Briefing: Is the AI hype cycle now approaching its ‘slope of enlightenment’?

As with many emerging technologies, the recent artificial intelligence (AI) rush has been fed on a healthy diet of hype.

Go to any conference – in the insurance sector or otherwise – and you are likely to be pitched on the myriad ways that AI can improve your business, its processes and your bottom line. 

Much of this chatter will even be true, but it is now much more common for those being pitched on the big picture of AI implementation to ask about specifics, use cases and how implementation works in practice. And that is no bad thing. 

Back in 1995, American consultancy firm Gartner noted that the cycle of hype around emerging technologies often looked very similar, leading it to propose a theory of technological adoption now known as the Gartner Hype Cycle.

In this theory, the five stages of adoption are, in order, the innovation trigger, the peak of inflated expectations, the trough of disillusionment, the slope of enlightenment and the plateau of productivity. 

AI maps neatly to this. There was the innovation trigger when ChatGPT was released to the public in November 2022, with the peak of inflated expectations following after that. 

The trough of disillusionment, meanwhile, is described as the stage where reality hits and firms realise that implementation is hard, with results potentially being underwhelming.

And while this could certainly describe the current moment that the insurance industry finds itself in with regards to AI, recent conversations and data do seem to suggest that the slope of enlightenment is in sight. 

Gartner describes this stage as when developers and users begin learning how to use the technology effectively, with realistic applications and second generation products emerging. Niche experiments in use cases also begin at this point, indicating possible long-term value. 

The here and now

In a comprehensive survey of the general insurance sector, due to be published in the coming weeks, IT and business consultancy Sollers provided further evidence that AI has reached the slope of enlightenment. 

In a preview of the report, shared exclusively with Insurance Times, Sollers showed that 51% of the surveyed 35 insurers had now incorporated AI into their company strategies, or intended to do this soon.

Experiments in use cases are also beginning to show value too, with the research revealing that insurers are focusing in on areas such as claims process automation, document and email data extraction and AI support for call centre agents. 

Speaking to Insurance Times, Sollers UK general insurance relationship manager Michał Litwiński explained: ”The main factor is cost and in UK general insurance, you have very severe price competition.

“Claims account for a relatively large part of the costs for an insurance company, so their AI journeys have focused first on the claims handling process.

“Claims are also relatively easy to split into micro processes that you can then use to build smaller proof of concepts with.” 

This focus on the practical upside of AI technology from UK insurance firms seems to indicate that the level of maturity around the technology has increased – a sure sign that the market is approaching the plateau of productivity, wherein mainstream adoption begins, technology standards emerge and hype is replaced by every day, normalised business use. 

Obviously, the applications and versions of AI are so varied and diffuse that it is not correct to generalise the technology as a whole as being at one point or another on the Gartner Hype Cycle. What can be said, however, is that some areas of insurance are already making effective use of the technology to improve processes and the customer experience.

Indeed, Sollers’ report showed that 97% of the surveyed insurance companies were already making regular use of at least one AI tool in their business. 

The Gartner Hype Cycle is an imperfect, subjective measure of progress, but it does provide a useful framework to think about the adoption of technology. 

The fact that AI is showing signs of climbing out of the trough of disillusionment is a positive for the insurance industry. What is left to be seen is just how far above sea level its plateau of productivity may sit.