Episode 148 - Explaining AI, Decision Science, and Augmented Intelligence with Lisa Palmer
To what extent can A.I. and machine learning be applied? We all know how integrated technology is becoming into our society — from predictive text to decision science. Today, we'll hear how they are being applied in places we would never have imagined.
Joining us for this episode is Lisa Palmer, CIO and marketing professor, as she discusses some of the latest trends in technology and A.I. She also talks about how the nature of innovation has been changing in the face of the pandemic.
For surprising insights on technology and innovation, tune in to this episode!
About Lisa Palmer
Lisa Palmer is Chief Technical Advisor AMER at Splunk and Marketing Professor at Southern Nazarene University. As a Gartner Executive Programs service delivery co-leader covering the western half of the United States, she provided executive advisory services, via a team of over 45 former CIOs and senior technology executives, to over 900 C-suite and technology executives.
You may connect with Lisa on LinkedIn.
Here are three reasons why you should listen to the full episode:
Learn more about decision science and its unexpected applications in real life.
Discover what it truly means to be innovative.
Find out how to effectively apply conversational A.I. to your business.
Resources
Sign up for free here to join the Local Maximum community! You can also support the show further by giving $4/month for additional behind-the-scenes content.
Marsbot by Foursquare
Splunk, the Data-to-Everything platform
Digital to the Core by Mark Raskino and Graham Waller
Related Episodes
Episode 60 on machine learning in action
Episode 79 on creating virtual worlds
Episode 87 on deep learning and A.I.
Episode 116 on cytometry and machine learning
Episode 117 on A.I. hype
Episode 135 on machine learning and language
Episode Highlights
Cautions in Emerging Technology
Lisa stresses the importance of keeping an eye on what's possible.
Map those into what you are trying to achieve from a corporate perspective.
These ideas apply whether you are in a commercial, non-profit, or government environment.
Real-Time Decision Making with Data and Decision Science
We can now push real-time information directly to those who need to make decisions instantly.
It's better for decision-making to happen closer to the source of choices and that there is no delay.
Arm your frontline employees with current data so they can make choices immediately on what they're going to do.
Real-life applications mean taking a stream of data and turning it into visualizations that can inform decision-making.
Real-Life Applications
One application of that data is helping to fight wildfires.
McLaren Racing collects tremendous amounts of data during each race, which decision science can help sort out.
It can also be applied to grain agriculture businesses.
Possible Problems in Real-Time Data Systems
You have to consider multiple factors since a lot of things can overlap.
It would be best if you thought about all the interwoven possibilities.
Defining Decision Science
Decision science is ensuring we make informed decisions through data.
It's a combination of using the best of technology—data, machine learning, and artificial intelligence—with the best of human abilities to make decisions.
One of the core issues people have around machine learning and artificial intelligence is that it makes the decisions for them.
It's best to be transparent on how machines arrive at decisions so people will be comfortable with the outcomes.
About Augmented Intelligence and Innovation
Augmented intelligence is weaving together creative problem solving from a human perspective and applying information from machine learning and A.I.
Most people don't know what to do with that kind of technology yet, so it's hard to get support or funding for augmented intelligence projects.
Innovation requires a lot of failures.
Something that has plagued the business community is the expectation that every effort in every project has to have some application, profitability, or impact on humanity.
Not all innovations will immediately lead to something.
The Agile Mentality
In the middle of the pandemic, customers now have more realistic expectations—if you want something fast, it's not necessarily going to be perfect.
There has been a fast shift toward acceptability of something good and done quickly over a perfect solution.
To sustain an Agile methodology, you need to set expectations and be upfront with your customer.
The point of creating things in an agile fashion is you fail fast, you learn, and then you jump forward with the next piece because of what you've learned.
Applying Conversational A.I.
By using conversational A.I. like chatbots, customers can receive answers to their questions quickly.
Conversational A.I. bots help create a happier customer experience and a happier employee experience.
Dealing with more complicated questions is an evolution with machine learning. The more you train the models to deal with those situations, the better they can address them.
Conversational A.I. is where the interweaving of human and technical capability is so critical. A bot should be able to discern whether it can answer a question and redirect customers to human agents.
Closing Thoughts on Technology
Think about how we can get comfortable with the capabilities of technology.
Don't be fearful—be engaged in dialogue!
Have a voice when it comes to governance in technology.
5 Powerful Quotes from This Episode
"We can use data to inform and to help provide high quality service that impacts not only profitability but [also] the purpose—the ability of these organizations to actually create the impact in the world that they're trying to create."
"We get the very best of…all of the things that…separate us from machines…and combine that with the very best capabilities at the table from a technological perspective. And if we can get that mix right…that's when we get really positive outcomes for customers and humanity."
"It's incredibly important we have a view that as humans, we're comfortable with understanding what information is being provided with the observability elements of what happens with machine learning."
"The really important thing about innovation is that we have to give ourselves the permission to fail."
"I just want to encourage people to think about how we get involved from a human perspective to get comfortable with the capabilities of technology."
Enjoy the Podcast?
Are you hungry to learn more about how decision science works? Do you want to expand your perspective further? Subscribe to this podcast to learn more about A.I., technology, and society.
Leave us a review! If you loved this episode, we want to hear from you! Help us reach more audiences to bring them fresh perspectives on society and technology.
Do you want more people to understand decision science? You can do it by simply sharing the takeaways you've learned from this episode on social media!
You can tune in to the show on Apple Podcasts, Soundcloud, and Stitcher. If you want to get in touch, visit the website, join our community on Locals, or find me on Twitter.
To expanding perspectives,
Max
Transcript
Max Sklar: You're listening to The Local Maximum Episode 148.
Time to expand your perspective. Welcome to The Local Maximum. Now here's your host, Max Sklar.
Max Sklar: Welcome, everyone. Welcome, you've reached another local maximum. I just want to say as we count down the final weeks of this year 2020, I really appreciate everyone who has stepped up to support The Local Maximum and join our local group at maximum.locals.com. Even if you just sign up, just put your name in there, and yes, download the app maybe and check it every once in a while. But just sign up really helps. And for those of you who became supporters on the platform with a small monthly donation, a double thanks. I really look forward to sharing all the plans we have for The Local Maximum going forward.
I don't know is that the royal we, I guess Aaron is involved and you know my friends and all the people who I have on the show and the guests and as well, all of you who I sometimes talk to about what we're going to do on this podcast and the related content in the website and all that. But we have the podcast, we have the learning concepts, we have the analysis of current events that are related to everything that we're discussing. And there is like a ton of work to do in 2021. So, I thank you all for listening and I thank you for all the support you have done there, and let's talk about it. Let's make it better.
So speaking of learning, we are going to get an update on applied artificial intelligence today. And if you listen to the whole program, you are going to get a broad view into areas where AI and machine learning and all that are being deployed. And, in some cases, in places where you might not have considered those things being deployed before, so that's that that's always fun.
I know I learned more about the world from today's guest. We talked about from real-time decision-making systems to the changing nature of innovation and the changing nature of work as we move into the 2020s. And I don't think anyone denies that it's changing. But it’s still an open question how, so we introduced some new thoughts there.
And finally, we discussed the concept of augmented intelligence as it relates to things like chatbots and all that fun stuff and what that really means augmented intelligence. So this stuff, regardless of who you are, this stuff affects your life, whether you like it or not, no matter who you are. But if you follow the AI business landscape, or if you're an engineer or product professional in the industry, today's discussion will be of particular interest to you. So let's get to it.
My next guest has a broad range of experience in technology, including applied artificial intelligence, research and business at the executive level, as well as academia and marketing. And from my discussion is really someone who has their fingers on the pulse of what's happening in the world of AI, especially when it comes to applied innovation today. Lisa Palmer, you've reached the local maximum. Welcome to the show.
Lisa Palmer: It is exciting to be here. Thanks for having me, Max.
Max: I'm actually really excited about this conversation. It’s been a few weeks since I've done an interview. I was like, “Whoah, when was the last time I've done this?” But I don't do as many interviews anymore. And I'm excited about this one, because there’s so many different areas in tech and AI that we could talk about today. So, we're just going to touch on a few things. But I want to start with your background. I know this might be going back a little ways, I saw you worked at Gartner, and I always loved their hype cycle charts. So, I was wondering if you could tell us a little bit about that and a little bit about what you do now.
Lisa: Absolutely. So just to give a little bit of context, I've got three primary pieces of my background. I was IT practitioner for years and years. All across the IT realm with my last role culminating as a chief innovation officer, which was a dual role of both CIO and CMO. And then the second kind of bucket of experience was in enterprise field selling and running a geo for Microsoft, for their enterprise sales territory. And then the third piece to your point, Max was a large period of time with Gartner and focused on—specifically on helping executives to use the fantastic research that Gartner has to inform their digital strategies. So how were they going to embrace all of the fantastic capabilities that now exist and that are constantly on the horizon to keep their businesses moving forward?
Max: Cool. So this is a question—I feel like a lot of people would be afraid to ask this question because it sounds like a dumb question. But you said you were a CIO, Chief Information Officer. What’s the difference between that and like a CTO or someone like that?
Lisa: So the definition of the roles can vary depending on the organization that you work for. There's certainly anomalies. But in my particular instance, the CIO role was responsible for everything that was technology, but also all of the people and process elements that you wrap around that as well. So sometimes there's a delineation between that—specifically the tech piece, and all of the other parts that are really necessary for your overall technology strategy to be pursued.
Max: Yes, so I'm trying to think like—you're talking about incorporating emerging technologies into businesses and this is a question just thinking off the top of my head, have you run into any cautionary tales? Like what do people do wrong in this area?
Lisa: With regard to emerging tech, it's really important for us to first of all, keep an eye on what's possible. Have somebody that is focused on keeping a pulse on the capabilities that are available and emerging in the market, and really mapping those into what you are trying to achieve from a corporate objectives’ perspective. And so, whether you're in a commercial environment, or you're in a not for profit, or government environment, where you're more mission oriented, having somebody that is really keeping a pulse on what is possible, and then translating that into what it means for your organization is really important.
Max: Gotcha, gotcha. Okay. So there's a lot of topics we can cover today in terms of data science and AI. We can go all the places if you want, but we spoke earlier about focusing on real time decision making with data. So can you give us some examples and tell me why this is an interesting topic right now?
Lisa: There’s been many years now where it's been popular to have dashboards, people at the—certainly at the managerial level, have information that they're gleaning from different types of charts and graphs, where everyone is comfortable, looking at things that have happened in the past and making future decisions based off of the historical information that you see on a kind of a classic dashboard.
What I believe is really important in today's times is that with the proliferation of data that there is now, we have an ability now to be able to push real time information directly to those individuals who are making decisions live right now today. And to be able to do that allows us to inform decisioning truly at the point that the choices are being made, instead of in retrospect. Or, a month from now, when we're looking at the dashboard, we'll look back, and we'll decide that next month, we're going to do something differently.
Well, what we've seen happen through the pandemic is that things are changing at this crazy rapid pace. And so, it's really important that we get that decisioning happening closer and closer to the source of the choices. And it not be so delayed in the way that we execute against information that we glean from our data. So, pushing that decisioning down to a frontline supervisor or even to somebody that is a frontline employee, if you can arm them with current data so they can make a choice right now about what they're going to do, that's when data becomes really powerful from top to bottom throughout your organization.
Max: Yes, now that I think about it, we kind of ran into that. In my job at Foursquare, we're trying to make some—put together some dashboards for COVID real fast. And it was like, wow, this is not the type of thing that we've thought about having to do. And so we were kind of...
Lisa: Exactly the timing is so tough, right? We've got to get that real time information into people's hands. So things like if you are a member of a supply chain organization, and you have historically sent an order to Supplier X, and you actually, today, know that as a result of COVID, for example, Supplier X is completely unable to fulfill your order. Instead of you putting that order in with them as you normally would have and then waiting to find out that “oops, they can't fill that order.” What if now you have real time information that tells you that that order could be fulfilled by Supplier Z that is normally number five on your list of suppliers, but they've got it, they have it right now. And so instead, at that very moment you place the order with Supplier Z instead. And now the impact on your customers is that you are able to provide them with the same quality and level of service that they have become accustomed to receiving from you. And that's all being enabled by real time data decisioning.
Max: So let's talk about some examples, which are a little bit more of it that are just like, interesting because I have some written down here that we talked about before. One was fighting wildfires, I think people want to hear about that.
Lisa: This is a fantastic example. And I love purpose-oriented examples of how we leverage and apply data. So in this example, as we all know, wildfires have been a significant problem along our west coast. And particularly, as a result of sparks coming off of our transmission lines in that area, right? So, in order to be able to control this more quickly, imagine just taking a metal stick with a sensor on the end of it, and those being placed in the ground along the entire transmission routes. And connecting those devices through IoT, or through IoT capabilities back into a core system that is providing up to the second information about the temperature of that particular sensor. And then as a result of that, if that sensor starts to show an elevation in temperature, we can make some assumptions that there's a problem there.
So now we overlay things like weather data over top of that, that shows where the wind patterns are putting are potentially going to push, if there is a fire outbreak in that area where directionally would that fire go and now you can deploy in real time, firefighting resources to that area without having to wait to see it, to have somebody physically come across that fire, etc. So you're able to, in real time, send resources to fight fires before they begin to spread.
Max: So it sounds like it's not just a real time data stream, where I'm just taking sensor data and pushing it to people in real time. There's also some real statistical modeling going on in real time as well.
Lisa: Absolutely. So this is about taking a stream of data in and turning it into visualizations on the fly that can inform decision making. So imagine that as that data comes in, that you're seeing a visualization—and it could depend on the situation, obviously—but a chart, a graph, some kind of a visual representation that shows people, “Hey, there's a problem right here.” So, this is how we get engaged.
And then literally on that visual, you've got a map that overlays what the wind patterns are, and you can physically see exactly where your threat is from a wildfire fighting perspective. So it's incredibly powerful to combine that real time data with time visualization.
Max: Is there anyone—and it could be wildfires, it could be something else—but is there anyone that comes to mind that implements something like this really well right now?
Lisa: So okay, Max, let me jump into an example of McLaren racing. So McLaren, obviously F1, fantastic racing organization. And they are recording petabytes of data during every race. This is a massive amount of data. And they're using every piece of equipment inside of their cars are intelligent. Every piece that's intelligent is sending this sensor information in through Splunk. And we're able to visualize for them what is happening with every piece of that super high-quality car, to enable this race team to help them to perform at their absolute maximum best quality. I mean, racing...
Max: Hold on. When these races are happening, are you allowed to, like gather and use real time data from like car and send it to the cloud back and all that, is that legitimate?
Lisa: Absolutely. Racing is absolutely a technical business in today's times. So there's an adage today that “every business is a technology business.” And I think that racing is a perfect example of how that has evolved through the years, and the McLaren team is just absolutely fantastic in this space.
Max: I haven't even thought about that.
Lisa: Yes. There are so many behind the scenes activities that happen with organizations that probably don't even immediately pop to mind. I have another example of an organization that is a grain agricultural business. So, imagine that they think of themselves as wanting to be the Amazon of the grain delivery business. Well, we don't necessarily think of agriculture every day as a highly technical business. And it absolutely is.
So those are the kinds of things that we can use data to inform and to help to provide high quality service that impacts not only profitability, but the purpose—the ability of these organizations to actually create the impact in the world that they're trying to create.
Max: What problems are unique to these real time data systems?
Lisa: Well, obviously, connectivity is huge, right? I mean—so there's a lot of conversation around 5G, and what are the 5G impacts going to be around things like real time data. And there's so much overlap in technology. So when we think about real time decision making, we've got to take into consideration things like, what are the capabilities of IoT? What are the capabilities of 5G? What are the capabilities of machine learning and artificial intelligence? And so, it's not just that we can take one of these technologies and approach it in a silo, we've got to really be thinking. So back to your question about emerging technologies, we've got to be thinking about what are the interwoven possibilities of all of these technologies?
Max: All right, so let's go to some other topics to get through, let's just go through them. So, the term data decision science has gained in popularity. What does that term mean? within our pandemic context, in particular?
Lisa: Yes, sure. So decision science is about making sure that we are informing the decisions that we make with data. And I like to think of it as a combination also of how are we going to make decisions by using the very best of what data machine learning artificial intelligence on the technology side, we've got to bring the very best capabilities of that together with the human abilities. And this is really an augmented intelligence approach. We've got to bring that together, where we get the very best out of human creativity, empathy, all of the things that really are core human capabilities that separate us from machines.
And we want to take the very best that we bring to the table and combine that with the very best capabilities at the table from a technology perspective. And if we can get that if we can get that mix right, where we've got the best of both being brought forward, that's when we get really positive outcomes for customers and humanity.
Max: I feel like a lot of machine learning algorithms, at least the one that I work with, they don't make decisions. And that's what people don't realize, they don't make decisions, they only make predictions. And it takes another step sometimes to turn those into decisions. And sometimes you have to think about what decisions you want to make based on what it says. And it's something that you have to do pretty carefully because it's not always obvious.
Lisa: It's absolutely not obvious. And I think a lot of the fear that people have around machine learning, around artificial intelligence—a lot of the fear is based on this, this concern about the actual decisions that are being made as a result of the information.
And so, it's incredibly important that we have a view that as humans, we're comfortable with understanding what information is being provided with, kind of with the observability elements of what happens with machine learning. We got to elevate the comfort level that people have. And I think you've just nailed one of the core issues, Max, that people are concerned that it's making decisions for them. So we gotta make sure that it is transparent, that what is happening is transparent and that people begin to gain comfort with that. And that in and of itself is not a simple solution. It's not a simple problem to solve because...
Max: And there are a lot of times when you're designing something and it's like, “Well, this is a Blackbox, that makes the that does the best on some metrics and training data,” let's say you're building like, a neural net, a deep learning system. And then, one of the downsides is people like, well, I don't know what it's doing.
Lisa: Exactly. And then you have a resistance to adopting whatever capability you've brought forward for your organization because people are anxious about “what is it doing? How is it deciding these things?” And that is an issue that has to be addressed for people to be comfortable with the outcomes that are created. So, it's a challenge that we have to take seriously. And as individuals trying to get this kind of capability embedded into your organization, that is a key element that you have to address.
Max: Okay, let's go to the term augmented intelligence now. And is this related to augmented reality? What is the difference?
Lisa: So, augmented intelligence is taking the best of both worlds. Let's make sure that to our prior conversation there around making sure that we understand how decisions are being made. Let's take the best of that, and the human creativity element and make sure that we're creating a cohesive alignment between things like creative problem solving from a human perspective, and the way that we apply the information that comes from machine learning and AI, that the way that it's applied, this kind of creative inner weaving is where augmented intelligence really plays. And if so...
Max: If I could try to summarize—tell me if this is correct, I'm not sure if it's correct. But it is trying to combine machine and human intelligence and while using the strengths of each.
Lisa: Yes! Absolutely. I like that, Max.
Max: I got it.
Lisa: I like that. And I'd love to hear your perspective about AR.
Max: Oh, yes. Well, I mean, it's basically related to the product that we just put out at Foursquare called Marsbot, where we're kind of using an audio platform to build some sort of an audio AR of the city where you walk around. And in real time, as you walk past a certain storefront, we can play an audio file that's attached to that storefront.
Now, it's really cool but I don't know what people are going to do with it yet. So, we're at the phase of that technology, where it's like, “we can do it, but what use is this?” And that's what hopefully that’s what people figure out. Right now, we're just trying to inspire people that, “Hey, we can do it.”
Lisa: And you know, Max, that's a great example of how AR and our augmented intelligence work together. So that's where the human creativity comes in. Like, “Here's this really super cool technical capability that we've created. Now, let's get a think tank group of people together and come up with some really strong use cases for how we can apply this.” And that's where the human creativity element plays in.
Max: Yes, I mean, one thing we're finding is Foursquare has lots of really great information about places that we've built up over the years, in a lot of it, honestly, has been, like, 70 years ago. Because it's kind of an older company where we can tell you, “Okay, what’s the highest thing to order at a cafe when you walk by the cafe? What's it known for?” Okay, that's interesting. That's something a machine can do. But I think a human could do a lot better. Like if we hire a comedian or an entertainer or something, or a musician or someone like that to give you an augmented reality view of the city. I think there are certain things we could do with the machines that are okay. But we’re finding we're up against the wall that only humans can solve at this point.
Lisa: And that's actually a great place to be because you're right on the cusp of you don't even know what is coming until you get that group of people together and brainstorm really, truly bring creativity to the table. And we see technologies either take off or die when they get to that point.
Now, sometimes the human creativity is leading the creation of technology, and we've already got some really strong use cases in mind when the technology is created, but that's not always the case. Sometimes we create technology and then find fantastic applications for it.
So, if we think about things like being able to put full books or resource elements into surgical operating rooms that a surgeon can wave their hand without ever touching anything and move forward pages in a book, or scroll down a web screen for some specific information, like literally in the middle of a surgery, that use case for AR technology certainly wasn't in the minds of those that created it. That was an application that came about after the technology already was created. And, there are things that have happened in history that model plays out over and over again. So, I'm sure there's cool applications right on the edge out there for you guys to get to, Max.
Max: That's one of the hardest things I found, a lot of times that the stuff that engineers in particular, were interested in building, if the applications aren't clear yet and perhaps rightfully so it's a lot harder to get support, whether that's internal supported in a company or funding to do these types of projects.
And sometimes it's good to switch off between something like that, and something that's like, obviously needed now that needs to be built. Or at least that I found kind of like a one for you one for me type approach to the career. But how do you hold these conversations ever where because there are a lot of times where, as engineers, we build stuff that we think is cool, that someone might use, and then it doesn't get used. So, it's that can be frustrating at times.
Lisa: But you know, the really important thing about innovation is that we have to give ourselves the permission to fail. Because innovation—really, truly being innovative requires a lot of failure. And that means that it may not ever actually be applied or take off in some massive way. However, I mean, we can look back to Edison in history, right? And in the thousands, of what he called failures, before coming to some of the most critical discoveries in history. So that is something that has plagued, in my opinion, the business community that there is an expectation that every effort in every project is going to have some application, some profitability, some impact on purpose or humanity.
And that's just simply not the case, we've got to elevate our willingness to accept that when we create things, that when we're truly being innovative, not every one of those innovations is going to lead to something high impact. However, it could and thus impact the body of understanding, the body of information, the body of work and knowledge that could create the next really big impact thing.
So, we've got to embrace a little bit of failure to make sure that we're really maximizing our innovative capabilities.
Max: Yes, I've been surprised—I've always run into the mindset of we have very strict priorities based on what the current needs are and we're always going to work on the number one priority. And seems like, “Hey, that should be the most efficient team.” And then it's always surprising when over the long run, it seems like, “Oh, it's not,” because it's, it's kind of you get kind of tunnel vision into whatever your current clients want, for example. It's like, well, yes, we built a ton of stuff that clients want in 2017. But now it's 2018 and we could have been building for the future.
Lisa: Yes, exactly. So I think that's it's important as we move into the next era, and I think that this idea of super fast innovation that's not perfect has been embraced during this pandemic period. We are seeing that...
Max: Why do you think that is? Because I feel like a lot of a lot of stuff—I'm sorry to cut you off—but I feel like there's been a lot of times in the pandemic period where people are like, “Oh, it's an emergency, you have to do this, this and this.” But it does seem like you're right, that there is kind of a shift in thinking toward—like you say. So what do you think's going on?
Lisa: I think It's a matter of done, available is better than waiting for perfect. Right? So, as we've been in the middle of this pandemic, period, customers are more accepting of the fact that it may not be perfect, but we're going to do the best we can to provide you with what you need. And so the customers have gotten a little bit more realistic about the fact that if you want something really fast, that it is not necessarily going to be perfect, but we're going to get it to you. Right?
And so I think that when society as a whole starts to be more open and accepting of done and provided is better than, “Oh, wait, we'll get you a perfect solution but it's going to be three months or six months from now.” We've seen a fast shift towards this acceptability of good and done is better than perfect and not here yet.
Max: Yes. It's the incremental approach is—so first of all, it's like too often more exciting because you're getting things done faster. But I think we've all seen situations where you're working on a project, but like, you can't show weakness to the customer. So you got to push it off, like a year until every “I” is dotted and every “T” is crossed. And by then the priorities could change completely.
Lisa: Right. I think that we've seen a fast forwarding of agile methodologies and agile mentality during this pandemic period. And I think that that is going to survive past this disruption period. I think that we've recreated expectations, and so from a technologist perspective, when you're having conversations, it's incredibly important that we set expectations up front.
So what if when you were in one of those situations with a client, what if you had the conversation up front, that if you hire us to do something high end innovative, then it's got to be fast iterations. You've got to be giving us feedback on a daily or weekly basis. And every expectation is set that this is going to be highly interactive, and that you are going to be behaving in an agile fashion. And that is how you are going to avoid the big failures that are possible if you wait to give a deliverable for six or 12 months. Because you might have a minor failure on a day or weekly basis, which you will have by the way when you're doing things highly innovative, right?
Max: We’re well aware.
Lisa: But then we have micro failures that we fast forward through and then the customer is engaged in that process the whole time. But you have to set expectations that it's not always going to be perfect and fruitful. And that's the point of creating things in an agile fashion is that you fail fast, you learn and then you jump forward with the next piece because of what you learned in the last effort.
Max: Okay, great. So, let's talk about conversational AI. I've created a chatbot once, that was actually the first version of Marsbot four years ago. It's kind of a benefit of where I work, I get to work on all these cool things. How can this be applied in B2B scenario?
Lisa: So conversational AI has so much, so many capabilities. And I'm always intrigued from a consumer perspective. I was on a website yesterday, obviously consumer website, and a chatbot popped up and was able to very successfully answer the three questions I had, in a matter of seconds. And it was incredible. And to see it applied in such a practical fashion, I got very quick answers to my questions, I came away a very happy customer.
And what happens on the backside of that, if you're the business provider, is that you did not have to have a cued up client because my additional experience in the past with those kind of customer service situations is and certainly during the pandemic, you could be waiting for over an hour. I was on a wait period one time for 72 minutes before somebody answered the phone. Right? That does not create a happy customer.
Max: No. And then the phone conversation is not happy either. Because you're kind of mad, or tired, or cranky.
Lisa: Right and being the person answering that call. And all day long, they're answering calls from people who've been on hold for over an hour. I mean, my heart goes out to these people.
So now instead of those individuals spending their time dealing with angry customers. Now you can re-allocate those customers to doing proactive customer service activities, customer delight activities, instead of them dealing with frustrated customers.
So now, by using conversational AI, you've accomplished a couple of really important things. You have made a very delighted customer, obviously critical because retaining customers’ needs to be the mantra of every organization. And you've also created a massive positive impact on your employee experience. And there are now mounts of research about the impact on your business of creating a positive employee experience. So, I think conversational AI has some fantastic capabilities in this space.
Max: So let me let me ask a question about that because I feel like there are certain types of questions that I expect a machine to be able to answer, and then certain types of questions where I don't even want to try. So, let me try to give an example. Like, let's suppose I'm doing online banking, okay? And I want to ask, “Hey, what is my balance? How much did I spend last month?” or something like that, that I expect to be able to do. But let's say I'm like, “I was trying to transfer funds, and there was an error on the website, and I don't know where it is right now. And someone told me this, but this doesn't make any sense. I'm like, I don't want to speak to a machine about that.” And sometimes it's like, I don't really know what the capabilities are of the chat bot? I don't know if we could deal with kind of like a delicate or complicated issue.
So how do you? Is there like a— I don't know. I feel like there might need to be a customer education part of it, where you know to ask the simple questions and the complicated questions. I don't know how good we are at the complicated questions. I don't know what to do about that.
Lisa: So I think the complicated questions, it’s an evolution. And with machine learning, and we know that the more of these situations that we're able to train the models on to deal with, the better that we'll be able to address more complicated questions.
But the reality is that this is where the interweaving of human and technical capability is so critical. So, on the chatbot example that I was sharing earlier, I had a fourth question. The fourth question was one of those. And I always ask the chatbots the fourth question, even though I highly doubt they'll be able to answer those high end critical questions because I want that question to be part of the data that they're training their models on.
So I make sure that I put the question in, and you know, what happens? It pops back and says that, “This isn't a question they're going to be able to answer for you please contact customer service, or would you like to speak…” And then they pop up another window with a live customer service agent, so you can interact, right?
Max: That’s a good solution, yes, rather than just breaking.
Lisa: Exactly.
Max: So that's probably the best if I were designing it like that you would want it to send you to the person before breaking down and just rather than taking a few key words out of that complicated question, and then answering something simpler. That could be very frustrating.
Lisa: Right. So the ones the bots that are doing well, right now, in my opinion, are able to discern when they can't answer the question and when they need human intervention. And then they're redirecting you to humans, giving you the option to either interact with a human via chat, or would you prefer to call? Or would you prefer us to call you? They give the customer the option of how they want to handle their next step.
And that same ideology is being applied in B2B environments as well, where a lot of the common questions going back and forth from business to business are also things that can be either answered really quickly, or they're too complex to do that. I do like to, as somebody who's a student in this area, I do like to make sure I ask the questions, so I can help to improve their models over time. Even when I feel like it's probably not something they're going to be able to answer. I always ask the question.
Max: Yes. Well, then there's always the model, like you said, of do I then go to the real human or do I not, and so it's certainly going to help with that. And that that just what you just said sounds like a perfect example coming back to like augmented intelligence. That's exactly what it is but perfect blending.
Okay, so you're in Tulsa, I just wanted to talk a little bit about that. There are probably a lot of us in like New York and San Francisco right now who might be curious about living in a smaller city. And I don't know what it's like to work in tech out there. But can you tell us a little bit about that?
Lisa: You know, it's interesting, I have spent most of my career living in Tulsa and been in the tech world in one way shape or another. And it actually provides me with a lot of benefit, obviously, being in the Central Time Zone is pretty fantastic place to be so that I can easily interact with those from both coasts. I'm the only employee for Splunk in Tulsa. So which pre-pandemic probably would have sounded like an odd thing. And now it doesn't even sound strange to people because we've seen the workforce shifts with support from employers, for people to be located physically anywhere. So, it's not as much of a surprise for people about where I live now as it used to be.
Max: Yes, now there are people who are a mile away, they're still on the little screen up front there.
Lisa: Exactly.
Max: Yes. All right. That's really cool. So I think we're about ready to wrap up. Do you have any closing thoughts? And where can people find out more about you?
Lisa: Sure. So for the closing thought perspective, I just want to encourage people to think about how we get involved from a human perspective to get comfortable with the capabilities of technology and don't think of it from a fearful place.
In a lot of conversations that I have people are concerned about what's going to happen with machine learning, with artificial intelligence. And instead of being fearful, let's be engaged. I want to see people get engaged in dialog. Learn what it means to have those inner woven capabilities between technology and humans, and have a voice with regard to governance around technology.
There are three core tipping points. And this is actually a book that was written by a couple of incredibly smart gentlemen from Gartner called Digital to the Core. And the book’s a few years old now, but the tenants or are so critical. Three tipping points, the technology's got to be capable, the regulatory inbox environment has to be conducive, and society has to be ready to embrace the change.
So along that continuum, we need to make sure that we are embracing technical capabilities, that we are creating a regulatory environment that makes it good for humanity, and that we as a society are deciding where do we want to apply tech and where do we not. And so those are areas where every individual and every leader can begin to educate themselves around those three different tipping points.
As far as being able to get engaged with me, I'm really active on LinkedIn, it's probably the best place to interact with me. I actively connect with people, chat with people offline. So, if I can ever be of help to you or your listeners, please feel free to contact me via LinkedIn.
Max: All right, great. And all those links, including your LinkedIn will be posted on the show notes page at localmaxradio.com/148. Lisa, thank you so much for coming on the show today.
Lisa: It's been a pleasure, Max. Thanks for having me.
Max Sklar: All right, I had a great time with that. I'm now counting down, we have three episodes until the end of this year. So, I've got to figure out what I'm going to do with that three. It'll probably include a December news update. And I don't know if we're going to do this, but it seems like the turmoil at Facebook and Google have taken another turn. So maybe we'll have another go at that. I know we've done a lot of that but I've never gotten any pushback so I assume people like that. But let me know.
And of course, one of them is going to be a look back for the year. So that might be an order. And maybe something else, maybe I'll have another guest, I don't know. But it's exciting. We're counting down. Have a great week, everyone.
Max Sklar: That's the show. To support The Local Maximum. Sign up for exclusive content and their online community at maximum.locals.com. The Local Maximum is available wherever podcasts are found. If you want to keep up, remember to subscribe on your podcast app. Also check out the website with show notes and additional materials at localradiomax.com. If you want to contact me the host, send an email to localmaxradio@gmail.com Have a great week.
