This article was produced in partnership with Netwealth
The exponential growth of AI means advice practices will miss out on a superstar team member if they are not working on integrating the technology into their processes, the Netwealth Accelerate Summit has heard.
Netwealth’s AdviceTech research found only 5 per cent of advice firms are using AI widely. There’s a further 44 per cent that are “dabbling”, piloting it or using chat pods.
Speaking at the summit in Melbourne, BlueRock chief information officer David Hanus said the difference between interactions with AI products like ChatGPT is likely to be at least 10 times or even 100 times better.
“When ChatGPT first came out it was the equivalent of a primary school student,” Hanus said.
“ChatGPT 4 came out as roughly a high school student. Now the next generation of models which is due to come out in the next three to six months is expected to be PhD level. And you can deploy these models 24/7 in your business.”
Hanus said the potential output from AI will lead to a “great reconfiguration” for all businesses, not just wealth firms.
“Imagine how you would have to rethink a basketball game if one of the players could now shoot from literally the defensive half of their court and get 80 to 90 per cent of those shots,” Hanus said.
“We’re all going to get capability within our business where you’re going to have superstar players available that can get 70 to 80 to 90 per cent of the time, outcomes and literally change the way the business operates.”
Dean Holmes, The Wealth Network co-founder who has spent 15 years as a practitioner, investor and business coach, said it’s important to provide the AI bot the context of question or task it is being asked of, to provide the most appropriate result.
Holmes said he used AI for three different streams – compliance, client services and sales – and each required to a different approach.
For example, typing in “generate a file note” will only give a “kindergarten response”, Holmes said, but adding the context of who it is meant to represent, like a compliance manager, it can generate a file note in those circumstances.
“With Copilot you can also do this strategy which is called ‘look like’ – so you can ask it to look like one of your templates that you already have saved on SharePoint and it’s going to format that meeting note in the context of a compliance manager,” Holmes said.
“From a compliance perspective that will give you one file note. The same transcript done different – is just asking to write a three-minute summary for a client. It’s the same transcript just two different questions that you’re asking of your AI in order to produce it.”
But far from utilising AI for advice from the get-go, Holmes’ beginnings with ChatGPT started by writing birthday cards or limericks, although he did not bring any examples to the session.
“What I was trying to do at the outset was just understand how the questions or the questions I had to ask it to get the right results first and foremost,” Holmes said.
“My journey started with ChatGPT because that was the one that was first available,” he said, noting his early application involved
“What I’m really to do is create small user cases. I’m not focused on trying to change my business overnight but if I can sharpen my knife and do something slightly faster each and every day by using their technology – that’s my current focus.”
BlueRock, which was named after pro-wrestler Dwayne ‘The Rock’ Johnson and Australian rules football club Carlton Blues, developed their own version of ChatGPT named Ask Dwayne.
Hanus said they had started with simple use cases but evolved over time to apply AI for more complex matters, like email summaries for various purposes.
“[To] roll them up into things like client profiles and then from there to be able to identify where lead opportunities were in a discussion,” Hanus said.
“If a client was speaking to a lead adviser…if in the course of that discussion that would be referred to another part of our business, we can start to spot those events within those conversations and quickly generate leads that could then be input into our CRM system into the future.”
The firm also used AI for transcriptions and file notes.
“The business started experimenting with that without necessarily the support of my team…and got some really good traction there for what tools worked and what didn’t,” Hanus said.
“When we started this [experimentation] there was a lot of apprehension in our business around using AI. The number of ChatGPT or Ask Dwayne screens around our office was maybe one in 30. You’d [now] struggle to walk around the office without seeing this tool open on the monitors.”
