Speaker 3
That's fascinating, thank you, thank you. So I've got a question here which is a little bit off-piece, but it goes back to the point you made about making your existing process a little bit easier, and of course we're at one of the potential risks if you can call it that of AI is there, there are fears out there that it will cause job losses, but it'd be good to hear about, you know, from where you're sitting, you know, is that something that you see happening, or is it a change of job function, you know, if you look at the business that you're operating in and the people that are conducting certain tasks at the moment, and could they be replaced by AI, or do you see other people coming into the business, is it a case of these people have to be retrained within financial services to use AI, it would be good to get some of your thoughts on that
Speaker 1
quickly. I think it's an interesting idea that people be up-skilled to use AI in their everyday roles, so within the product function, we get a lot of customer feedback, we get different ideas, we get suggestions, we get the push and pull of the market, and we can use AI in a walled garden sense of only our feedback to say, can you summarize this into the key points? So I've got seven, eight, four pages worth of feedback from various customers all around the same topic, can you summarize the exact pain points they're describing here, and then we can take that away, validate, sense check in, use it as a hypothesis. That's a great example where we're learning and educating ourselves to make our lives easier, to actually spend more time doing the creative process of our jobs. So for us, and certainly on my side of things, when we talk about AI, it's all around how can you spend the time on the real high value stuff, on the stuff that as a human you really bring to this organization, your experience, your knowledge, your expertise and intuition, that's the stuff I want more of my team to spend their time doing, less of their time on what is essentially admin, joining together data entry processing, where you can do it in a smarter way if you've got a bit more intelligence and a quick away and create those shortcuts, but fundamentally it still takes time and that's where I see people actually getting to spend less time on what might be the more laborious tasks that they have to do, freeing them up for the real value add stuff and actually giving people more time to innovate and think about what to do next, rather than being bogged down in the BAU, the business as usual of what we just need to do here and now. I don't know Dave from your side how you see
Speaker 2
it. Yeah, it's very similar. It's augmenting people. So one of the hardest parts about running an application isn't necessarily building it. That's tough, but looking after it and looking after it over a number of years, there will be lines and files with coding that we won't look at or touch for potentially five, six, seven years because it runs and it keeps running, but one day we'll have to come back and look at it. And code is incredibly explicit. It does what it does. It's necessarily binary, but it follows a set of rules and it either passes over or fails those rules. Human language is far more ambiguous. You have to turn that code into something a human can understand and quite a lot of time it takes hours to process something that even you might have written. It may have been three months ago, three years ago, and you've got to get that back into your kind of mental construct as to how that code works. AI can do that for us. It can help us explain how that code works if it's overly complex. We try not to make it complex, obviously. In order to turn it into something that's far more palatable and speeds up the process of understanding. Similarly, we can do it the other way as well. So you can explain what you're after and it can help you code those things. But again, converting an ambiguous set of or could be ambiguous set of instructions in a language like English into something far more explicit isn't necessarily something you could rely on AI doing perfectly. It might be able to get you started. We can be writing code and we can get suggestions as to what to write next based on what it's learned from the past, but it doesn't necessarily mean it's right every time. And we'd quite quickly find out that it wasn't if we just accepted what it gave us and pushed it out. I would definitely not be happy with me if we started. Thank you.