Podcast 003 Crash Course in AI and ML Part 2

Episode 003 Crash Course on AI and ML Part 2 Transcript

Heather McKee:              Welcome to the Modern Polymath where we discuss topics in technology, economics, marketing, organizational behavior, market research, human resources, psychology, algorithms, higher education, cybersecurity.

Heather McKee:              Hey, podcast universe. Thanks for tuning in. On today’s episode of the Modern Polymath we’re going to give you part two of our crash course on AI and ML, aka artificial intelligence and machine learning. On the last episode in part one, we covered the definition of AI, how AI has developed over time and went into a few sub fields like machine learning and speech processing.

Will Callaway:                  AI is not artificial. It’s still designed by humans. It’s intelligent design and there are still curated processes that need to be done before the AI can be optimized to perform a certain task.

Heather McKee:              For part two, we’ll discuss how organizations are using AI in their business and give some helpful tips and insights for business leaders who are exploring the idea of integrating AI into their business. Today again with us we have Dr. John [Christiansen 00:01:20], John-David McKee, Will [Callaway 00:01:24] and I’m Heather McKee. Let’s get part two started.

John-David M.:                To start with, we need to understand how it can be used within organizations, the impact that it can have because it can create significant competitive advantages to companies that figure out the right application in the right moment. It also can be an opportunity or a putting you in a position where you fall well behind your competition because your competitors have adopted AI effectively and you haven’t and therefore are missing out. You’re basically stuck in the stone ages and everybody’s moved on to bronze. But before we go into how you can use it as an organization, because this isn’t just important to the technical crowd and that’s what I want to… actually it’s most important that we can do here is remember that it has to start at the strategic level. So I’m talking to you leaders, executives, directors levels. You can have analysts in there all day long that are great at math and great and analyzing this stuff, but at the end of the day, the data doesn’t answer questions for you.

Speaker 3:                        You have to ask it questions and then get the answer out of it. And to do that, you have to start at the strategic question level of what are we trying to answer? What’s important to our business? What’s important to us strategically? And an analyst isn’t going to know that or it’s unlikely they’re going to know the whole picture. They’re not in those meetings. They’re not sitting there, they’re doing, they’re there to do a job. So you as an executive, and this is everybody outside, we think a lot of companies are very sophisticated with their data and most are not at all. So you really have to start at that strategic level. It starts at the top and then works its way down. And a data literate manager is something that everybody needs to be at this point because you have those resources there.

John-David M.:                To elaborate, it’s really important to understand what that data-driven manager needs to look like because this is an important role to creating a data-driven culture. I mean, you have to have the leadership capabilities to be able to go all the way from the board room to the marketing team to who’s going to be reaching out to the customers, to the product team is going to be implementing the data at the product level, the sales team to communicate this in the sales process to the specific analysts working on all of this process, all has to be pulled together so you have to have that. But you also have to have enough chops technically to explain how this is going to work and be able to get that message across. But again, to that point, that’s why it has to be someone who is strategic, who can be that change agent, who can lead that company into the data literate culture.

John-David M.:                What that really means is you have to leverage critical thinking and to leverage critical thinking, you have to be at the right position to be able to do that across the organization because we’ll keep coming back to this throughout. This is one of the core principles that we work from. Critical thinking is imperative to solving any kind of complex problem, but certainly it’s at the forefront of any analytics undertaking. To find any kind of opportunities that are out there, to identify weaknesses in your business or create a competitive advantage or something along those lines, you have to start by formulating the right questions and then understanding the need and what you have in terms of your resources. Then making selections like what’s the right technology? What’s the right platform that we should use here? Things along those lines. But that all starts with critical thinking.

Will Callaway:                  So let’s talk about how do you get to those right questions? What does the team look like? Or how do you formulate a team as a manager who’s been put in charge of this task? Who do you get? Who do you get on your team?

John-David M.:                Well, that’s why you can’t isolate the quote unquote analytics team unto itself. I mean, you have to have management representative who understands the strategic goals and able to keep things going, which includes project management. You have to have the marketing sales team involved because they are going to be inevitably using the data to some to some extent if not completely. You obviously have to have the analytical people that are involved with that. It’s not just one singular analyst. There’s the process of getting the data ready, data cleansing, blending and all that. And then you have to have the really the collaboration between that person with the technical chops who is that deep analyst, that sophisticated analyst, along with the strategic element that comes together into being able to ask the right questions and answer them through the analysis.

Will Callaway:                  And or the guy who’s going to champion it and maybe be the data translator, the person who curates the message for the different sales team, marketing, product team, whoever it may be that needs to hear the message of the insights that maybe the data science team or the analytics team has derived.

John-David M.:                Precisely. I mean, it’s got to be that. You have… that person’s imperative cause they have to take that then to the board and explain it to them. Because there’s ultimately an investment being made here that the money could be put elsewhere if it wasn’t, if this wasn’t the goal, so the executive team has to have a clear vision for what this advanced analytics program’s going to look like. What does a success look like? How do we define that over time? And then how do we translate that across the board to all the different shareholders within the organization that need to play their part as a part of a bigger team?

Will Callaway:                  Right. Yeah. And I think even like you were saying about how do we start measuring this, the person who’s championing it and the analytics, let’s say, team who’s really doing the hardcore research is probably different from the project manager. And the project manager is more in control of the budget. Well, that project manager might need to actually go talk to the analysts and say, “Hey, can you guys help me create a framework of how we’re going to measure if this is a success, if this is a failure, why was this a failure?” What is our iterative process to maybe build out an 80/20 thing where we approach 80% of our problems only building out 20% of an analytical process and those will be our quick wins, right? Those are our six month wins. Then what does the next iteration of the data culture in the company look like to where maybe the first year, the second year in all of that, those are then maybe long-term projects where you really start building out your infrastructure.

John-David M.:                Yeah, that’s all necessary and that’s why you can’t just yank a part-time project manager over here and have them come in and not be a part of the team and not understand the goals and not have them clearly defined. There has… you can’t put those silos up and expect everybody to work together coming from these different departments, these different backgrounds and various amounts of capabilities and knowledge within the organization and expect them to gel and these things work. This is, again, why that change agent, that leader, that champion is so important to this whole process.

Will Callaway:                  And like you were saying, you can’t have any silos. You need to have different collaborative teams. Well, that’s pretty much the modern polymath, right? Be multidimensional with your critical thinking. Be able to go across different sectors, different departments in your company and either be able to speak the same language as the marketing and sales team right before you move to the product team or vice versa, or talk to the C level guys right before you start talking to your underlings. Something like that. That is the true definition of maybe a polymath to be able to be multidimensional, solving different problems at different stages of the problem process.

John-David M.:                Absolutely. And you don’t have to be the expert in every one of those fields. In fact, if you are, you haven’t hired people very well. You should hire people that know more than you in each one of those areas. But you, as that leader, you as the embracing the polymath ideal need to be able to speak to each one of them and most importantly be able to see how all those pieces fit together into the big picture. That’s the difference between a successful initiative and something that’s going to tank from the start because no one understands the big picture.

John-David M.:                Now that said, to leverage AI, you as an organization first have to understand your data. You have to know what you have and how to use it. And to use data effectively is what we call data literacy.

John-David M.:                So if you’re asking yourself, are we a data literate company? You’re probably not. And that’s okay, that’s very common, but you have to start there. Make it simple at first, take all the data and distill it down into what are the strategic objectives we’re trying to accomplish? And it really needs to come from the top. It needs to come from leadership thinking strategically. And then once you get literate at a very elementary level, then you can start to layer on top more sophistication. Otherwise you’re throwing resources at a problem that you haven’t actually defined yet.

Heather McKee:              Well, so what’s a good example of a data literate company?

Speaker 3:                        So one of the best examples, and honestly one of the earliest examples that I can think of, of a forward thinking company using data to streamline their efficiencies to become data literate, well ahead of their competition was Walmart. You think of Sam Walton when he comes along as this folksy guy from Bentonville, Arkansas, but he was way ahead of the game. He used computers, those little things that came along in the 80s and gave you the opportunity to actually analyze data. And through that they were able to get their inventory control processes down to such a point where they were far more sophisticated in terms of that with competitors. And through that they were able to ensure that they had the inventory in stock and that they were able to have price cuts below the competition that they could then pass on to the consumer.

John-David M.:                So being a data literate organization is really taking yourself from a knowing culture to a learning culture. Knowing culture, meaning you make decisions based on heuristics and best guess anecdotes, the things that have worked before. To a learning culture, which is a culture that uses data to make objective data-driven decisions. Once you’ve gotten to that point where you can make decisions based off data and be objective in your decision making, then you really at a point where you can leverage the power of AI and technology to make your organization much smarter, much more effective, and just continue to make your organization better relative to its current position and to your competitors.

Heather McKee:              So with that said, what does it take?

John-David M.:                You got to have the data. Data has to be there. You have to be gathering data from somewhere. Big data is thrown around all the time, obnoxiously so. And we’ll talk about what that is at a later episode. But if you have good data and you can teach the machine to learn, it can analyze it billions of times faster than we can as people.

Dr. Jon C.:                         Yeah. Which is the goal.

John-David M.:                And you’re going to have the observations that you’re going to have are so much broader than any human mind can compute. Right?

Will Callaway:                  And to get that good data, you usually have data engineering teams who create the data infrastructure, the pipelines, wrangle the data and basically scour the internet trying to find data on whatever their target is to put in the hands of the data scientists, which is your next step in your data analytics team.

John-David M.:                But the human has to still be behind it. It has to be driving it. Human intelligence is necessary to correctly guide the AI however it’s being applied.

Dr. Jon C.:                         Because AI, all AI is trying to do is mimic what you’re telling it to do. It’s going to… you give it the set of rules and it’s going to play by them. That’s the criteria and that’s something that we, you will hear us talk about a lot, which is kind of lesson zero of the world. When in doubt, first learn the rules. If… I can’t walk out to a basketball court and expect to win if I don’t know how to play.

John-David M.:                So everyday we’re dealing with this stuff. We are, we’re seeing it happen. It’s impacting our lives whether we know what it is or not. And it’s one of the reasons we wanted to talk about the broad concept of AI and hopefully you get some value out of this because it’s all around you. That’s nothing to be intimidated by. I mean, it’s not, it’s a value add. Yeah, there’s some scary movies made on this and sure The Terminator is a scary idea that in theory could potentially happen one day but probably not because people are controlling that. And the upside of it is significant. And you don’t have to be a computer scientist, you don’t have to be a mathematician to understand what these things mean. You don’t need that. You don’t have to know any lines of code. What you have to understand is they are programmed by people. People are driving this and we’re using it to solve problems.

John-David M.:                So you’re using it everyday whether you know it or not. Get over the intimidation side of this and embrace it as the future, it’s where we’re going and it’s going to… it may create some problems. Anything that has any good is going to create a bad, the yin and yang situation. But it’s going to be a continually present part of our lives. And everyday that we grow, every day that goes on, we’re going to know more and more about them. We’re going to see it more and more.

John-David M.:                In our real world example, it’s the whole idea of stop guessing. The data’s there. We can use AI to remove the guesswork out of it. Human intuition is great, but it’s flawed. It’s been studied and studied and studied. We’re not great at that and at seeing things objectively and seeing enough things to actually be able to make a decision intelligently.

John-David M.:                But a computer can do that if it’s trained to do it. It can look at millions of observations that help you make a strategic decision. That’s one of the things that we help our clients do is we look at their data and we try to understand what they’re trying to accomplish with, be it their business goal or the biggest challenge or whatever it may be, and then let the computer do what it does well by guiding it in the right direction. And then you can answer some really, really sophisticated questions with a very small margin of error.

Dr. Jon C.:                         Which is what’s amazing about what it’s doing for our world and how it’s simplifying it and how we can continue to leverage it to make us all smarter and better at our jobs and spend more time effectively and efficiently doing what we were hired to do. And when we get into how resume screening is done through AI and machine learning, we’ll really highlight how it works and how it’s really good for organizations and companies to be using this type of technology.

John-David M.:                That’ll be an interesting one.

John-David M.:                All right. I think we pretty well beat that topic to death. Not really. We could go forever on that and there’s so much to learn but what I really want to do is just, again, demystify. It’s a crash course, right? It’s what you need to know to get comfortable with it. Hopefully somebody listens to this and decides they want to go into this for a living or something, but the goal is to make sure that it’s a difficult concept to understand made less difficult.

Heather McKee:              Yes, hopefully we’ve done that for you all, but if not, we actually have some in-depth content that will dig a little bit deeper into these topics on our website, insandouts.org. You can go there and visit our blog and see our recent posts for all things that we discuss on the podcast. So that’s it for us today. We thank you for tuning in and hope that you will join us for our next episode, which will be a crash course on algorithms.