Advanced technology will augment the insurance workplace and improve services without replacing industry workers, QBE Ventures expert Alex Taylor says.

The global head of emerging technology at QBE’s investment arm addressed almost 150 professionals at the Insurance News technology seminar in Sydney yesterday.

He says while generative artificial intelligence is expected to transform the industry, insurance employees need not feel threatened. 

“We are not going to see role replacement,” Mr Taylor said. “Augmentation is the word that I like here, because a lot of the things that you’re about to see make the jobs of underwriters, brokers and analysts working in insurance fundamentally easier. 

“As much as you get these bold predictions of people’s jobs going away, I don’t think this is actually true. I think AI is going to lead to less friction, more ability to serve our customers better, and the ability to understand the risks they bring to us in a much more effective way.

“Apart from being scary, this is a massive opportunity to mature our sector. To do this at scale is tremendously powerful.”

Mr Taylor – who said at age five his parents “sat me in front of a Commodore 64, and I never really left for the next decade” – explained how the development of technology has vastly outpaced predictions.

“We didn’t think things were going to take off as rapidly as they did. Over the past three to five years, we’ve seen an exponential pick-up ... We have seen one of the biggest breakthroughs in the past 30 years as a result of this uptick in capability. Things that we couldn’t even imagine would be possible are here and now.”

He says a modern iPhone is 200 times more powerful – and 80,000 times cheaper – than the Deep Blue computer from 1997 that beat Russian grandmaster Garry Kasparov at chess. 

Mr Taylor says AI will be able to analyse unlimited numbers of insurance documents, answering questions such as: what are the most frequent types of medication prescribed as a result of slip and fall accidents when somebody falls off a ladder in Utah?

“We can really understand what risks we’re already writing. Finding and summarising is a superpower, but also being able to do it extremely quickly and cheaply has extraordinary implications.

“Our problem is data, particularly unstructured data. It’s the documents that are attached to submissions, documents attached to claims, all these other things that we struggle with because they’re extremely voluminous and we can’t ingest it as fast as we might want to.”

New generative AI models are demonstrating reasoning and can “take understanding and learning from one segment and apply it arbitrarily to a new segment ... Essentially the same process you might go through training a junior member of staff to understand workers’ comp, for example, and then apply that knowledge to a risk they’ve never seen before. 

“We should all be paying attention to this. A lot of things that were the sole preserve of humanity until literally two years ago, we’re starting to see as emergent properties in these models.”