Ready for the Robots?
Artificial Intelligence (AI) has been a buzzword for quite some time. Banks are announcing AI-enabled programs, but how does that actually play out in financial services? In this episode, Eric speaks with Tracey Fieber and Gary Melling of Acquired Insights about AI adoption in North America, the prevalence of AI throughout the rest of the world, and the fact that, in Tracey’s words: “AI is here now.”
Tracey Fieber is the Chief Strategy Officer of Acquired Insights Inc., an AI company with extensive experience in the finance sector. She is also an author and speaker on the topics of AI, ML, and Business Transformation.
Gary Melling is the President & CEO of Acquired Insights, and also leads other startup-teams in the AI and ML space. He is a speaker and strategist, and has been published in numerous academic publications including Harvard journals.
Eric: 00:05 Welcome to the Finance Frontier. I’m your host Eric Hathaway. Today we’re exploring the topic of artificial intelligence and we have the pleasure of speaking with Tracey Fieber and Gary Melling. Tracey’s the Chief Strategy Officer of Acquired Insights Inc. an AI company with extensive experience in the finance sector. She’s also an author and speaker on the topics of AI, ML and business transformation. And Gary Melling is the President and CEO of Acquired Insights. He’s both a speaker and strategist and has been publishing numerous academic publications including Harvard journals. Welcome to both of you and thanks for joining us today.
Gary: 00:41 Thank you Eric.
Tracey: 00:41 Thank you Eric. Pleasure to be here.
Eric: 00:49 So you’ve been working at the intersection of AI and finance industry for quite some time, what are you seeing happen in relation to both fintech and large institutions?
Gary: 00:59 Sure Eric I’ll take that one. If we take a look at financial services most of the effort as far as what’s currently existing in the market, in artificial intelligence, finance services these days appears to be using structured data and credit scores. So what we’re seeing is a general trend where there are changes occurring real time as we speak. For the fintechs and the large institutions the problem is that they’re basically using the same measures and KPI’s that they’ve been using for about the last 30 years. Of all the data that’s currently available in business and financial services, basically anywhere, manufacturing, retail, etc, only 20% of it is actually contained in a structured data source so that means organizations by and large are missing out on the opportunity to access this other 80% of data that could be used for enhancing, refining, decision support, around a lot of issues as we mentioned like credit worthiness. So while most appear to be moving towards artificial intelligence and machine learning to try to get better resolution, better answers and so on, in many instances what they’re really doing is using AI/ML to get the same old information faster. And I consider this to be a lost opportunity.
Eric: 02:29 So are you seeing the same thing outside of the U.S.? What about the rest of the world?
Gary: 02:34 We’ve been taking particular note of what’s been going on in the financial services sector in South Korea, Singapore, Hong Kong. There are places where the adoption of artificial intelligence and machine learning is far and away more integrated, more adopted by the customers, by the citizens than what we’re seeing in North America.
Eric: 02:59 And as we see parts of the globe moving faster than the U.S., do you feel that this has to do with regulation, access to technology or other issues?
Tracey: 03:09 I think Eric that really the U.S. is not an early adopter when it comes to countries, looking at each country. With the U.S. and Canada being slower to adopt the other countries are much quicker. We’ve seen in financial industry the Mag strike on the back of a card, we started to see smart cards and different technology in tap and all sorts of things. In other countries far way and ahead we’re gonna start to see where the adoption rate is coming on faster and faster within U.S. and Canada.
Eric: 03:47 Absolutely and the adoption rate is extremely fast but there’s also this conversation about the hype. What are we seeing as far as broad use across the financial sector?
Tracey: 03:57 It’s not something that’s coming, it’s here now. So whether it’s the fintech or the large institutions, that adoption rate is starting to get to that tipping point where the large masses will start to embrace it.
Gary: 04:09 In my opinion it seems to me that much of the reserve and hesitation often comes down to culture. When we take a look at hiring, for example, a python coder we can take a look in North America at some of the PhD programs and they’re excellent programs and we had people who’ve dedicated years and years to their academic studies but the truth of the matter is if I need a python coder, a hardcore python coder, python is one of the languages used in artificial intelligence, I can probably go to South Korea or Singapore and if the only ting I need are coding skills I can find a 16 or 17 year old that can be just as competent as someone who spent a good number of years in post secondary education and the coding will be at the same level. The culture embraces, in these countries that we’ve mentioned, often what they’re doing is they’re running towards this new technology. They’re looking at holograms, they’re looking at all kinds of things. And here we kind of think of them a little bit as, a little bit freaky, a little bit creepy. There’s a curiosity but we haven’t really jumped in with both feet.
Eric: 05:28 There’s been a little bit of bad press about AI taking over workforces. Do you think there is a mix or do you really think that it’s going to take over and replace human beings?
Gary: 05:39 Well I think when we take a look at the implications of technology on the human form we’ve seen these things before. If Henry Ford was quoted as saying if he had asked people what they wanted they would’ve said faster horses. There have been many jobs along, especially the last hundred years or so that have tremendously benefited from a new technology but over time they’ve been displaced. And I frankly don’t think that artificial intelligence machine learning is going to be a whole lot different. We’ll see a period I think of, especially in the finance services industry, of the next two to five years with sophisticated chat bots, natural language processing, a variety of different ways of analyzing people and moods and emotions and so on with a variety of technologies that we’re currently using. It’s only going to get better and better.
Eric: 06:37 So Tracey can you give us a little insight from the customer perspective as to whether AI is being used as more of an inclusive technology or is it divisive?
Tracey: 06:47 Yeah, we’ve seen actually a mixture of both and each company and even each industry determines whether it is inclusive or divisive, right? So think about back to computers when they were first brought in, the fear was that they were going to get rid of paper, the whole paper industry was going to be gone, right? Well we’ve seen what happened to that. They said there’s no need to print because you have everything on your computer right in front of you. Well instead it’s created actually more paper because we all print on-demand at home and divisions and office buildings. We print just to take a look at something and then throw it away or shred it. I think that artificial intelligence, the fear might be there but I don’t think that that fear is valid because what it’s going to do, yes, it’s going to redeploy people and some of those routine tasks that you do and probably hate to do, those are the things that artificial intelligence does wonderfully at.
Eric: 07:43 What about the issue of bias? You mentioned early on that most of the data from the last couple years as culture has morphed over the last 50 or so years, how is bias being addressed?
Gary: 07:55 Well great question Eric. I mean if we think of bias issues, part of the attraction of using AI/ML is for gleaning insights that would rather otherwise go undiscovered. Some insights have been identified suggesting that we currently work to anywhere of up to 24 different cognitive biases. The first task is always to be aware that we can be cognitively biased, to try to codify them to understand them, to categorize them, once we can do that we can actually investigate the data and the decisions that are coming out of the AI/ML and basically have the AI proof itself to ensure that it’s not subscribing to those biases in the decision making process. We have a link to a white paper that describes those 24 biases but just being aware of what the biases are I think is really the first challenge. So knowing that they exist, knowing that we can codify them, we can classify them and that we can then interrogate the decision making process for those biases.
Gary: 09:10 Like everything else it’s early days, that will get better over time but we’re already aware of it and we’re already starting to codify it so we can detect it.
Eric: 09:27 So can you speak to the pros and cons of regulation and where regulation sits in regards to AI? Today in the U.S. in the financial sector especially it’s one of the most heavily regulated sectors on the planet so I’d love to hear about whether you see benefit or negatives in what’s happening with regulation today.
Tracey: 09:48 Sure. Coming from, remember I was 15 years in the financial industry so I see regulation as an necessity. Now could that be my upbringing? It’s possible but with artificial intelligence if you think about it regulations can be programmed and when I think back to when some of the regulations were coming in whether it was for mutual funds that were just being introduced or many other things, a lot of times if you are unsure where that line is, what the regulations, even though it says something there’s still always interpretation. So with artificial intelligence the machine, the system, can learn, machine learning can learn where those lines are. We used to have to wait for an audit and then we would find out are we within the lines or just outside of the lines. We had the guidelines and we knew that were close to those guidelines but we still didn’t know if we were really in compliance and we didn’t know where to improve or how to improve it. Whereas with artificial intelligence you can get there faster.
Tracey: 10:49 Some think that if it’s not broken don’t fix it. Well personally I think that you fix it before it becomes broken because I don’t want to be on the wrong side of regulations, right? So now that you know better you want to do better and we’ve created within our own company and I think many companies, we’re hearing more and more whether it’s regulatory that’s going to cause it or people are coming on site, we’ve created an ethics management office in our company. So what we do is we evaluate all of the projects that come through to ensure that the business we do is staying on the right side of ethics. So its not just a regulation, that it’s a law or rule but its also on that right side of ethics.
Eric: 11:30 I agree with that. And one of the things that we’ve seen is that regulators are actually not trying to block the technology by any means but are almost reaching out and saying help us, teach us, let us know what’s going on so that they can provide regulation that protects the consumer as well as continue to let the industry drive forward. Are you seeing that same trend?
Tracey: 11:52 Yes absolutely. We’ve had some regulators and some high level government officials come to us and talk to us at different events and conferences or even just pick up the phone and talk about okay so what are you guys doing? Because they have to create the regulations themselves and sometimes, you know some of the things that they put out we have even reached out and contacted them to say, hey, you’re kind of missing a large part of this, right? You’ve only got a small portion for the portion that you’re looking at, now you need to expand it and think about all these other things. Absolutely, the regulators aren’t the we-know-all, they’re trying to figure it out as well and many times it’s a collaborative effort.
Eric: 12:35 So let me transition a bit and as the two of you are on the ground and involved in AI globally, where do you see the biggest advances in the finance sector and the use of AI over the next, say, 12 to 24 months?
Gary: 12:50 I think as we look at the next 12 to 24 months particularly in financial services and particularly in North America we’re going to see much, much more use of behavioral data in helping customers understand and service providers, financial service providers, understand how they can do the best matching possible.
Tracey: 13:13 Yeah I was just thinking too, Eric, it’s really-operationally, the financial industry is starting to use artificial intelligence to be able to streamline the processes. So think about the media events that have happened for misuse of account openings, increased fraud, the economy and the impacts, artificial intelligence helps with all of that as well. For example, fraud, we can use the artificial intelligence to pinpoint the exact location, the exact machine and the exact time that the fraud happened. It can also take and link, it might be let’s say if Gary had a computer in his office and I had one in mine and you had one in yours and we’re all linked to this fraud, it can pinpoint all three of us. So that’s a good thing from a fraud perspective, from a law perspective, a regulation perspective. And we need to make sure that the financial industry as they start to use more and more of the artificial intelligence they can help and be collaborators just like with the regulators with law enforcement, etc.
Eric: 14:18 So that’s interesting is we just finished a mini series on fraud and there’s always two sides to the equation. We highlighted a situation where fraudsters are using AI to fight the same AI that banks are using to fight the fraudsters.
Tracey: 14:37 We always used to say with the card industry, we weren’t sure if the fraudsters were ahead of us or if we were ahead of the fraudsters. So it’s still the same whether it’s artificial intelligence or any other means.
Eric: 14:50 Gary and Tracey thank you so much for your time today and we really enjoyed hearing about where AI is headed from both a data customer experience and financial services sector point of view. As well it was fascinating to get some insight into what’s happening across the rest of the globe.
Tracey: 15:07 Thank you so much.
Gary: 15:08 Thanks Eric, it was a pleasure.
Eric: 15:13 We hope you’ve enjoyed this episode of the Finance Frontier. In our next interview we’ll be continuing to explore the topic of artificial intelligence so tune in every other Wednesday for new episodes and until next time subscribe on your favorite podcast app. Since we depend on listeners like you to help us spread the word we’d love it if you’d take the time and post a review of our podcast on iTunes. Until next time I’m your host Eric Hathaway.
Love the show? Want to be featured as a guest? We’d love to hear your questions and comments and welcome guest recommendations. Our producer Sara Tatnall can be reached at sara.tatnall [at] zootweb.com.