AI in the Workforce

 Join our esteemed panelists as they discuss the vast potential of AI, and the ethical and moral questions around it. Also, hear some remarkable examples of how AI is contributing to our workforce.

Transcription:

Paola Peralta (00:07):

Hello, and thank you so much for joining me.

(00:12)

And I am joined today by my lovely panelists, Lindsay Jurist-Rosner, CEO of Wealthy, and Jeri Hawthorne, Senior Vice President and Chief Human Resources Officer at Aflac. And today we're talking about a topic that I personally cover a lot and I think is very important for I think everyone in every industry right now, which is AI in the workforce and kind of what that means for businesses, what that means for business leaders. And I have both my panelists are very excited to, I think get into this topic. And so to kind of kick us off, I just wanted to ask the very general question of how has AI kind of disrupted the workforce in general? And then also what has it meant for your guys as companies specifically?

Jeri Hawthorne (00:58):

Do you want to start or do you want me to start?

Lindsay Jurist-Rosner (00:59):

Either one.

Jeri Hawthorne (01:01):

I'll start. So it's interesting when I think about AI in the workforce, because AI is such a broad term that encompasses so many things in actuality, and it's getting a lot of airtime now, but it's actually been disrupting the workforce for a number of years. I worked at a quantitative hedge fund a few years ago, and they were using modeling to not predict the stock market, but to remove the human element from the traders. So we had engineers that would, or excuse me, we had quants that would design models based on trends that we would see in the marketplace. And then we would have engineers that would build those models and we would have humans who were traders who would monitor those models to say, okay, is this doing what we expect it to do? So that was about seven or eight years ago. So it's been in existence for a while.

(01:57)

I think it's getting a lot of airtime now in particular because of chat, GPT and the advances related to being able to interpret emotion, things like that. So I think for us, it's been more of an enhancement to our workplace. A good example for us is that Aflac, we are a supplemental insurance company. We have other products, but our core is supplemental insurance. On average, we process about 650,000 claims per month. Right now, about half of those claims are automated. The way they're structured today, the other half are less automated because we operate in 50 states and in Puerto Rico, there are different regulations in each of those states. People have bought claims at different times or policies at different times in the lifecycle of our company. And so we're working now with AI to say, how can we use artificial intelligence in order to help us become more efficient with these claims that now we're processing manually?

(03:06)

So we have a team in Northern Ireland. We opened up a tech hub there about two years ago, and they're doing things that I was just there last week, and they're doing things that are amazing from an AI perspective where they have actually said, okay, we're going to automate. So any claim that's paid out under $200, we're going to automate the adjudication process of that claim so that you no longer have to do it manually, but the system will do it. So right now that will automate about 50 to 60,000 claims per month, and then by 2027, we anticipate that that will get up to about 200,000 claims. So if you think about that from a human perspective, what that does is it removes that remote or me that rote work from the hands of humans, and it allows us to use our humans to do more. I think with that, aligned with what your organization at Wealthy does to really do that high touch type of engagement with our customers.

Lindsay Jurist-Rosner (04:06):

Well, Jeri, I think you hit on a couple of points that I was probably going to hit on too. We're very aligned. No, it's perfect. And I think the coupling of these two companies talking about AI is very exciting because I don't know that you think about the organizations and our core offerings as maybe being AI first or the type of company. So I think it's very cool to be on the stage. But yeah, so I run a company called Wealthy and unlike Aflac, which everybody knows, I'll tell you just quickly what we do. I think that's important context. So we help families that are dealing with complex, chronic and ongoing care needs. So think of us sort of like a care concierge. So when an employee is dealing with an aging parent or a child with autism or genetic condition or down syndrome or a spouse or a sibling or their own tough diagnosis, wealthy hand holds them through finding the right resources and services and programs that are the best fit for them.

(05:03)

And we work with employers. We've been doing this for almost eight nine years at this point. So been at it for a while, and we've seen a pretty dramatic and exciting acceleration in some of our tech capabilities, really just even in the last year due to generative AI. And so we are very actively using generative AI, and it's really about, so very similar to what you shared Jeri, not about changing our core business or our core offering, but really just kind of supercharging our incredible team that's doing incredible work. How do we just help people do the stuff they do best and the other pieces that are more rote in your words, which I think is the perfect description or more monotonous or kind of routine, how do we help take some of that off their plate? So yeah, I was pushing our CTO from the time we started, when are we going to start using AI? When are we, and all of a sudden this kind of door opened with generative AI, and we were able to almost overnight stand up a private environment that's secure and then start playing and building tools and futures. And so it's been very exciting and very accelerating for our business.

Paola Peralta (06:27):

And I think you hit a really important point, which is that I think when people think of AI, they think of the big tech companies, they think of a lot of Microsoft, Google, but it's affecting every company. And so you kind of touched on this a little bit, Lindsay, but when talking about AI, how did you guys decide, or how did the conversation come up where you needed ai? How did you as to leaders make the decision on where, what needed AI? How did that kind of thought process work?

Lindsay Jurist-Rosner (06:55):

For us, it was very organic. We didn't sit around and say, Hmm, there's new technology called generative AI. How do we apply that to our business? That's not what we did. We've always had a roadmap of technological tools and features that allow us to scale. We know that the experience we provide is ultimately human led, but tech powered. And to be able to scale and provide a consistent exceptional experience to families, we rely on technology. And so we've always had this big roadmap of all of these tools and features and functionalities that we knew would help us be able to scale that human led experience. The challenge for us in implementing AI before generative AI was just that the skills required to do kind of machine learning. It just was a different skillset. These are highly, before generative AI required highly specialized talent and engineers and data scientists who know machine learning have implemented machine learning tools.

(08:09)

And as a growing company, it felt hard to invest in a team that just did that. So there was this massive unlock for us with generative AI because all of a sudden our current team could use these powerful tools and we didn't have to hire specialized talent to be able to complete what was already on our roadmap. So for us, it was accelerating our roadmap, things that we wanted to get done that we thought would take months and quarters and years, all of a sudden took two weeks. So it was this very powerful kind of transformation of our team, but not really doing anything we hadn't already planned. Yeah.

Jeri Hawthorne (08:50):

It's interesting. I actually think that the juxtaposition of our companies is we work together, but we have very different size scope sort of journeys because we've been in existence for 65 years. And so ours, Aflac is a 4.2 billion revenue company. We operate in Japan, and the US have been in existence for 65 years. Our Japan is actually the biggest part of our revenue. Most people don't know that. So we were very deliberate. And so we have been able to invest in data scientists, machine learning technologists for years. It's been part of our culture, not an Insurtech company, but we sure have a heavy investment in technology and have for many, many years. And we've done it through the lens of how do we make our customer experience as simple as intuitive and as easy as possible? So if I'm a policy holder and I go to the doctor and my company bought a wellness policy for me, and I just go get my physical, how do we make it easy so that person can just answer a couple of questions and we pay their wellness benefit?

(10:04)

That's how we've been leveraging AI in order to enhance that experience for probably about a decade. And you're right, it really has accelerated very, very recently. And where we've seen it expand is even areas outside of our traditional business. And so what do I mean by that? I mean, I lead human resources, sorry, I lead human resources and we have engaged simple. It's to chatbot, we've named her Amelia, it's for our applicants. So if they go to, they come to our career site and they want to apply for jobs they can engage with Amelia. And Amelia got, when we first implemented Amelia, she was not very smart. And I kept saying to the team, we can't launch this until it's smarter because it learns. It is using generative AI to learn. And so I was like, I keep asking it questions and it keeps just pointing me to the careers page, this is not getting smarter, but I was wrong.

(11:06)

It got smarter. And so now it's our candidate care. And we've seen usage go up month over month, and we've actually just rolled out so that not only can candidates use it just to answer general questions, but they can also actually ask a question of a hiring manager and they can use it to schedule interviews. If they're selected for an interview, they can actually use it to schedule an interview with that hiring manager or with that recruiter. We've started to see it go into other areas of the organization, which has been phenomenal. But the other piece that we found is that we've really had to think about our governance as a large company, the governance of that amount of data. Of course, all of the data that we all in the insurance space have, it's all HIPAA protected. There's so much governance around it. And so I talked to our head of security and just said, what do you think about this? And he said, really? He said, we still think about it as a data issue. He said, we don't think about it as an AI or a chatbot issue. This is about how do we ensure that we are protecting the data of all of our policy holders and our customers in this time where it's now there's a lot more sophisticated phishing, there's a lot more sophisticated hacking, there's a lot more sophisticated use of technology in a bad way. So I thought that was actually a very interesting observation

Paola Peralta (12:35):

And something you guys both have in common is that you deal largely with clientele and you have a lot of people to make happy and to lives make easier. And so how has that been received, the implementations you've done in the rollouts you've done? How has it been received by clients and how has it made their experience a lot better on the user side?

Lindsay Jurist-Rosner (12:54):

Well, we're being really careful. This is still sort of, I mean, to Jeri's point, I mean there isn't really clear regulatory boundaries. There's still questionable legal and ethical and even moral considerations here. And so it was interesting in the early days of chat, GPT and generative AI, as our team started to get excited, we went to our board and said, here are our plans with generative AI. And our board gave us a very clear directive saying, we do not want to be cutting edge. We do not want to be bleeding edge with AI. And so we've made the decision with AI, we're going to use AI to supercharge our humans to help behind the scenes. We do not do any AI in a member facing environment. And that's really important. I mean, until we're really clear on where all this lands. And then the other piece for us was just almost doing kind of self-regulation, right?

(13:55)

So if there were to be a regulatory body that kind of puts parameters around AI, what might that look like and how do we put that in place early? And so we have an ethics committee and they put together guidelines for responsible use of AI. And so we sort of did that preemptively before there's any sort of governing body that tells us what that really looks like. And so that was a really helpful process. And then what we did is really fully embraced training our team on responsible use of chat, GPT and generative AI. So not just the product and engineering orgs who are playing with it and tinkering and building tools to support our operations, but even for marketing, for HR, for all of the different use cases where chat GPT can kind of accelerate people's days and allow them to be more productive and efficient in marketing.

(14:47)

Maybe it's a first draft of a content piece in HR, it's first draft of a job description. But we did really careful training to say that chat GPT should be used in certain ways and not used in other ways, and really educating about privacy and security. And then when we did stand up generative AI within wealthy and using it for our own business, we have our own environment. So we're not using general learnings from the internet as part of what we serve. It's a closed gated environment that we can make sure is really incredibly private and secure. And so those are the actions we've taken just to be overly careful and cautious, but also start using it and experiment and embrace that this is the future and really empower our team to use it.

Jeri Hawthorne (15:43):

Yep, same. It's interesting. We're the same. So our philosophy is always kind of stay within the herd. We don't want to be out front. We're an insurance company risk averse by our nature. We know we need to embrace, get on board, but at the same time do it in a very responsible way because of our size. We have a lot of governance committees in existence. And so we have our global IT and security committee that's a committee of the board. And as all of this has been rolled out and gotten more media attention, our guardrails have just sort of followed that. So we don't even allow our employees to have access to chat GPT on company provided devices. If you want to do that on your personal phone, that's fine, but we know people are using it and people are using it to play around and research things.

(16:37)

And what I've observed myself with all of that is that chat GPT is really just a super fast, very efficient search engine, but it's only as good or as useful as the questions that you put into it, sort of like garbage in, garbage out. So it's that it can become a bit of a self-reinforcing vacuum. So if you want to ask chat GPT something very technical that has an answer, you will get that technical answer. If you want to ask chat GPT, who is the best presidential candidate? Well, that's an opinion. And so that's really the line. And you talked about the moral and ethical and that right now it's great from a perspective of finding factual data really, really fast, but when it gets to the pieces of opinion-based, more emotion-based, it's just not there. And it's not reliable. It's probably, and even with factual things, it may not be as reliable because it may lag still. So there's a lot of evolution to come, and it's coming very fast. It's coming very, very fast. And I think organizations, all companies need to be thinking about it from the lens of what's our governance? What's our perspective? You need to have a point of view at the very least as to how it can influence your companies. But it's still not new, but there's still a lot of improvement that can happen with it.

Paola Peralta (18:17):

And I think that's a perfect segue to kind of the employee side of it, which it'd be remiss not to talk about the fact that I think when a lot of people think of AI, there's still a lot of doubt, a lot of fear that robots are going to take over our jobs or that they're going to be let go in order for automation to happen. And so internally, how did you guys handle those fears? Was there kind of pushback from employees? Did you see a little bit of hesitancy from any of them? And so how did you guys handle that and how should employers be having these conversations as well?

Lindsay Jurist-Rosner (18:50):

We were really conscious of that exact point when we started talking about chat, GBT, generative AI and the larger space within the company. We wanted to make sure that our communication always led with appreciating the truths that will never change about our business. And frankly, most businesses still wealthy. And our experience will be human led, brand matters, content matters, employee experiences matter. There are some truths that are really regardless of where AI goes. We did a lot of work of thinking about 10 years from now in an AI first world where AI is much more involved, what does wealthy look like in that AI first futuristic world? And we still landed on the fact that brand matters, experiences matter, humans matter. And so we really got focused on that as the kind of upfront communication and even the branding around using generative AI within our company.

(19:59)

We made the decision to brand it as copilot, which you see a lot of the big tech companies doing because we loved that, implied that it was going to ride alongside our team, right? Team doesn't change, but we have an extra tool in our toolkit. And so the communication went really beautifully and it was kind of fun. And we focused on education and bringing people along on our journey as we were starting to explore and experiment with generative AI. And we really didn't face feedback that people felt like their jobs or their roles were threatened. It was something we preempted. I guess in our communications,

Jeri Hawthorne (20:43):

I would say we have 6,000 folks in the US and I do think there was concern among some of our call center and claims employees. These are our frontline people who are dealing directly with the customers. And when all of this started happening, there were conversations about will there be layoffs? What's the future of the employee experience at Aflac? And we did the same. We said, look, this isn't about getting rid of jobs. This is about repurposing people for a better and higher experience for our customers. So we don't want customers talking to customer care reps about changing addresses that should be automated. We want them talking to customer care reps about, I have this medical experience and I want to understand what my policy covers. I'm sitting with my significant other, they're in the hospital and I need this money, and what can I do and what is covered under my policy? And I'm confused and I am emotional and I'm tired. That's where we want our humans to intervene because that's really the experience that we want our customers to have. So it was more for us about saying to people, look, this isn't that sort of a play. It's about yes, we have to get more efficient, absolutely, but it's not about getting efficient for the sake of getting rid of humans. It's about repurposing those humans to really give our customers a better experience.

Paola Peralta (22:11):

And I like Lindsay that you mentioned this futuristic world that I think a lot of people like to think of when they think of AI, when cars are floating and everything like that. And so as two leaders, how do you guys plan for the future when it comes with that? Because we've been talking about how it's still semi new, there's still a lot we don't know. There's still a lot for it to go, even though right now it feels like this is as far as it will go. It always feels so new and so advanced. And so how do you guys go into those conversations and how do you guys move into business thinking of the future and how you're going to continue to use it and have those conversations, and how should leaders go about those conversations themselves?

Lindsay Jurist-Rosner (22:49):

I think that's a fascinating question, and it's something we grapple with, not just on the AI topic, but of course across all areas of our business is how do you future proof the business? How do you think about what's going to happen in 5, 10, 15 years? And that could be trends, just different trends in the workforce or whatever kind of macro trends we're facing. But couple that with the work that has to get done more near term, and it's a constant push and pull. But I think the biggest kind of learning that we've had is just it's critical to think about both. So where do we want to go? Where do we want to go and where do we think the world will be and how do we make sure we're kind of oriented toward that future world, but then what are the kind of blocking and tackling tactics and builds that we have to do over the next quarter or two to satisfy our clients, to scale our delivery or meet those kind of near term objectives? So it's both, frankly.

Jeri Hawthorne (23:51):

And I would say I lead an HR organization and my perspective to them is, look, we have to have a point of view, so we can't stick our heads in the sand. We need to understand this. We need to have at the very least a point of view on where we think this could impact our organization, our employee experience. Do we need to run head first? No, but we need to have a point of view When Aflac thinks about everything, the lens through the lens of a customer, our customer experience, HR, we think about everything through the lens of what's our employee experience? That doesn't change, that's sort of the north star. So what we want that to be. So if you keep that sort of north star consistent, then it's really about what are the tools that you are using in your arrows, in your quiver, so to speak, to enhance what that experience is. And that's really how we've started to talk about this with the team, with the business. It's that this is just another arrow, like pull this arrow out. It used to be that you could put people on hold. That was a big deal in the 1970s. You could put a customer on hold when they were researching their policy and going into the microfiche to do that. It's the same. It's just how do we use this to enhance that experience?

Paola Peralta (25:15):

And just because of the nature of the topic, I did want to leave time if anybody had any questions for either Jeri or Lindsay on this topic. Nope.

Lindsay Jurist-Rosner (25:29):

I mean, I'd be remiss. I feel like I didn't mention our main AI features. And in case that's interesting, I'll mention some of the features that we've deployed that use generative AI. So we've launched two features in the last couple months after creating this private environment. So there was a lot of engineering and development built there. And then the features that we built, the first feature we launched was sentiment analysis, which it's actually a profound difference for our business. So the concept is as we scale and have more and more families, we're touching and supporting, and our business has scaled considerably over even just the last couple of years, ensuring that that member experience just is as high quality as we want it to be, meets our standards. And in the early days of the business, we could do that with human oversight. So we had care managers overseeing our care coordinators and manually sort of checking messages and interactions to make sure the member experience was good.

(26:33)

That doesn't scale. And so one of the use cases we're really excited about is basically sentiment analysis. We trained the machine, trained the generative AI tool to basically understand what language could be perceived as positive or negative or indifferent. And so what words, what combination of words, what types of specific things would we see in our business? And training the tool to be able to flag negative sentiment as soon as it happens to our care managers. So our care managers can instantly go into those member experiences and basically determine is the member upset about something? Well they did, or is the member upset about the healthcare system in general and be able to help our team take the appropriate action. And so we scan, so the tool scans messages all day long and just flags those negative sentiment for us, which is basically quality assurance, quality control in a really efficient way. The coolest thing is we stood up that feature in a hackathon, so that's the power of generative AI, our CTO as part of a hackathon, which are really fun to do. If folks don't do that at their organization, I recommend him. And he built it in a day. And that became a feature that we then fully developed and launched in our platform. So that's just one kind of use case, for example, of the tools and features we've launched.

Jeri Hawthorne (27:59):

I love that. I love that. And I wish I could say something as interesting as that. I can't. That's amazing. We have a group in our Northern Ireland that is, they do hackathons. I was impressed when I was over there, but we don't have that. I can speak to something that creative on the horizon anytime soon, but we could, and I just may not be in the loop on that type of activity.

Paola Peralta (28:24):

And to close this out, I think that also raises the question of how does AI play a role in recruiting? When you guys look at, it's well known that young people are very tech forward and tech savvy. And so when you guys look into recruiting, is that something that you guys have in the back burner? Is that something you guys think about when you're looking at people to join the team that they're AI savvy? Is that something you keep an eye out on in resumes?

Jeri Hawthorne (28:46):

It definitely is for us and in particular in our technology spaces, but also from an analytics and quant perspective. So we're always looking for people with innovative experiences, people with innovation within their resumes to help to evolve our footprint in that way.

Lindsay Jurist-Rosner (29:05):

I agree with Jeri, and I would just add one point, which is I've followed closely Microsoft's journey and sat Satya Nadella's leadership, and they talk about learn it all versus know-it-alls, which I love. And so I think that's more of the spirit with which we're doing more hiring is looking for people who are comfortable learning new tools because generative AI and chat GPT is the tool today. In two years, it could be something totally different. And it's really a personality and a mindset for people who are comfortable not knowing something and then figuring out how to go about learning it and just embracing those new challenges and those new learnings. And so that's really the personality trait that we're very focused on in recruitment today, more so than kind of AI specific Technical. Yeah. Yep.

Paola Peralta (30:03):

Well, Lindsay and Jeri, thank you so, so much for kind of walking us through how AI has changed the workforce in general and how you guys are incorporating in your business. It has been an absolute pleasure.

Lindsay Jurist-Rosner (30:13):

Thank you. Thanks. Wonderful.