Poole Podcast Season 2, Episode 4: How AI is Driving Data Decisions in Business, with Haroon Abbu and Bill Rand
Artificial Intelligence (AI) can be leveraged in business for so many things including data analytics and enhancing the customer experience. But there are concerns about AI as well, such as loss of actual human jobs and facial recognition bias. Today, we dive into AI and how it’s taught in the classroom.
Dr. Haroon Abbu is the vice president of Digital Data and Analytics at Bell & Howell and the co-author of TRUST: The Winning Formula for Digital Leaders. A Practical Guide for Companies Engaged in Digital Transformation.
Jenny: All right, Bill, let’s start with you. Welcome back by the way, it’s so nice to have a returning guest in our second season of the podcast. You’ve been busy this past year building and working on the mission of the Business Analytics Initiative, which we affectionately call the BAI here at Poole College.
Could you give us a quick review of what the BAI is and what are some of the things you’ve been working on in year 2?
Bill: Yeah, well, first of all, it’s good to be back. Thanks for having me back on the podcast. So we have been having a lot of fun with the BAI and kind of getting everything set up and getting started. And just to quickly recap, you know, the Business Analytics Initiative brings together all the great analytics work at Poole College.
On the education front that includes our undergraduate business analytics certificate, our graduate certificate in business analytics, and our masters of management and marketing analytics and on the research front we’re helping to coordinate the state of the art research that’s being done by our world-class scholars here at the college when it comes to business analytics, and on the thought leadership outreach front, we are assembling a top-notch advisory board that represents a diverse set of industries.
So ruined here will be a member is a member of that, and we are really committed to helping them understand and kind of talk about cutting edge analytics and data-driven decision-making. In terms of next steps, we’re launching a masters of management risk and analytics this fall, which comes out of our prestigious enterprise risk initiative here at Poole College and the great work being done in the accounting department there around risk and analytics.
And we are planning our first business analytics roundtable, which we affectionately called the BAR or the Bar if you are interested in joining us at the Bar, right? We ask that you kind of look for the materials that we’ll be putting up on our website and it will be launched and hosted on May 19th here in Raleigh at NC state’s campus.
And we have lots more great opportunities that we’re working on for industry and potential students to interact with the Poole college and interact with the Business Analytics Initiative that we’re, we’re planning for the future. So lots more to come, and we’re really excited about all the possibilities of things we could do in this space.
Jenny: Yeah, it’s a busy time for you guys, but as you mentioned, very exciting. So Bill, you mentioned that Haroon, our guests joined the inaugural advisory board for business analytics and somehow got talked in to teaching an MBA class this semester on artificial intelligence. So Haroon tell us a little bit what, why and how you got kind of drawn into Poole college?
Haroon: Yeah. First of all, thanks for having me. It’s great to be with you today. Bell and Howell and Poole College of management we have been working together for a number of years, actually since 2014. It began with CIMS center for innovation management studies. That was then part of Poole college of management.
And Bell and Howell was an advisory board member of CIMS. So it was a great industry academy collaboration. Our company was in the cusp of transforming from manufacturing into services first company. So we needed a structured way to evaluate new ideas. For business impact, and that can match to our capabilities and strategy, right?
So we use innovation management process that CIMS has developed and changed it a bit to reflect our own unique business needs. And then partnering with NC state, we have done multiple workshops on innovation management. And today we have institutionalized that process internally, so we have a 90-day evaluation process from ideas to go no-go type of a project decision.
So this has resulted in significant business impact for us, both for our business model and new services model. Right? So in addition, we have participated in number of MBA practicums, projects. It has helped us onboard some good talent from NC State. So when Bill and Christine presented us the mission and objectives of business analytics initiative, BAI, our CEO COO and I quickly realized that we shared the same mission, right, which is data innovation leadership to tackle today’s leadership challenges. So I’m happy to be part of the inaugural advisory board of the BAI. It’s already working out great. Actually, we have two sets of business analytics honor program students doing practicums with us at the current time. It’s a semester long practicum.
I mean, it’s amazing to get their perspectives on some of the things that we’re working on. As you mentioned, yes, thanks to Bill I’m teaching an MBA class on artificial intelligence, with the same spirit of partnership, bridging the gap between education and industry, I’m really enjoying it.
Jenny: Bill, why is it so important to have industry experts like Haroon in the classroom, especially in a space like analytics and AI.
Bill: Yeah. So I think Haroon did a great job at hinting at some of the reasons there, which is that, you know, in the educational world, you know, we’re of course great at teaching the theory and the concepts behind analytics and AI, but it’s really only through these kinds of practical experiences, whether they be through the practicums, and all of our programs require a practicum that Bell has been part of for quite a while now.
in addition, bringing in outside experts like Haroon to kind of teach directly into our program about the concepts of AI and analytics, these are ever changing fields and our faculty general do a great job staying on top of the current trends and the cutting edge in these fields.
But Haroon is able to discuss not only what is theoretically possible in the space and what is, what the research says, but also how it’s being done in the in industry. You know, they could talk about questions like what problems do companies actually run into when trying to implement particular methods or particular techniques, which techniques, even though they may be similar, theoretically, work better in practice for different areas.
And finally, in some ways, most importantly, our students are constantly interested in kind of the organizational and governance nature of analytics and AI, like who makes the decision to implement it? Who has to do you have to get buy-in from, even all the way down to what are the job titles that we should be looking for in this space of AI and analytics, right?
And a new field like this, things are organizations changing, even more kind of organizational front, then they are in some ways on the technique and methods front. And Haroon is able to speak to that within his own company and his own organization, but also to similar realizations and industry wide practices that exist in this space.
Jenny: Living the think and do spirit. I know that I had to throw that in there.
Bill: Yeah. Yeah. 100%
Jenny: Tossed up the softball, had to take it. But Haroon, I’d be curious, so you’re teaching this course, this semester with the MBA students. If, if Bill and I were joining this class, what are some of the things that you would be teaching in the AI space?
What would we be learning?
Haroon: So this course is about AI and its applications to help make data driven decisions in business. Things are changing at a rapid pace, so students are going to learn about, you know, the introduction to AI, what does AI mean? Discuss different applications of AI and work with some of the actual AI tools to develop new insights.
One of the big things that, you know, I’m talking in the class is no code AI platforms, which means that, you know, you don’t need to know R or Python or coding languages, but how do you use some of the commercially available artificial intelligence platforms out there to do artificial intelligence development, as well as deploy and scale them.
Right. So in a nutshell, students will be able to explain the importance of AI for business, get familiarized with artificial intelligence, terminologies, machine learning, et cetera, and how they used in management, understand how to use AI to get gain insights, to make a business decision. And then there is a lot of talk about governance as Bill pointed out, governance models, AI ethics, and how do we minimize bias in artificial intelligence, as we develop algorithms and implement them.
And at the same time, students will be able to develop the skills to present an actual AI use case and analysis to their idea of stakeholders. In addition to that, we have some guest speakers as well. These are industry experts who will talk about, you know, deploying and scaling AI in their companies. So I would say it’s a good introductory course. It’s a good mixture of technical and management type of a course that would peak your interest and curiosity on AI topics and maybe motivate you to take more AI courses or maybe even start a carrier in artificial intelligence, machine learning or data science.
Jenny: And this kind of transpires a little bit to this question, I’d be curious to see what you both think about this. What are you seeing as some of the major trends in the AI space and what industries are currently doing well with AI?
Bill: I would highlight one of the trends Haroon already mentioned, which is no code AI, right? This, the idea that we can take a lot of the kind of basic machine learning algorithms, which are still extremely advanced. And then imbed them into a tool, almost like Excel. Right? So that, so that a company can basically, take us at a data that they have, put it as an input to these algorithms and get a result out of it.
And that allows people who may not be, you know, computer programmers may not be that sophisticated in terms of their ability to work with these toolkits to develop AI solutions for their companies and their organizations and, and, and, and solutions that are not just one-off but that can be repeated on a regular basis. and so I think that’s a trend that I definitely see continuing to accelerate.
One other trend in the kind of actual, in the terms of the techniques and the cutting edge I’ve seen kind of growing is the use of more and more advanced, deep learning methods.
Right? And these are the things that allow, for instance, computers to see, for them to take an image and identify objects in it, right. Or to take a passage of text and understand the concept behind the passage of text. and in particular, there’s a technique known as generative adversarial networks. I’ve been following for a while. That’s very interesting.
These tools allow AI to not just do simple classification like you know, for instance, determining if this email that was just written into me into my customer service line is on a broken product or that they don’t like the way the product actually works or, you know, classified in different ways to direct it directly, which is kind of a current application of AI.
But generative adversarial networks or GANS actually allows computers to start creating content. So we’re starting to see computers that can create, for instance, a social media post. Right then are having as much engagement, if not more than human posts on the same topics. And I’m really excited to see where that continues to go in the future.
Right? Can we use these tools to help marketers do a better job in some of their spaces, right? Or to help, you know, HR do a better job in terms of writing postings for future job listings or, you know, all of these kinds of ability of the computer to work as a creativity engine that then compliments the human who’s making the final decisions in a lot of these cases.
And I would, you know, in terms of what industries are using this, I would say almost every industry is starting to embrace AI at some level, from healthcare to finance, to manufacturing, but you know, even given the examples above my own field of marketing, I feel like has really embraced AI, you know, the move toward data driven decision-making and, and marketing started 10 years ago and around there’s more and it’s continuing to trend, right?
this is partially due to the complex decision space in marketing. Right? Marketing is a field where the old Wannamaker quote is that half the dollars I spend on marketing are wasted, I just don’t know which half. And, and trying to understand where that money is actually fruitful and useful is very difficult to do.
But the other aspect of marketing that makes it really right for AI is the fact that there’s quick feedback, right? So we quickly see, are we getting more purchases? Are we getting more sales? Are people clicking on this website? Are people engaging with the social media posts, right? At some of the other industries, you may not see that as much, right or as frequently. And so it’s going to take longer for AI to prove its value in those spaces, but I, I feel like almost all the industries are embracing it to the extent which they can.
Haroon: No. I agree. I think that’s a very good overview of the various fields of AI.
Jenny: And I have to ask this question because I have been a marketer, everything you said Bill was absolutely true, but I’ve heard people say, I guess, people who aren’t as tied or invested or they don’t understand all the power of AI, that someone had made a comment several weeks ago, to me that, well, AI is just removing the human.
There’ll be no need for a human anymore. What do you, what are you to say to that when, I’m sure you’ve people have said that to you as well?
Haroon: Yeah, no, that that’s, that’s mostly hype. You know, the AI is used for augmenting human employees basically. So it’s automating some of the repetitive tasks, robotic process automation, et cetera, et cetera. So that’s where AI will be most useful, to augment human intelligence, not to replace human intelligence.
I’m sure. you know, bill has a lot to say on this topic as well.
Bill: Yeah. I mean, I like to think of AI, not as replacing humans or jobs, but instead replacing tasks.
So what, what tasks are we currently doing that don’t really require the human imagination and human creativity to really start to take apart, right? What are, what are kind of repetitive tasks that can kind of just be replaced automatically?
What we’re going to see, I think in the future is AI replacing a lot of those tasks that are not necessary for a human to do right, or that are hard for humans to do, right. So for instance, searching a large set of a large database, right. For a particular past image that look like an image you’re trying to find, right?
Like that’s not something we need a human to do, and we’re not very good at doing it right. Whereas an AI can replace that task, get help, us, get supplement us can start to make our, our dollars that we’re spending on human resources, even more valuable, Right, because we’re augmenting those humans with computers around them to help them be better at their jobs.
Jenny: Yeah, I think if anything, it, it, it provides a level of efficiency that we don’t even know we needed. Haroon. I want to shift a little bit when I was doing my homework on you, I read in your bio that you recently completed a research project about the bias in AI.
What is that? And what are companies doing to avoid that bias moving forward?
Haroon: Yeah. I mean, it’s a good question. It’s one of my favorite topics as well. Yes. my coauthors and I have done a lot of work on digital leadership and its role on successful digital transformation. So one of the key aspects of that is ensuring ethical AI. It’s a key dimension of digital leadership, right?
AI makes us inroads into various domains. You know, some of those Bill talked about, human resources, criminal justice, healthcare, autonomous transportation, et cetera, the credibility and trustworthiness of AI takes the center stage. You know, the surveys have said that consumers are more likely to choose services from a company that offers an ethical framework on how the data and AI models are built and managed.
AI or machine learning is only as good as its underlying model that can be easily tainted, sometimes unconscious bias of the program or the engineer who’s writing the model or who is creating the model creeps in. Sometimes the bias can come from incomplete data or incomplete training data.
In both cases, it can create unintended results. Right? There are multiple examples of how facial recognition was used, which raised some ethical concerns, even apple card credit limit, you know, the, between husband and wife, husband got higher credit limit compared to wife there was, you know, an FTC investigation into that.
So industry is actually taking a note. There’s actually a new ISO standard on artificial intelligence, that articulates current best practices to detect and treat bias in AI. There are a number of their government agencies, international agencies, even FTC is stepping up their game to ensure trustworthy AI, right?
So at the end of the day, it’s up to senior management and may even be at the board of directors’ level to ensure that AI models are unbiased. The recommendations these models make are indeed fair. Fairness has to be a conscious endeavor, and some of the steps that we can take to reduce bias or companies are taking steps to reduce bias would be to make sure that a data set is complete, it’s free of bias.
And the other important thing is to reinforce diversity and inclusiveness within the AI teams. So as long as the AI teams are diverse, various aspects and various perspectives are considered in building a model and interpreting it. And there’s also a talk in some companies to establish institutional review boards, IRBs just like what educational institutions have so that these IRBs can actually ensure that there is some kind of an ethical review of algorithms, making sure that you know, they’re not biased. They are explainable and they are transparent.
Bill: I think what Haroon’s talking about is, is, is makes him, if he makes a great point, right that bias is going to be increasingly a concern within AI and its use in industry. And in fact, we’re doing some research right now that kind of explores consumer reactions to biased algorithms in industry.
Right. and one of the early findings that we have is that it’s very important to have something like a watchdog, like an IRB or something like that that’s looking over the algorithms and whether that be external to the company or not right, is it’s a matter to decide in the future, but, but it’s, it’s critical.
And the reason why it’s critical and the reason why it needs to be trusted by consumers is because, consumers only proceed the outcome of these decisions. They don’t actually perceive the actual algorithms themselves. And so if an algorithm consistently is making a decision that might seem to be unfairly biasing, one group of individuals over another, like for instance, the credit cards being offered, higher limits to men over women, right.
Then, you know, people will be upset about that. They will have very negative reactions, they’ll call for boycotts of that company, et cetera. And it could be that the algorithm is just bad at making decisions, right? It’s not biased at all. It’s just inherently a bad algorithm, right. As to how it makes decisions.
And we need some sort of internal review board or a watchdog or something like that to come in and check the out rooms to provide evidence to the consumers that that is not the case. otherwise I think it’s inevitable that companies that are making fair and equal decisions wind up facing backlashes against them for just having bad algorithms and bad decision making in their systems.
Right, and that companies that may be, you know, have actually worst algorithms kind of hide among companies that are just have bad decision-making right. And so having some sort of check and balance on the bias in algorithms, I think it’s an inevitable next step in the, in the future.
Jenny: I think about that a little bit more too, in the sense of, and I’ll, I’ll put it in a layman’s term, it’s kind of like a bank, right. You look down and you see the FDIC, you know it’s been held accountable for different things. I don’t want to say we’re years away from it, but I don’t think people quite understand that they are being put into an algorithm and that they are being assessed. it’s a little big, brother-ish a little creepy sometimes.
You know, insights on kind of where you think we are as a society of making sure that that is something that we’re leaning toward and someone is going to say, or some entity is going to say, this has to go through this in order to be here. Where are we in that? Do you think in the sphere of, of AI?
Haroon: I think it’s happening; it’s happening fast. The standard that I mentioned ISO standard is very recent as a 2021 standard. So as industry, the government agency is international agencies are taking a look at it, it’s getting more and more prominent that bias or making sure that the algorithms are explainable and transparent, at least internally is you know, becoming more and more important.
And as the more cases of, you know, like the Apple card, Goldman Sachs being investigated by FTC. These are high profile cases, that other examples of that, and it’s, it’s again taking the center stage. So I think we are at a point where in the next one to two years, there will be some regulations around it, and then, you know, the people, citizens are going to take note and then it will become more important for the company’s reputation.
So they will build that trust in AI into the companies, you know, top goals and mission statements so that, you know, that trust for me, it’s not just organizational styles trust. It’s also trust in algorithms.
Bill: Yeah, totally agree. I think we’re seeing more and more high-profile examples of algorithm bias becoming problematic right, so the Apple card examples, the gender bias with respect to the credit card limit being given. Google ads was accused of being racially biased because it was only showing high profile jobs to, more, was more likely to show high profile jobs to non-minorities right. To white people, essentially. There were concerns with Amazon’s recruitment tool being biased against women, right.
Haroon: Number of examples on Facebook algorithms, right?
Bill: Yeah. Yeah. Facebook algorithms, right? Like all of these, like they keep coming up. Even like the, the algorithm that Britain used during the pandemic to determine the school placements right, of students. Right, and not to mention all the, the use of algorithms, in crime that has been criticized, like PredPol is being, a predictive policing tool, that’s that had many accusations of being racially biased.
I think it’s almost inevitable that enough of these events will happen that the demand almost comes from the bottom up. And we’re, you know, we’re starting to see legislative interest in this as well. And I think that unless industry comes up with a solution on its own, right? Currently the standard is that the legislatures have decided that it’s up to industry, to self-regulate in this space. right?
And I think that if industry doesn’t come up with a, a kind of solution that solves this, then regulation, it’s almost inevitable. And, and we’re starting to see some of that in Europe and other places as well.
Jenny: Bill the BAI had the opportunity to work with the Clinton health initiative to develop an AI tool, to help prepare countries for COVID-19 support. As you’re building the center and initiative in the college, is there a wish list of other things that you want to do or partnerships you want to have that you think could not only benefit BAI, but even maybe a student experience.
Bill: Yeah. So that’s a great question and we’ve been thinking a lot about where our research initiatives really are focused. And so, we find in our education side, we try and teach all aspects of analytics. But on the research side, we really have to take advantage of the resources we have here.
And one are the resources that we have is that we’ve really developed a suite of tools in the BAI to work with unstructured data in an interesting way and convert it to structured data. So by unstructured data, I basically mean anything that’s not in a standard database format. So, pictures, audio video, these kinds of things that we can’t just put in a roll in column format right away.
And our work with Chai, with the Clinton Health Access Initiative was exactly that we’re scanning social media and we’re scanning news reports. We’re scanning scientific journal articles about COVID and kind of helping them quickly prioritize what pieces of information they need to make aware to their members as quickly as possible.
Right. And so I would say my, my wish list is really to work with more companies that are interested in the app. They’re interested in, how do we take this large amount of data that we know we collected and we don’t know how to use it, to actually make a business decision. And that’s not a particular industry necessarily but it is a, is a problem that we’re looking to solve.
We’ve had a lot of success working kind of with data for good, right? Our in our students are very interested in kind of how can we use data science and business analytics to make decisions that benefit society. So companies are interested in that is something we’re definitely interested in partnering with, but you know, we’re also interested in helping companies make decisions in general, around unstructured data.
One of the nice outcomes of this development is that we have this ability now going forward to collect similar data sets for our students to work with. So if a company came to us and said, hey, we’d really like to, you know, better understand our competitors positioning in this space, but we don’t really have a data set that really describes that we could go out and turn this tool that flex on unstructured data and analyze it on the competitive landscape, it kind of collect that data and have our students work for a semester helping the, the focal company answer that problem.
Right, and really that, you know, any data that’s out there on the web, any unstructured data, any instructor, data that a company can make available to us and try and turn into a decision that’s kind of where we’re looking to expand our collaboration.
Jenny: Feel like endless opportunities. You could go
Bill: Yeah, 100%.
Jenny: really. It’s really endless. Well, Haroon, I’d be curious to know what you think. You know, we talk a lot about this in Poole college, especially with employers about making sure that we’re preparing our students with the skills to be successful, but in a space like analytics, specifically in AI, how do we prepare students for jobs that don’t even exist yet?
What can we be doing in higher education to make sure that we’re preparing them to leave with the skills to be competitive for those jobs?
Haroon: Yeah. And this is just my opinion with what I’ve seen. So as I say, digital innovation requires digital leaders, right? A lot of rapidly changing technology, open collaboration, name it, cloud artificial intelligence, internet of things, et cetera. So it demands, digital leadership demands, certain side of new qualities that are both human traits and technical skills, right to meet the challenge of digital age.
You know, nowadays, as we touched upon no code AI platforms are replacing programming languages with, you know, simple drag and drop interfaces. It’s, you know, when Microsoft, you know, previously DOS problems, which then immediately after that got replaced with windows, then you know, it, it became so easy to use for everybody.
So these no code platforms means that powerful AI technologies, which only large well-resourced businesses have been able to afford is now suddenly available to even smaller and medium sized companies. Right? So the point I’m making is machines can do algorithms, but algorithms need managers too. Right?
So algorithms need managers who can consider long-term implications of the data to make sure that the data is right, the right inputs are selected. So algorithms don’t make decisions that can be biased or that can lead decision makers astray. So it’s the requirement is for the managers who can connect the dots, understand the risks and limitations.
I’m looking for leaders who can leverage digital technologies in a way that they are used ethically that it respects privacy, promote sustainability, and ultimately, you know, the AI that benefits society. So in my opinion it’s a blend of human skills, human leadership skills that is complemented with technical skills.
That’s what is needed for tomorrow’s digital leadership skills. And I think, you know, as, as education providers, we need to focus on that blend of technical and human skills going forward.
Jenny: Yeah, I don’t think I’m, at least what we’re hearing in Poole college is there isn’t an industry that does not have an analytic component to it, and you know, some of the soft skills I might add to this is that curiosity is a big one, and the ability to be agile, because the pace at which analytics is moving we could be teaching something now that could be vastly different, right in a year or two, and being able to have students to have the flexibility, to, to, to bend with it.
Bill: Yeah, I very much agree Jenny, in many ways, I think that, you know, I would almost phrase it as the kind of 50 cent word is autodidactic right, someone who is willing to teach themselves about these topics and these concepts. I think that’s very important, but I, I think one thing is to ruin mentioned, that’s terribly important as a skill set is, is this ability to blend data driven decision-making with managerial acumen. And that’s really what we strive for here in the business analytics initiative is how do you both balance what the, what the data is telling what the models are telling us? With the need to make a decision in a landscape that involves a lot of human actors.
Right. And, and that’s, pretty difficult to do. I mean, in many ways, that’s one of the reasons why the social science as a whole is the harder science, right. It’s, you know, the, the, the humans don’t act directly like what the models tell you they’re going to do. And so trying to balance Those things, that’s, that’s the real goal of an excellent manager, I think in many respects and something we’re trying to teach the students here at the BAI.
Jenny: Are all good points. So I want to wrap up with two final questions for the both of you and you both can think about it and answer it, whoever wants to tag in and go first. But I’m curious if we had a crystal ball, looking five years down the road, where will AI be most prominent in your opinion?
Haroon: Okay. So basically in my opinion, companies are re-imagining business processes using AI to increase their operational flexibility, better decision-making and increasing personalization of products and services. Right now, AI is offering you many more revenue opportunities with your existing hardware products by adding a layer of data, adding a layer of intelligence based on AI that will interact directly with consumers.
Right? So it’s basically in the next five years. I think AI will become ubiquitous. More and more companies will use it will become table stakes, stakes, table stakes for processes like process automation. There’ll be more computer vision technologies in retail. it will become prominent in manufacturing, security, automotives, even education, you know, agriculture, customer support and human resources, hiring talent, retaining talent, making sure that, you know, there is inclusiveness in the workforce, et cetera, et cetera.
So I think, you know, there’s lot more that’ll happen in five years, but it’s mostly in improving business processes, right? Reducing the human tasks, the low-level you know, the repidity of human tasks and augmenting human workers, but it’s not going to be anything like, you know what some people may think artificial intelligence will, you know, take over humans and et cetera, et cetera.
I don’t think it’s going to get that. But in the next five years, we will see more of this technology used in various industry. There’ll be, you know, autonomous vehicles and security industry, et cetera.
Bill: Yeah, I very much agree with Haroon, and maybe I’ll take a slightly different spin on it and think about how will it affect the consumer, right? How will the individual in society, you know, where will they most be disrupted or changed as a result of AI? Right. And I think it is very, this notion that we talked about earlier, which is the replacement of repetitive tasks with AI.
Right? So I think in five years. we’ll see partially autonomous of long-haul trucking, right. Starting to become a major player. That doesn’t mean the elimination of truck drivers which some people speculated on. AI is still not good enough to navigate a semi around the downtown area.
Right. You’re still going to need humans in that space. but these kinds of aspects where, you know, we see some of these changes occurring, all right will happen. So another example might be, you know, searching through large numbers of realtor listings, right to identify the perfect house for you.
Right. Is something I’m going to see, you know, and companies like Zillow and Redfin are always already trying to do this. We’re going to see more and more of that become automated. And so essentially from the consumer perspective, I think you’re going to see less than less, Interaction in terms of amount of time with a human agent in many respects and more interaction with these virtual agents. In some cases you might not even know during interacting with virtual agent, right?
Jenny: I was going to jump in and say that Bill too, cause I think most people have no idea sometimes that they’re using or they’re, they’re being exposed to AI. And I think that’s kind of funny because you hear people say, oh, it’s, you know, it’s scary to me. I don’t know what AI is, but you’re using it every day and you don’t even know, which is kind of mind blowing.
Bill: Yeah. I mean already, we’re seeing like, I mean, I think probably the easiest place where AI first implemented itself was in the personalization of content recommendations. Right? If you use Netflix or Hulu or Kindle or Amazon, or at Best Buy or any of that, right, you’re getting personalized recommendations, pushed to you by an AI algorithm.
Right. And that is AI, it’s very poor. And that was probably the first place. It really had a prominent impact on the average consumer. The second, you know, recently chatbots, right? The vast majority of chat bot interactions you had, at least for the first two or three exchanges are AI. And then they hand them over to a human, interactive agent, right who, who takes over. And that allows one human to handle a lot more customer service interactions than they ever were before. you know, in the. Case of the personalization allows you to get better content than you were able to find before. And So AI is improving our lives to a large extent. But we’re going to continue to see those small changes, some of which may not even be noticeable to the average consumer as we go through.
Jenny: So final question for the both of you, and I think you both have very different lenses. So I’ll start with you Haroon. What excites you the most about AI? Definitely where we’re in now in our space and where we could potentially be down the road.
Haroon: Yeah, this question, you know, I want to relate it back to Bell and Howell where I work because we are in the business of providing solutions to grocery retail business through our automated grocery pickup solutions, robotic pick-up solutions and services, right? So I’m fascinated with the use of AI in the retail space, whether it is, you know, smart shopping carts like Amazon go shopping cards where you basically don’t have to do anything.
But while you shop other than picking up the groceries or items that you want, cashier less checkout technology, product recommendation systems, even sentiment analysis at the checkout counter, you know, these AIs can now determine based on the length of the, you know, the queue based on how many people are waiting in line and their sentiments emotions can be measured and the associates can be sent to help them.
You know, that similar technology can also be used to prevent theft, et cetera. So, you know, and, and personalization in the, you know, the advertising and even metaverse, for example, the whole thing about metaverse, how it can add an angle of interactivity while you shop so that, you know, well, we’ll pick up Milky, you know, everything about that, how it is sourced, where it is from, you know, all the Nutri nutrition, content, et cetera, through some of these you know, AI and augmented technology. So that’s where that excites me. The field is actually, you know, rapidly this grocery space, the retail grocery space, or even retail space in general is reshaped by some of these new technologies. So, we are living in exciting times and that’s what excites me about AI space at this time.
Bill: Yeah, I’ll, I’ll pick up on one of the last notes Haroon made there about the metaverse. Honestly, that’s what most excites me about AI. I am a science fiction reader from way back in the day I was reading William Gibson and his cyberpunk novels. And then Neil Stefansson who actually invented the word metaverse in his book, Snow Crash.
The idea of having this like virtual world that parallels our own world and that we can interact with. It’s only possible with the use of AI really, to a large extent. And it has So many interesting implications for the business world. Marketing, you know, like, you know, there was, there was, there was a bunch of advertising over the Superbowl around the metaverse right.
And different possibilities or instance of having like virtual beers in the metaverse, stuff like that. But also, you know, in more serious applications. Um in fact, I have an assignment in a census just recently about learning and development in the metaverse. And how can companies take advantage of a virtual environment to train their employees?
Right. If you think about it, if we can create this metaverse type of world, that parallels in many ways, our, our actual reality, it brings a new level of sophistication, the training for service enterprises, right. And how they’re dealing with customers. Some of whom may be very pleasant to deal with someone who may be less pleasant to deal with and trying to teach our, our staff and our key players, our key customers facing personnel how to interact with customers on a regular basis. Um in a, in a, in a more appropriate way. And I think, I think the metaverse really does bring this ability of us to really expand what we’re doing in the real world to a whole new level that I think will be really interesting to track. and now I’m excited to see where it goes, honestly.
Jenny: So to summarize, I’m hearing, embrace it. Don’t fear it. Right? This is a good thing that a lot of great things in the works. Thank you both for being here today. This was a great conversation. I definitely want people to check out the BAI round table, May 19th. This is what we talked about, Bill, and Haroon, I think if students aren’t already excited about it, they probably are going to need to think about signing up for your class.
Right. I think what’s a given after this one. But thank you for your willingness to join the journey with BAI as an advisory board member and bringing your expertise and knowledge to the classroom. I know we, we benefit from that big time at pool, so, but thank you both. And, look forward to seeing all the things that at AI can do for us in the future.
About the Poole Podcast
The Poole Podcast is a think and do conversation about the relationship between academics and industry. Each episode will share research and ideas from inside the classroom from our incredible NC State faculty and explore how it’s being translated into practice. Released every two weeks on Monday.