed for years, and which is consistently underutilised – especially in forecasting and simulations of future demand.
In years to come, 2023 will be looked back on as an inflexion point in the development of modern AI. Indeed, just this week, the Cambridge Dictionary named ‘hallucinate’ – the term used for when AI produces false information – as its word of the year. The technology is set to dominate the governance, risk and commercial agendas, and companies must adapt to a new set of priorities. Today, too few organisations have leadership teams which truly grasp the topic, let alone non-executive Directors who can ask the right questions. Without the right people around the table, companies will fall behind. Bets are on for the first time that we get a brief for a CEO to lead a consumer-facing organisation with experience in AI.
Thirza.Danielson@thembsgroup.co.uk | Liana.Osborne@thembsgroup.co.uk | @TheMBSGroup