How are artificial intelligence systems making a difference in people’s lives? In this episode, Dr. Wanda Curlee talks to APU logistics professor Dr. Oliver Hedgepeth about the advancement of robotics and the use of AI in everything from household goods to self-driving vehicles. Also learn how various companies like Amazon and Walmart are incorporating AI into every aspect of their business including customer service, logistics, and delivery systems. Hear about the growing pains of this technology, including some of the funny and not-so-funny failures, as well as the legal challenges as the government tries to keep up with quickly advancing technology.
Listen to the Episode:
Read the Transcript:
Dr. Wanda Curlee: Welcome to the podcast. I am your host, Wanda Curlee. Today, my guest is Dr. Oliver Hedgepeth. He is a professor at American Public University, and he teaches in the area of reverse logistics, the AI courses, transportation and logistics, and many, many others. Oliver, welcome back. It’s great to have you here.
Dr. Oliver Hedgepeth: Wanda, thank you very much for inviting me here and looking forward to it.
Dr. Wanda Curlee: Our topic today is artificial intelligence, AI. Robots, and various applications of AI. Why is this topic important for our listeners to understand?
Dr. Oliver Hedgepeth: Well, I think it’s really important because AI is in everything we do today. You are involved in AI when you turn on your computer. You’re involved in AI when you turn on your television set and want to search something. You buy a new vehicle today, you’re involved in AI because not just that it drives by itself, but even the ones that don’t drive by itself, if you’re driving down, and you’ve seen the commercials where you’re looking off somewhere, eating popcorn or French fries, and all of a sudden you’re going to run off the road or run into an 18-wheeler, or something runs in front of you, the car is going to say, “We’re stopping.” And that’s AI, it’s all over the place. So I think it’s very important to discuss some of these subjects today.
Dr. Wanda Curlee: What is happening with AI and robots in 2021 business?
Dr. Oliver Hedgepeth: Well, wow. I look at it from a significant point of view. Academic researchers, we talk about things that are significant, but really it’s kind of like you’re asking a question, “So what? So what about AI?” And you really kind of answer the questions about it by saying, “So what? Or is it solving a problem for me? Is it some significance in my life?” And the answer is, well, it helps you is what we need to talk about.
There is a mystery of AI, not everybody’s a computer technologist and we use the term AI a lot for robots and new machine intelligence. There is kind of a mystery for a lot of people. It’s like computer programming, computer programming is a mystery to a lot of folks unless you are doing it. And so I think AI does matter today, but we do need to address the “So what” questions. And that’s what we’re going to talk about today, I think. That sound okay?
Dr. Wanda Curlee: Sure. So let’s get back to robots. How do you see AI and robots interfacing with people? Do you think that’s different than the robots you have in maybe logistics in a manufacturing warehouse?
Dr. Oliver Hedgepeth: Yeah. Robots are interfacing with people. Over in Japan, robots are teaching kindergarten to about third or fourth graders. I’ve seen the videos of robots teaching. And the students know it’s a robot because it’s a little-bitty robot, about two feet high and it’s helping them understand a principle, ask them a question, answer a question, tell them to do something. There’s a teacher in the back of the room and they’re not messing with it. The students love to play with it. So robots are there.
You see these commercials on TV of pizza being delivered by a robotic vehicle. Well, that’s real. It may be just a little gimmick right now, but the gimmick is leading a way. Up in Radford University here in Virginia, there’s a company up there, who’s delivering packages, delivering pharmaceuticals and medicines from a drug store to your house, with a drone. It’s a robotic-controlled drone, it’s got a little AI system. It knows where it’s going. It knows what not to bump into, as it comes to your house. So it’s happening more and more. And I think we’re going to get more involved in it.
Now, it doesn’t mean robots are going to replace all our people. Because Walmart got rid of its 500 robots that were pulling boxes off the shelf, I think about a year or two ago. They got those, like it would be kind of cool. Six foot high robots would take a box off, put a box on, put some macaroni and cheese up and take it down whatever it needed to.
But they found out, it is a little better when we use humans and the human seem to like it. And we seem to like it and doggone it wound up being cheaper than the robots. So we’re going to see them a lot. I mean, we really are.
Dr. Wanda Curlee: Wow. That’s kind of interesting that Walmart took away the robots and replaced them with humans. So speaking of that, how are human AI applications evolving?
Dr. Oliver Hedgepeth: Well, human AI applications, you’ve got them inside the surgery. There are AI applications and robotic applications, helping the surgeon cut you open and do things inside of you. I am very impressed with how it’s improved over the last 10 years.
There are robotic arms that are doing some of the factors for surgery inside of you, that the doctor doesn’t want to do because there is total 100% control of what the robot’s doing there. Where a human, the human doctor is not going to make a mistake, but they are human and something could happen, mistakes have happened.
So they try to get rid of these standard procedure things, standard cuts, standard putting something inside you, standard sewing up. It’s happening out there in the medical field. I am really impressed how they are involved in that. And there’s a lot of investment in the medical field.
Dr. Wanda Curlee: Wow. That’s very interesting. What are the expectations for artificial intelligence and the robot and other machine learning applications for the next five years? What are we going to see in the next five years, if you had a crystal ball?
Dr. Oliver Hedgepeth: Well, the crystal ball, it shows me that, like I said before, we will see a lot of things in the automotive industry and in manufacturing. The pandemic of 2020, 2021 has shut down many warehouses and manufacturing firms. And the manufacturing just stopped, they just kind of shut down. And a lot of investment was put into replacing some humans in the manufacturing with robots. And you’re seeing more and more of that.
They’re routine things and one thing I’ve noticed over the last, oh gosh, 10 years or so is robots in warehouse. I used to go to a warehouse and I would see 20 or 30 men and then later women and men running down the aisle, driving a truck there to pick up something and put it down. Well, they’re gone. They have lost their jobs doing that. They’ve got a robot that’s following a yellow line. It goes to where it needs to pick up something, picks it up and pulls it back to where it needs to be put on the truck. The humans are still there, but not as many.
Instead of hiring 20 or 30 humans, they may hire five or six. And those five or six, they’ve got a little better skill too, because they’ve got to learn to deal with these robots, these computers. And then like Amazon, good gracious Amazon. It’s got like 50 warehouses and they’re doing another one I think up in Northern Virginia. And they’re really promoting how much robotic stuff is there. They’re still going to hire 2,000 people, but they could have hired 6,000 to 10,000 people, if it was 10 years ago. So you see there’s a difference there. Go online and look up Amazon warehouse and just sit back. It’s like, “Wow, okay.” It’s amazing to see what’s happening with the robots.
Dr. Wanda Curlee: So how about some other industries like education, media and customer service. How do you see that changing in the next five years with AI and robots?
Dr. Oliver Hedgepeth: Well, it’s happening and it gets kind of funny also, but in academia, let me tell you. I don’t think we’ll be replaced. I’m a professor here, you’re a professor. I don’t think will be replaced by robots, but I do see that robots will be helping us out a lot.
I also noticed that there are many universities like American Public University that teaches AI and how it can be used. And universities are doing research on little robots, how they work to do something for different levels of people. But they’re also teaching courses on it. I mean, everybody from MIT to Carnegie Mellon to Stanford and Harvard and Yale, all sorts of universities. I think I’ve got a list of about 30 universities, when I was looking at it, that are doing AI and robotic applications. And they’re not just teaching, here’s what AI is. They’re doing demonstrations. They’re showing how they teach somebody so they can leave the university and go to work.
A lot of these aren’t bachelor’s degrees or master’s degrees, they’re certificates in a two-year program. They’re training you to go do a job in front of a robot or help manage a robot. So the education is going like crazy. I mean, it really is so exciting. In fact, the two-year school down from where I live, they have now started an AI course, and I was just shocked how they were trying to do it.
Dr. Wanda Curlee: You covered education. How about the media and customer service?
Dr. Oliver Hedgepeth: Well, the media, oh my goodness. The media is fun. Let’s say, newspapers. My wife’s a newspaper editor, retired. And artificial intelligence is being used in the media to figure out the weather. Weather is being done by AI. They don’t do it by hand anymore. It’s all done by AI. And it gives you the reports, quarterly reports what’s happening.
It doesn’t mean the AI system’s going to do it. I think it was in 2017, the Los Angeles Times, big newspaper. They had a story about a 6.8 earthquake. It just happened. Wow. In Santa Barbara, California. And I was like, “Ooh, earthquake.” The problem is the AI system they involved had the chance to look at what was going on and it found that it had an earthquake in 1925. But for some reason, the computer program generated an article that was published in the newspaper in 2017, that we just had a big earthquake. Wow. It forgot to put 1925 in 2017. Yeah, they’ve corrected it. But there are some funny things that happened there in the media.
There’s another one for, I love this one. I remember reading years ago about someone developed an AI system and fed 30,000 cookbooks into an AI system, cookbooks. All kinds of food. And it said, “Okay, create a recipe.” And I remember it created something that was, I don’t know like “Beothurtreed Tuna Pie” or something. It’s some strange name.
And I remember the ingredients. I think it said, all you need is three ingredients. “One hard cooked apple mayonnaise.” I have no idea what that is. And then it said, “Add five cups of lumps, thinly sliced.” It doesn’t say what lumps. And then it said, I got a note here. It says, “Surround it with one and a half dozen heavy water by high, whatever that is and drain it and cut it to one quarter inch remaining in the skillet.” It made no sense at all. But they thought it was kind of cool. They were going to use AI to read all this stuff and do that.
Now, AI can do some good things. AI beats chess champions. I’ve seen AI trained from ground zero on how to play chess and it loses, but it learns the little areas that creates their solution. And AI will beat chess champions. Now, in fact, I’m a chess player. I gave up playing chess online about 10 years ago because I would go play chess by myself with a computer and it could pick the beginning stage of the AI system or the expert stage. I couldn’t beat the beginning stage. I thought I was smart. I mean, I play my grandson all the time and beat him and he beats me.
Dr. Wanda Curlee: It’s nice to have it like that, yes. So Google is doing a lot with customer service and AI. Let’s face it, we have chat bots and when you call somebody it’s a Google AI system. How do you see that evolving in about five years?
Dr. Oliver Hedgepeth: Well, I think Google is really doing a lot of customer service. They can place human-like calls to a customer and try to understand what’s going on. You get a phone call and it sounds like you’re talking to a human. It asks you a question. You answer a question. Yes or no. And it gives you another piece. And it says, “Okay, here’s how we can solve this. Bring in your item you didn’t tomorrow and we’ll refund it.”
And you think you’d been talking to a person and you have been talking to an AI system all along. So customer service is getting better. What it’s doing is replacing the human to be hired, to be trained, to listen to the irate customer because a customer can come there screaming and hollering about something they bought, a pie and it’s messed up. Somebody put it in the bag upside down and sideways, and it looks like an apple cobbler now. And I want my pie and I’m angry that the clerk didn’t do it.
You can scream and holler at the customer service person. If it’s human, they’ll listen to quietly, but you could get them upset. But the AI system would just calmly listen to you and get the facts and say, “Well, bring on in the pie and we’ll get you another one or keep it and we’ll get you another pie free. No problem. Come on in.” It’s really great how customer service is being enhanced. I’m not saying, it’s replacing humans, but it is being enhanced by such AI systems. It is really exciting to see that.
Dr. Wanda Curlee: Let’s go to, what are the legal issues with the use of AI and robots and related machine learning applications? I’m sure there’s quite a few.
Dr. Oliver Hedgepeth: Well, it is most interesting because AI and robots are replacing humans. And humans make decisions and there can be legal issues. And I looked up some laws on the books about laws using AI and robots. And there are Congressmen who are, Congresswoman, who are working on developing laws to control decisions being made by robots, by AI systems. And the thing is, good luck finding those laws.
The question is, it relates to laws more than ethics and laws are, what if an AI system, what if a robot system makes a decision? What if that surgeon has that robot do something as a deciding factor to do A or B, to cut here or cut there or to sew here or sew there? What if it does something and because of that, the person dies? People do die on the operating table and that’s just too bad.
Or it could be another decision where you make a decision for finance. We have AI systems in the financial sector, advising you to invest your hard-earned money. What if you take the advice of a nice piece of software, you’re not talking to human and you invest it. And all of a sudden your $100,000 goes to zero because the investment was like, “Oh, bad data.” Maybe the bad data came from a human that AI system ran, doesn’t matter.
The legal aspect of who you sue is not there. The thing is, the law is evolving right now. And it is very difficult because you’ve got to decide, is this piece of software and hardware legally liable. And if so, oh, how about the computer programmer, who programmed those decision rules? Because all AI is, it’s like computer programming from the ’60s. It’s a bunch of rules. If A happens, then B happens. And then, if B happens, then C happens and if, C happens then buy cornbread instead of muffins, whatever it is.
And that decision that is being offered to you, you as a human have to make the decision to take it. But if someone’s offering it to you, if it’s human to human, you can sue them. I’m sorry. You can get angry at them and maybe get something back. But that legal aspect is still up in the air. And very good question and something we should all keep asking our Congressmen and Senators about and seeing what’s going on, do a little search.
What are the laws against AI or laws for AI? What are the laws for robots? So anybody listening out there, go search up the laws and then you find out what your Congressmen might be, or Congresswoman might be thinking about in AI. Give them a call or send them a text message and say, “I’d like to know about that law.” We really do something out there, but that’s a long ways coming.
Dr. Wanda Curlee: Yeah. It takes a while for the law to catch up with technology. We’ve seen that with other technologies that are rolled out. But we were talking about universities a little while ago, Oliver. How is the application of AI and robots being used in colleges and university courses?
Dr. Oliver Hedgepeth: AI is being used by professors. They help understand maybe complex problems. They can bring them into a classroom and do that. There are some examples of robotic systems at a lower level producing a course, giving some instructions on how to do something, they’re basic instructions.
The actual replacement of a human with a robot, it’s overseas more than here, but people are looking at it. MIT, University of Maryland, Colorado State. Goodness, California Institute of Technology. They’re looking at it as replacement. So they are thinking about it. I would expect in the next, I wouldn’t say five years, maybe 5 to 10 years, we might have some neat courses, where it’s being taught by a robot. I wouldn’t be surprised. And it’d be kind of fun to see.
I would expect it might be a low-level course, like a 100, 200 level course. Like here are the basics for how to write an article, an English course. Here’s what an adverb is. Here’s what a verb is. Here’s what a noun is. Some basic structural rules that are very easy to put into a robotic system and have someone work on that, It would be something simple, I would think at first. Versus complex, like how to do surgery for a person.
Dr. Wanda Curlee: So Oliver, you and I are both at American Public University. And there are a couple of AI courses in the business area and you teach them. What do you teach in those courses? And how are the students taking it?
Dr. Oliver Hedgepeth: I find it very exciting to teach the course because the students teaching it are very interested. Some of the students don’t really believe in AI. They’re kind of interested in it. And once we get involved in it, they say, “Well, AI doesn’t work here or there.” And I obviously have a lot of examples of where it does work and they’re surprised.
They’re taking the course to find out how their role as a human fits into manufacturing or the military or customer service, as we mentioned before. They’ve got a job. All these students, they’re working. Our students are like between 20 and 70 years of age, they got a job, they’re working or they’re retired and going to another job. They want to understand how it’s impacting their work. And maybe because they’ve read something that says, “You might lose your job.”
So let’s talk about that. And how can I increase my ability to work with AI and robotic systems? It is most exciting when we get into what is AI and what are the applications? We get into ethics. What are the ethics of AI? They don’t think about that at first. And then they realize it might be something ethical, a decision that AI gives me.
The legal aspect we discussed also, that comes up. We always bring that up because we want them to be involved and understanding the laws will change. There will be changes out there for laws, just like the laws for use of drones. Drones that are robotic systems have changed so that certain companies can now move goods with drones. We didn’t know that until a couple of years ago, now it’s happening now.
So they get excited about it, to see how it impacts their life. And that’s probably why all these other universities that I mentioned are doing AI as well. It’s not that these professors or the university wants to teach it, it’s that the students are asking for it and I see this all the time. I’ve got flyers that come to me all the time from MIT, for example, about a two-week course or four-week course on AI and how to use it in your work.
I see advertisement for two year college degrees, on here’s how AI can help you do your job in manufacturing or your job in customer service. The students are asking for it and the universities are responding.
AI is really a common, new course. It’s part of the school of business, as well as I find one department of economics, an economics department professor, one at Penn State, I think was talking about AI and robotics and how it’s used because in economics people need to understand how they’re going to spend money with these systems.
Dr. Wanda Curlee: And you mentioned some failures of AI and they were kind of comical. Do you have some other examples of failures in the use of AI and robots?
Dr. Oliver Hedgepeth: Let’s see. Yeah, there was one, I forget when it was, not too long ago. Where a couple, man and wife, bought a robot cleaner, a vacuum cleaner thing. It’s going to clean the house and they go to sleep at night and they turned the vacuum cleaner on at 1:30. So at 1:30, the little vacuum cleaner, it turns on, it’s going to sweep up the floor.
Well, the AI system didn’t understand what dog poop was. You know what dog poop is. Well, at 1:30 in the morning, the little robot turned on and it’s going around and cleaning the floor. And for some reason, the AI software didn’t understand dog poop. And what it did, it picked up the poop and it started spreading it throughout the floor. It even decorated the floorboards, it decorated the furniture. It put poop on the legs of the furniture. It put poop all over the rugs. It tried to spread it everywhere. That was one of the silly mistakes that happened. So we have those that happen like that. It’s kind of silly.
And then the presses that I mentioned before were at AI system back in California said, “Oh, we have an earthquake.” But it was like a 100 years ago almost. So there are those things that do happen.
And online, I’ve seen news of robotic cars, a few years ago that killed people. The robotic cars didn’t really understand either turn left, turn right, slam on brakes, whatever it was. The driver was asleep, obviously, they tell them go to sleep and the car takes you to your place and that’s fine. But then you see where it did have a wreck. It didn’t turn correctly and it has a wreck and it burns up.
So there are some mistakes that do happen that are not funny, but those serious mistakes are getting fewer and far between. I really believe it.
There was also, I think in China, there was a robot in China that was supposed to be delivering things around shopping malls, from one store to another. And it crashed and fell down. It fell down the escalator, it didn’t know how to go down the escalator properly. And it just rolled down, rolled down. It hit people, it dropped everything. So the robot wasn’t really trained to take care of that little escalator.
Oh, there was one software that would look at your face and would scan it and look at characteristics and say, “This person’s going to be a criminal, we’re going to put this person in their criminal category.” And the police really looked at it that way. Somebody who just happened to have the characteristics. The people they programmed it for, weren’t really like other people that it had never seen before, because we’re different races, different colors, different features, hair, everything. It does that.
And let’s see, the soccer ball, there was a camera in a soccer match program to follow the ball, it followed the round ball and it followed the ball. But for 90 minutes in one of these games, it followed the baldheaded man in the stands. As he got up and walk, it followed baldheaded men. It found one man, then it found another man. It was chasing baldheaded men and the ball in the field, it just forgot about that.
Dr. Wanda Curlee: Some of those are hilarious examples. How about AI and older people? I think there’s a failure there. What are your thoughts about that?
Dr. Oliver Hedgepeth: Well, there could be, it could be. If AI systems are supposed to recognize older people, there have been examples where it didn’t recognize the older person, because a lot of older people do have a lot of wrinkles on their faces. The faces do change. You get a certain age, I’ll say 70, 80, 90, a 100. And your face it just might change from nice little smooth face, it might not look the same. And it can’t recognize you properly.
The other aspect is, there’s a good thing too. There’s some bad things there, if you’re supposed to be facial recognitions. I have found that there’s some good factors, too. And I’ll relate this to nursing homes, where I know there are some women that they can only talk about a few topics all day and they talk about the same topic all day long. “Oh look, the dessert was great. I liked this. Did you like the dessert?” And then the next thing they say is, “The dessert was great. Did you like this dessert?”
And then they’ll do this all day long. Or they say, “Oh, I got brown shoes on, do you like brown shoes?” And they’ll just say it over and over. A little a robot is sitting next to them and the little robots like, “Wow, that’s really cool. Tell me more.” And go on and on, whereas if it was you talking to your mother, you might get upset. You might leave them alone. You wouldn’t help make them feel happy.
What the little robot is doing is making that woman have a nice life. I mean, it does happen. And I will say this from personal experience because my mother died, such a woman in a nursing home. And she said the same thing over and over all day. And they did have a little teeny miniature robot at one time doing that. And that was the coolest thing.
My momma liked the robot and I found that other people, and I’ve seen videos of women and men in nursing homes, walking around holding the robot by hand, as it walks with them. And they think of it as real, just like a five-year-old or seven-year-old thinks of it as real in Japan. They think it’s a real thing. So the older generation, there’s some good and bad there.
Dr. Wanda Curlee: My mother is very scared of technology and she won’t pay bills online and things of that sort because she’s scared of it and maybe rightfully so. She goes into the bank, she won’t do the drive-through. She has to have a teller. She won’t do online banking.
So I think AI has got some aspects that it could help the older generation, to help them understand that it’s not so scary, but that’s training. I think it’s incumbent upon us to train them. So how do researchers major the significance of AI technology? Research helps us to understand if we’re getting to where we want to be.
Dr. Oliver Hedgepeth: All these universities I mentioned, and the others who are doing AI courses, AI research. You do research to find out what people think about it. For example, American Public University did research of, I forget, about 150 military people who were students at the university and asked them, “What do they think about AI? Give me examples of it. Did you like it? Did you understand it? What’s good about it, bad about it?”
And they had a nice analysis, just like MIT is doing the same thing for their students and general population and customers and customer service. But how do you measure? At the university level, you don’t just write a report like a newspaper report and say, “Well, here’s what a 100 people thought.”
That’s not the same, at a university, you have to find out what’s called the significance. And the significance is really measurable. In fact, I think since the 1980s at universities, they were like 10 different significant factors you’ve got to measure against. If you’re going to make your research, make sense.
For example, you might have something that was subject-based claim. If your AI has impacted a person, if you did some research and talked to people and that AI system or robot system impacted a group or a person, or just answered a question that nobody realized, it can answer this question? Or maybe it solved a problem.
For example, it could solve a problem in five seconds versus five days because there’s 20,000 pieces of data that have to be analyzed. And all of a sudden the AI system analyzes that data in a matter of seconds and gives you a decision, go here or go there, buy A or buy B.
Well, that’s called subject-based significance. Things like that happen. Disciplinary, a field-based claim, for example, there’s 10 things. One more, disciplinary. Well, that’s where AI might be used in law enforcement to help them convict a criminal or something else, catch lawbreakers. And if that is so, if the AI system actually helps the policemen find the criminal, then you can do research on that. And that would be called disciplinary or field based research. You’ve actually done something.
There are a significant factors and that’s all academic language, and if anybody’s listening who’s not an academic professor is like, “Oh, just turn off?” There are ways to measure it because you must be able to measure it. It’s like a yardstick. You must be able to measure its significance to make a difference in a human’s life, a group of human’s life, a business life, military. If it makes a difference and you can measure it, then it’s called significance. All these universities are trying to find something that does that versus just use AI and say, “Oh, that’s kind of cool.” That’s good enough for us common folks. It’s kind of cool. If it makes it better milkshake, that’s cool. But we don’t really worry about its significant factor.
Dr. Wanda Curlee: Thank you, Oliver. This has been a great discussion. Do you have any last words for our listeners before we sign off?
Dr. Oliver Hedgepeth: I would say that AI and robotics, it’s not something new, much like barcodes and radio-frequency identification tags, little chips, computer chips that are tracking the purchases in a store. You don’t really think about that little device on a gown or something or suit you pick up at the store. It’s just there, but it’s been around for a while. And it’s technology, it’s part of AI, it’s tracking you. It’s here and it’s for our benefit really.
There are some funny mistakes I’ve mentioned, but it’s really here to help us much. When you’re on your laptop computer or your iPhone and you’re typing a text message and to say, “Hello, Wanda. How are you doing today?” And you mistyped the word, it will come back up and correct your word. And then you come back and say, “Oh, I’m doing fine.” It will give me three or four possible answers based on who you are, what we’ve talked about in the past. And I’ll just select the answer the AI system has given me, instead of typing an answer.
It is all over our area here and it’s making our job, our living, I think a little easier. We can sit back and enjoy the breeze and enjoy the birds out back because a lot of the AI has taken away a lot of manual labor you might say, or manual thinking that we had to do in the past. So yeah, I think it’s here to help us, I really do.
Dr. Wanda Curlee: Thank you again, Oliver. And thank you to our listeners. We look forward to the next one on AI. We have all kinds of different topics that we can talk about. Thank you, and stay well.