Fintech Insider Podcast by 11:FS
00:04:17 · Kate Moody
[00:04:17] S0: Yeah. I think it's, it's been such an interesting one to see how all the different people rolling it out are approaching it. I mean, I guess from a Plum specific perspective, how has this changed what you're doing with your customers versus what you're able to do before, you know, with the financial guidance you were offering before?
00:04:32 · Daniel Wills
[00:04:32] S2: Of course. I think how it how it differs, it's filling a really big gap. The Financial Conduct Authority estimates around 23,000,000 people in The UK are underserved by the financial advice market. And that's, of course, a huge number of people making important money decisions, largely doing that on their own. And it's a confidence gap as much as a cost gap. Right? So picture someone with £23,000, for example, sitting in a current account. They know they should probably do something with it, but general guidance leaves them with a lot of homework, to do themselves. So comparing, various rates, and various providers, researching, working out for themselves. Of course, paying for full advice can feel like a big step, for that amount. So, in turn, making that relatively intimidating. So what usually happens is they end up doing nothing, right, sometimes for years on end. So what the regulator is trying to do here, I guess, is help people get the right level of support, and, doing that at an affordable price. And that's obviously where, Plum feel, we can offer a decent level of support at the same time. I suppose, with, you know, a lot of consumers facing household, large household budgets and bills, and they're under real pressure. So I think this kind of help, really matters more than ever.
00:05:52 · Kate Moody, Daniel Wills
[00:05:52] S0: Yeah. And you've emphasized a lot, I suppose, in your releases around this, how this is going to integrate with the AI functionality that you're building out into the app. Again, I suppose it would be good to understand a bit more about that, like how is, how have you used AI with targeted support and how are you kind of bringing that into the proposition? [00:06:10] S2: Absolutely. I think the easiest way to picture it is as a ladder. So today, our Plum AI assistant has really been used many, many times, as you say, more than 270,000 customers, we've helped with their with their finances. So, through that Plum AI assistant, people can set those financial goals and get step by step plans to help, achieve those goals, But that's guidance. So I think what really is the big difference here for us is we're going to start helping people more specifically through this targeted support, which is how we see it the next rung up. So rather than just laying out the options and leaving the decision entirely with, the consumer, we can offer that ready made suggestion for people in that situation based on their, circumstances and similar circumstances to other people in their group. So customers can get more help, practical steps, and can put them in a much better position.
00:06:10 · Daniel Wills, Kate Moody
[00:06:10] S2: Absolutely. I think the easiest way to picture it is as a ladder. So today, our Plum AI assistant has really been used many, many times, as you say, more than 270,000 customers, we've helped with their with their finances. So, through that Plum AI assistant, people can set those financial goals and get step by step plans to help, achieve those goals, But that's guidance. So I think what really is the big difference here for us is we're going to start helping people more specifically through this targeted support, which is how we see it the next rung up. So rather than just laying out the options and leaving the decision entirely with, the consumer, we can offer that ready made suggestion for people in that situation based on their, circumstances and similar circumstances to other people in their group. So customers can get more help, practical steps, and can put them in a much better position. [00:07:07] S0: Okay. No. No. It makes sense. Matthew, keen to bring you in. Obviously, you know, Target Support is a UK regulatory initiative. What does it look like to you on on your side of the pond? Are you seeing sort of similar problems in The US? And and kind of how is... How are companies approaching on on your side of pond?
00:07:24 · Matthew Goldman
[00:07:24] S3: I think absolutely the problem still exists. The, you know, the currency may be different, but the aspect of consumers have money, need advice, but don't maybe have millions of of dollars, you know, to pay a financial adviser is very much the same challenge. And I think with AI, everyone's just doing this for themselves. Right? They're asking their their chatbot of choice. And in The US, we've seen all of them, all... You know, ChatGPT, Claude, Perplexity investing in things like, you know, open banking links so that they can bring in that data and provide it to you. But what we don't know is if it's any good. Right? Or if they're misleading consumers. We know that AI can do that with other aspects like medicine and mental health, and we've seen the the downsides of that. I have not seen any indication that there's the kind of regulatory innovation here of this middle ground. And I think right now, we're still stuck. And officially, there is all these sites that give you financial advice and then have a disclaimer that says, but this is not advice because they don't wanna get, you know, sued or or fined or whatever. And then very expensive financial planning. And some of the big firms have tried something between where you get a financial plan for hundreds of dollars, but those are actually quite poorly done. Right? They're very generic. So I'm I'm really excited. I was really excited to learn about this from from The UK because I think a similar thing would be very helpful here. People want that advice, but they also don't want to make mistakes. And an interesting kind of last point is I started to see the marketing from large financial firms echo this. Like, there's a new Fidelity Investments campaign, which is basically like, I wanted help from a person along with AI. So it's kind of talking in the same language. And the the message is I can use Fidelity's great web tools, but I can also still talk to a person. And I think that's still where a lot of consumers want to go.
00:09:20 · Kate Moody, Brad Wiskirchen
[00:09:20] S0: Yeah. Absolutely. Brad, what was your what was your take on this? Can you take your perspective? [00:09:24] S4: You know, I I I think I think you hit it head on. First of all, explainable AI is difficult with LLMs and and neural nets in general. And that's because LLMs aren't necessarily deterministic. They give different answer... Different answers if asked the same question. So, you know, one of the big questions is how do you ensure AI generated solutions are reliable, explainable, and based on appropriate customer information. But, Dan, I'd I'd ask you, how are how are you approaching the balance that was just referenced between automation and human oversight as Plum develops a service? Then what happens when the AI system produces a recommendation that isn't appropriate for a particular customer?