Banks and insurers have built-in hype detectors and pride themselves on their ability to differentiate between fads and genuine value-adding innovation. Over the past 15 months, their antennae have been tested as never before by the unrestrained fervor surrounding generative AI.
Consider this: since OpenAI unleashed ChatGPT in November 2022, we have tracked the launch of a further 623 foundation models that either address a specific business use case or provide a generic tool for tasks such as code testing. Meanwhile, the FinTech Innovation Lab New York, a fintech accelerator program that offers mentorship and advice from leading financial services firms, including JPMorgan Chase, BlackRock and Prudential, among others, is close to finalizing its next cohort. Ninety percent of the finalists – which are selected directly by the FS institutions – are focused on generative AI solutions.
One of the tests to distinguish between fool’s gold and the real deal is ‘show me the money’. There’s no shortage of that, from Microsoft’s investment of $13 billion in OpenAI and Meta’s announcement of its $33 billion commitment to the technology, to the recent news that SoftBank’s Masayoshi Son and OpenAI’s Sam Altman are both hoping to raise huge amounts – $100 billion and up to $7 trillion respectively – to accelerate the production of AI chips. If that sounds excessive – global computer chip sales last year amounted to a little over $500 billion.
Generative AI really is different from most of the major technological innovations the world has seen to date – could even the internet claim to have changed virtually every part of every business? Add to this universality the staggering pace at which the technology is advancing and diversifying, and even the greatest skeptic has to admit that we are witnessing an exceptional phenomenon.
Most financial services professionals agree. Throughout 2023, virtually every financial firm worked on identifying use cases and running proofs of concept. 2024 should be the year when they shift from experimental mode to the systematic, cross-enterprise pursuit of value.
There’s a lot more to this than just shifting gears and ramping up investment. To unlock the full value of generative AI, banks, insurers and capital markets firms will need to examine the fundamentals of their business. Most will have to upgrade their digital core because the technology cannot fire on all cylinders if the company doesn’t have a flexible IT architecture based on cloud and a modern data foundation.
All will need to rethink what work will be required and how it will be performed, and the impact of this on skills, roles and the organization itself. Given the pace and extent of change, traditionally conservative firms will have to radically transform their culture to become more fluid, adaptive and innovative. Their leaders will need a solid understanding of the potential of AI if they are to continue to influence the direction and evolution of their organizations.
These firms will have to bring their workers along for the journey and many will be uneasy. In a study we carried out late last year, we found that only 31% of financial services firms have comprehensive strategies in place to ensure positive worker outcomes and experiences with generative AI.
The reality is that if you succeed at capturing the full potential of generative AI you will, by definition, have reinvented your business. Few are ready for this. Workers are eager to be trained in generative AI, but not many financial services employers are truly meeting this demand.
This, of course, is good news for the front-runners, but all is not lost for those falling behind. The democratization of generative AI technology makes it easier for firms to catch up, provided they make the right strategic choices.
We all enjoyed playing with ChatGPT when it first came out, posing improbable challenges and marveling at its ability to swiftly deliver impressive responses. Implementing the technology to transform your operating model, customer experience and workforce is an altogether more demanding proposition. It demands a systematic strategy that reaches into every part of the value chain and a multi-year program that enables continuous reinvention.
A crucial part of this strategy is selecting from a growing array of increasingly sophisticated – and increasingly specialized – AI models. We talk about the need for a ‘model garden’ which allows you to nurture, try out and capitalize on a variety of models for different purposes. Another is understanding the vital role of the cloud. This includes provision of the necessary compute power, as well as making your data securely accessible to your preferred models.
There’s also no getting away from the fact that a great deal of preparatory work will be needed by most companies. Ultimately, if your data is not organized and secure, you can’t expect to make full use of it. If your culture doesn’t encourage curiosity and experimentation, you will always trail the industry leaders. If your people don’t have the necessary skills and willingness to change, you’ll struggle to scale the technology throughout the organization. And if you haven’t considered proper guardrails and responsible AI, you’re likely to risk introducing bias and losing the confidence of your customers.
The excitement around generative AI is palpable. Seldom have we faced a future where so many of the pipe dreams of today are likely to become tomorrow’s reality. However, the full value of this innovation won’t be realized if financial services firms don’t first put in place the essential building blocks.