Jon Scheele (00:00)

We're all trying to figure out our place in this new world of AI, how we can get better results from our organization for our customers, as well as our own personal careers. I'm very pleased to be joined by Keith Carter. Keith, we've known each other for about five years now.

and through a couple of iterations but please for the audience, introduce yourself.

Keith Carter (00:28)

Jon, yes, it's been a few years and I love that we get a chance to talk about something near and dear to my heart. Hey everyone, I'm Keith Carter and I've had 52 years in the making. Wall Street technology is my background, then New York fashion and then recently, of course, Singapore academia. Being a professor at National University of Singapore and finally,

These days I teach businesses and governments how they should be thinking about AI and digital finance and more. I'm a member of the Bretton Woods Committee in Washington DC. And I'm with a group of partners called KDA Capabilities here in Singapore, where we focus on KDA means know, more, learn, and decide, and then act, take action. So I hope whatever Jon and I talk about today, there'll be something you can take action on.

Jon Scheele (01:23)

Yeah, so Keith, you've had a wealth of experience across different aspects. So you mentioned Wall Street. You've been heavily involved with data and analytics for some time. We met through the National University of Singapore's Advanced Professional Certificate in FinTech, which you set up. And you really amplified...

Keith Carter (01:41)

Yes.

Jon Scheele (01:45)

your reach now with your understanding of how AI can be employed in organizations. But we hear a lot about possibilities and also some of the dangers of AI. But what do you think about when you advise an organization who is really trying to make sure that they stay ahead or don't miss out or want to achieve a financial return on investment from an AI initiative.

Keith Carter (02:16)

I'd like to start off with that we're not trying to do digital transformation and AI implementation. All those are fancy words, but what we're really trying to do is get one more customer to buy one more product at the right price profitably. So we used to have a phrase at Estee Lauder, which was around

How do we profitably maximize revenue?

So this type of idea is where we start from. And so when people talk about AI and I talk to businesses, I start off on the business side. What is the workflow that is taking up your time and preventing you from delivering the type of customer service, the type of sales, the type of business outcomes that you'd like to achieve?

Let's see how we can use AI or other technologies to improve that.

And then there's a direct connection to customer service, net promoter score, revenue, attribution, et cetera.

Jon Scheele (03:21)

I guess this is an age-old problem for anyone in technology that we have quite large technology organizations in an enterprise now, and there is often a concern among the business leaders that the technology is not really supporting the business strategy. And the technology teams themselves constantly make sure that they not only align,

with the business strategy, also show business leaders what technology can do for them in terms of capability. This is always a challenge though, as you know, having from a technical background and also having an extensive business background, there is this divide between business and technology.

in sense, you need to make sure that the technical people can describe things in terms that the business people can understand. But then also, the business leaders need to be able to describe what they want to achieve in terms that the technical people are able to not only understand that, but can match their technology capability to help to achieve.

that. So you mentioned some business terms there about customers, about profitability and maximizing revenue. I guess there are number of technologies that help to do that and we've discussed before that APIs if used with partners can help

your partners bringing you more business, more customers than you would have been able to achieve or a more streamlined customer experience so that it's...

Keith Carter (05:12)

That's right, Jon. Like the airline industry, Expedia, using API to find out what next Singapore Airlines ticket price will be. That's where API is come in.

Jon Scheele (05:22)

Yeah, so the opportunities for AI, a specific technology is good for some things in business and not necessarily good for others. So where do you see the different types of AI applications helping with that, that customer experience or gaining new customers profitably?

Keith Carter (05:43)

Sure, I I like to look first of all at business as a war. You're trying to position yourself in the right market and place and price and you're trying to take over customers, get them to buy your product. So that's attrition, position plus attrition of your competitors. So if we think about it like that, where are our

points of getting better position. Number one is that as a business owner, I might have certain markets that I work in comfortably because of language. Now all of a sudden with AI, can write, speak, do marketing materials in at least 40 other languages like that. So what were my markets before?

And the other markets I needed to add some barrier to entry and so forth. Now the barrier is a bit lower. Not gone.

So that can be my positioning. The other positioning aspect is as I go to market the company and get people to understand what's available from us. And this is where Amazon has always used the technology. What is my competitor doing? What should I offer that's going to be marginally better return for the customer?

and marginally better availability. And ideally at a better profitable price for me. So that's where again that position and nutrition can be used for AI. AI can look at something called hyper personalization. Look at the market basket analysis. What did people buy before? Where should I be selling them now? It can be looking externally. What are my competitors selling and at? Do they have inventory? How long does it take to get there? Can I beat that?

those areas there. Of course, I can also look at quality. What are the customer reviews like? AI can help understand that as well. So that's kind of the front facing and also what I would consider somewhat low risk activities. The other side is on the back end.

I call AI an exponential technology because it distributes creativity. Where solar panels help distribute power anywhere in the world. And Starlink helps distribute internet anywhere in the world, which is great. If you take a cruise and you have a Starlink cruise, you're very happy with your internet service. And AI distributes

Creativity. I can be anywhere in the world and I can make a song, I can make art, can make an essay, or even, and more. know, respond to a customer. And so what does that mean though is I can have my team anywhere in the world respond with a higher quality output to my customer. Not simply the response or emotional aspect or struggle to get information, but literally have the AI get the information.

write it in a style that I most likely like. Now they make mistakes and there's an opportunity for humans to review. But man, my quality and speed to solve a customer issue can be higher. But notice I haven't gotten rid of the human in any of these scenarios.

So those are the ways that AI can be used, but that's at the tip. The real reason why AI is exciting and no one talks about this, it almost as if people forgot what made businesses and governments really excited about AI. It was AlphaGo. AlphaGo is a Go game kind of like chess. You

it was able to beat a human consistently. In a game of position and attrition, it could anticipate the person's next move. Now, what companies are using AI to anticipate their competitors next move, the move after that, and so forth? Those are the ones that are winning. And the problem is though, in order for AlphaGo to win, they had to

See, not only its information, its pieces, it also had to look at the competitors pieces and see where they were. And it had to also get trained on how did the competitor move before? What happened last season when we discounted? Did they discount also? When we didn't discount, what happened? You know, all those things, all those moves need to be there. And then of course you're optimized that.

And that's where AI is really meant to playing in terms of business and government and of course military. And so I just remind us that this is not just for chat and writing emails. You're supposed to be able to write an email and then ask, what's the other person going to say? What are they going to ask? That's move number two.

What should my answer be? That's move number three. How will they feel about the answer? That's move number four. Will this get to my objective of them agreeing to whatever I want them to agree to? That's move number five, six, seven. That's what AI is supposed to do

Jon Scheele (10:43)

There's a lot in that so let's just start to unpack a couple of things. So you started off by talking about a reduction of barriers to doing new things. And I think about the classic market. Do I put new products in an existing market? Do I take my existing products into a new market? And the reason

companies contemplate this is they're trying to figure out what are the adjacencies? What are the things that they, that the capabilities they have that are unique to them and that they can translate into more. Am I going to get more customers by going into a new market or am I going to get, deliver more value or more products to my existing customers? And you talked about being able to reduce some of the barriers because there are

Keith Carter (11:16)

street.

Yes.

Yes.

Jon Scheele (11:35)

language differences or there are challenges in translating something from one geography or one field to another, then you can leverage tools that will reduce the effort involved in doing that. So it makes it easier to move into new areas and potentially serve more customers who are slightly different or in a different area.

than we would before. The other part I think, we typically often think about new products, but what you're really describing is more value to the existing customers by being able to solve their problem more comprehensively or faster. And that...

That makes the customer relationship more sticky, perhaps even allows you to earn more money because you're delivering more value to your customer or certainly makes the customer less likely to leave for a competitor. Then you started to also talk about the capabilities of people, I think, because you talked about the distribution of creativity and...

Keith Carter (12:31)

Yeah.

Jon Scheele (12:44)

I think you sort of also alluded to you can do this from lots of different places. don't have to be in a centralized place in order to achieve that. What are some of the capabilities that business leaders, what are the skills, the capabilities that they should be looking to build in their team in order to deliver those sorts of experiences for their customers?

Keith Carter (13:09)

What are the capabilities that a business leader needs to have? Really they need to first of all know their own workflow and the second thing they have to have is a great handle on the data necessary for the workflow. Let me give you an example and one thing let me caveat this doesn't mean that you have a data center or a data warehouse or data lake.

or a lake house. These are technical things. If a customer calls in and the employee experience to solve that customer request is instant, like, I can just look up the order and see where the thing is and give them a response. Beautiful. That means the employee who's dealing with the customer has the access they need.

has the, you know, sees where everything is and answers the customer query. The implication of that is that the manager has anticipated the type of questions that the customer may have. You're not going to ask, you don't need to answer every question in the same way. You don't need all of the data in your company. You need to be able to, the data that answers the questions that are going to be asked. That's called actionable intelligence.

Every manager today needs to be thinking along those lines, how can I answer, how can I anticipate the questions that are going to be asked so that I can give the most accurate answer possible? Now, here's where people get angry about AI and why I have mixed feelings about it. It's not because of the technology, it's because of the way it's handled.

If I know what customers are going to ask, I can train an agent on the data that's necessary to answer that question. And I can also decide whether I need an agent at all. Maybe it's not an agent, it's just a database lookup and clear crisp black and white response. We need to pick out when we use technology, whether it's automation or AI. And those are two different things. Or analytics, let's put all three of those on.

table.

So if a good manager will not look at AI as a productivity, wow, customer service answers faster. No, They'll look at it as, hey, customer service answered better, answered more, was able to cross sell and upsell, left it, built goodwill. Those are all the things that AI should enable. But as I said at the beginning,

That AI has to have the right data. You have to choose whether it's automation or whether it's analytics, really actionable intelligence to answer the question.

Jon Scheele (15:51)

So I guess there are several generations of thinking there because there's still the basics. There's still understanding the customer needs, what the customer experience and the workflow is. The data that's going to support that. And you mentioned actionable intelligence. I think you were writing about this in the days when mostly what we're thinking about was predictive analytics. And it was really before people were starting to get excited about.

generative AI. And then I think what you alluded to is the new possibilities are to explore what if scenarios better in order to anticipate what some customer questions or needs may be and already formulate potential responses.

Keith Carter (16:40)

Yeah Exactly exactly that's When I called USAA USAA is one of the top banks in the US and My father had passed away unlike the other banks. I called Citibank. I Bank of America said my father passed away. They helped me close my dad's USAA the guy said Do you need any help with the funeral arrangements now? This is a surprising request from the bank

Jon Scheele (16:40)

to that.

Keith Carter (17:03)

I said, well, because he knew where I was, he knew where I was calling from, that's how I authenticated, right? They know where you're calling from. He said, you're in Singapore, but father passed away in Maryland, that's what the death certificate says, put two and two together. do you need help with the funeral? This is just a human thing, right? And I said, yeah. He said, okay, well, we can suggest some places, some funeral home. I said, that's really nice.

Jon Scheele (17:08)

Yeah.

Yeah.

Keith Carter (17:29)

Then after that, after he gave me that information, said, and by the way, how are you planning to pay? Well, I was going to round trip the money, you know, from New York, Singapore, and then pay my credit card. He said, no, no, we can just pay directly through your insurance to the funeral home. I said, that's wonderful. So I could concentrate on my father's passing, the logistics, the payment, everything was taken care of. And so.

Jon Scheele (17:49)

No.

Keith Carter (17:54)

And the customer service agent had the things at his fingertips because I hate to say it like this, sorry Jon on your podcast, everyone dies at some point. So why are we surprised at the process, right? You don't want a funeral, sir? I mean, come on, you don't have to ask that question. And so it's just a matter of what I call the moment of truth. And the moment of truth is in any product.

or service is when you expect it, when you have the need, it's there anticipating you. And not just simply, we're going to pay you the insurance money. It's, this is a major disruption in your life. We're not just going to move the money. We're going to handle the things that are the reason why you need the money, right? Wonderful, simple, just human.

is what we need to have. And when people say, know, customer service, answer him faster, get off, go to the next customer, that doesn't happen. Now, what's the difference? That customer service agent, just taking us a little bit of time, having everything at his fingertips, he didn't have to put me on hold, he didn't have to say, let me transfer you to the insurance department, nothing. What does it mean though, that I'm gonna be a customer for life for that bank, right?

my lifetime value to that bank and all the way up.

not because of productivity, but because of effectiveness.

Jon Scheele (19:16)

And the very human element because humans want to deal with humans and the more that that human can feel for and accommodate that customer, the better that relationship is going to be. I mean, that's

Keith Carter (19:20)

That's great.

Jon Scheele (19:33)

That's a challenge for any organization, of course, because they need to meet some financial results in the near term as well as the long term. ⁓ clearly, organizations that think longer term will be more concerned with maintaining that human connection with their customers in addition to achieving productivity gains and...

Keith Carter (19:42)

Yeah.

Exactly.

Jon Scheele (19:56)

and better ways to serve.

So from an individual standpoint, so we talked about what organizations should be thinking about how to apply new technologies to achieve better outcomes for their customers and for their companies. From a career standpoint, what as individuals should we be doing to ensure that we have a career?

You joked just before the call that we don't have a career anymore. I think we do have careers if we're prepared to constantly evolve and adapt and think about the next steps. What are your thoughts on those next steps?

Keith Carter (20:34)

Jon, I agree with you. And if I go back a hundred plus years ago to the career of a horseshoe maker, if they didn't become a tire changer, they didn't have much of a career left. And by the way, people say, there's going to be much less, you know, there's a lot more tire changers than there ever were horseshoe makers. So and why is that?

Because to change a tire takes less skill and less effort than making and changing a horseshoe.

And so more people can do it. And let me back into something that's a, you're right, Jon, we have to continue to transform our thinking. We have to be open to new ways of doing work. We have to figure out how to create new value.

And the issue that we have is the speed at which this is coming for some large companies may be too fast for the society to handle that change. Some companies want to let go of 4,000 people right away. Some have.

Jon Scheele (21:40)

Yeah.

Stop hiring new graduates for a while.

Keith Carter (21:45)

That's right and yet though you know that's such a short-sighted view. Really you're gonna have the AI run do everything and not yet. It can't.

As we look at this, anyway, I'm gonna stop there. I think the most important point is career change is a must as much as horse shoe to tire changer is from programmer to AI wrangler is from, copier editor is now copy editing hundreds of documents very quickly with AI assistance.

know, volume plus quality plus maybe imagination is gonna change.

Jon Scheele (22:24)

Yeah, well thanks for sharing that. guess my feeling is that we still need to be well-rounded individuals. That we should have a specialization, something that we're really good at, but we also need to be able to communicate with other people. And you've heard the term T-shaped individual, where most of us have a deep specialization in one field, but we should have a broad understanding of...

a number of others. think constant learning and adaptation can help us to deepen some of those broader, the top of the T a little more, as well as going deeper. But really appreciate that perspective you shared and look forward to discussing it even further in future.

Keith Carter (23:02)

Yep.

Jon, thank you so much. It's been a pleasure to share some thoughts with you. And I hope to inspire us all to know a little bit more, decide what we're going to do, and then take action. There's no time to sit on our laurels. We all have 24 hours in the day. And we better use it wisely. to...

I meant what I meant by widely is to widely study various things. And one last thing I can leave is if we think about the next generation too, you have kids, I have kids, Jon, they can either be makers today or consumers. It's so important that we teach them, that we inspire them, that we encourage them to be makers.

You know, with AI today, anyone can make music, but to really have studied it and to have, and make it, make it good and production quality is even easier than ever before. Many people can write, but few write books and publish. Now's a great time to do that. You can program, you have an idea, you can go to an AI and it can start you off on the programming side. So.

There's so much that AI provides in terms of distributed creativity now. And it's really distributed to anyone who has access to this. And it's a matter of us making sure that we get inspired by creativity we see to do our own thing, not just get into a spiral of consuming that creativity and not really going and doing some of our own.

Jon Scheele (24:42)

That's a great thought for us to consider and we probably should be looking to do that as much as we can, not only for ourselves but to inspire others to also. So thanks very much for sharing that.

Keith Carter (24:57)

Thank you, Jon.

​

Start with the customer – find out what they want and give it to them.