Podcast with Dr. Wanda Curlee
Program Director, School of Business, American Public University
and Dr. William Oliver Hedgepeth
Faculty Member, Transportation and Logistics Management, American Public University
Could artificial intelligence, or AI, one day replace human project managers? In this podcast, Dr. Oliver Hedgepeth talks with APU’s Program Director of Business Administration, Dr. Wanda Curlee, about her 30-year career as a project manager and how she anticipates AI will change the future of project management.
Start a transportation and logistics management degree at American Public University.
Learn what tasks AI will likely take over, the benefits of using this technology and her projections about advancements in the next five years. Also, learn why it’s critical for companies to start conducting research and planning for how to incorporate AI into their project management systems in order to assist with bidding, budgeting, operations and more.
Subscribe to Innovations in the Workplace
Apple Podcasts | Spotify | Google Podcasts
Read the Transcript
Dr. Oliver Hedgepeth: Welcome to this podcast. It’s Innovations in the Workplace. I’m your host, Oliver Hedgepeth. And today we’re going to be chatting about artificial intelligence and its impact within project management, something that people don’t think about a lot today.
And today, my guest is Dr. Wanda Curlee. She’s a program manager of business administration and is also a professor at American Public University. And she’s got many years of experience and research in artificial intelligence and especially project management.
Wanda, welcome to Innovations in the Workplace. Tell me a little bit about yourself.
Dr. Wanda Curlee: Well, thank you, Oliver. It’s great to be here. Well, I was a project manager, program manager, project portfolio manager for over 30 years in business. I also did as a contractor to the government for many years as well.
I’ve worked in companies like AT&T, Deloitte, doing all kinds of different projects that were kind of fascinating. I also am doing a lot of stuff within project management with my organization at American Public University. So I’m glad to be here. I’m glad to talk about how artificial intelligence and project management do intersect and should intersect. So let’s go for it.
Dr. Oliver Hedgepeth: There’s a lot of stuff out there, you’re right, in project management world and AI world. Let’s just start off simply with what do you think the creators really of project management, the people who develop the software, the project management software, have been doing incorporating AI? Or have they been incorporating AI into the project management software?
Dr. Wanda Curlee: Well, it’s interesting that you say that. I am very disappointed with the amount of AI that there is in project management, especially in the softwares, the tools that you can buy off the shelf, such as MS Project.
MS Project is now on a server kind of basis. You have to buy it on a monthly basis from Microsoft. Microsoft has gone to that different type of buying. You can’t buy their software anymore.
So they have maybe just a little bit, and I mean just maybe a thumbnail sketch, of AI within scheduling. They do some scheduling things that it will come back and ask you, “Does this make sense?” It’ll do some what-if scenarios, but it’s very limited in what it does and it’s only within the system. It doesn’t go out and look for data in the company, so to speak, of what you’re working in, because that’s where you want to look at all that data, that massive amount of data that your company most likely has.
But I was doing some research for this podcast and I came across something that Accenture is doing, which I found quite fascinating. They understand that many companies have mega data. They have thousands and hundreds of projects every year, if it’s a large company, and they use multiple vendors, clients, and partners.
They realize that companies can understand that why projects didn’t come out with the budget and the schedule that they expected. So the decision makers couldn’t understand all the complexities and risks that were inherent in their projects.
So Accenture decided to do something about that. They wanted to come up with a system that predicted the risk tier of each contractor project, but also they wanted to have an actionable explanation of these predictions.
So what they did was in Accenture Labs they did what they called Applied Explainable AI. And it developed a five-stage process as to explain the risk tier of each project and contract.
This is what we need. We need smart systems that can reach out into the corporation’s information. That’s what AI is good at, is slicing and dicing the data and maybe providing recommendations.
We as humans, project managers don’t have the time, the bandwidth or the capability to take in all this data and understand what it means. It would take us hundreds, if not thousands of hours or a PMO — a project management office — that would be staffed with hundreds, if not thousands of people looking at this and coming up with information.
So we need software out there, the risk-management software, the contracting software, the invoice payment software, all of that, to understand what our partners are doing, what the subcontractors are doing and how is it affecting our project. And we’re just not there yet.
Obviously, Accenture has done a baby step towards it. I’d like to see us get to a toddler, if not a teenage system in place.
Dr. Oliver Hedgepeth: That’s very interesting. I know that AI is used, as you mentioned, for a lot of this mega data. That’s very useful. It’s how it started out. But you did mention risk and there are risks and problems of projects that go off track. If a project is successful, everybody goes home happy.
And you’re talking about the projects that fail and the many hours of time to try to figure out why, what went wrong. You have any examples of like a project that you’ve done in that past, that it took so long and you might be able to say, “Yeah, this AI, if it had come up and said, ‘This is what’s wrong.'” Is that what you’re trying to get this risk AI software to be?
Dr. Wanda Curlee: In a sense, but even small projects today have a lot of data that they should have access to. For example, let’s say you’re on a small project and you have no idea, anything about the vendor or the partners on this small project. Let’s say it’s a three- or four-month project.
And [you] come to find out through serendipitous means, one of the subcontractors has been notoriously late, has notoriously sent out bad information. If you had had AI to go out there and scour the data for this contractor or subcontractor, you would have known that and you would have been able to micromanage this subcontractor.
Let me go into one that I did many, many, many, many years ago. It was one of the first projects I was ever on. And it was done by the Air Force and we were contracted to the Air Force. And the Air Force wanted to come up with a paperless office and a non-hackable email.
Now, this was in the ‘80s. So you can imagine what that meant. I mean, most people hadn’t even heard of email in the ‘80s. And the project did fail, by the way. Congress canceled it because it got so expensive.
But we were constantly coming up with roadblocks. They were technical roadblocks. And I can’t help to think that if the engineers had had AI to go out and scour everything that had been done within our company, and the company was AT&T, if they had gone out and scoured all the information out there on similar engineering problems, that we would have had something that would have been feasible for the Air Force or we would have found out very quickly to tell the Air Force, “This is not doable. We can’t do it.”
What we ended up doing was floundering around for 18 months, costing the taxpayers money, costing us embarrassment, and having Congress finally coming in and say, “You’re 100% over budget. You’re nowhere near what your schedule should have been.” We were showing that we were going to be behind two or three years.
So if AI had been available back then. A non-mature type of AI was available back then, but it wasn’t cost-effective to put it on our project. And most likely, many of our engineers hadn’t even heard about it. So AI is there to help us with gathering all the data.
Dr. Oliver Hedgepeth: Yeah, I really like the way you brought up the example of the paperless office back in the ‘80s, because the idea was great. I was involved in that too; as you know, we both worked in AI in those timeframes.
And the paperless office has always made more sense. I remember being told we wouldn’t be using No. 2 wooden pencils anymore or paper to write on because computers are going to replace all that. I understand it. But the paperless office is a great metaphor and a segue in what you’re talking about, in terms of AI being part of project management.
Project management is a process, many steps to follow to make sure your project is done properly and it’s beneficial and there are risks identified, as you’ve mentioned.
But can the concept, the metaphor of the paperless office be used today with AI being used in project management? Or let me pose this one, can AI replace the project manager or part of what the project manager does?
I mean, project manager is you. You’re a project manager. People hire you to manage a government or civilian project, to build a new software system or a new bridge. Can AI, the AI system, not only capture the data, but be used to manage that project, taking you out or pushing to a side, or you do something else? What do you think about that?
Dr. Wanda Curlee: Interesting concept. I would like to think that AI would be an adjunct to the project manager, allowing the project manager to do the more value-add to the project.
I see AI going out and doing the mundane things and repetitive things such as updating the schedule, chasing down people to put in their information about the schedule, looking at risks, getting to know me and know my biases.
So for example, if I always look at risks that are going to happen maybe in the next three months, AI might go out and look at all the risks on similar projects and what’s going on on this project and say, “Hey, Wanda, you’re only looking at risk for the next three months. How about this risk that’s going to happen most likely in month six, which will topple down your project?”
So AI has now given me a segue to say, “Oh, gee, I hadn’t even thought about that.” So it will push us to do the value-add. So now I can go and look at that risk in the next six months and say, “How am I going to mitigate it?” Or, “Am I going to accept it? What am I going to do?” I can brainstorm with the team and bring everybody around.
Dr. Wanda Curlee: Other things that I see AI doing as well is, and I don’t think we’re quite here yet, but I think it will be here probably within the next five years, if not sooner, to draft emails for me. So for example, if I have a meeting coming up, it would automatically look at my schedule and send out the invites to everybody. I might have a chat bot that goes out to go find somebody that I need to talk to immediately. I could say, “Hey Alexa, go find so-and-so. I need to talk to them.”
Another thing that they could do, if I do PowerPoints, it would start to understand how I like to do PowerPoints, how to do status PowerPoints for me, send it out or let me, once it’s put together, give it back to me, even for more complex PowerPoints, send it back to me and then I could have what I like to do. And the AI would be understanding that, understand what I’ve made changes and would know how to format the next PowerPoint. So these are some of the things that I see AI doing. Will it eventually replace me? I hope not. I don’t think so.
Dr. Oliver Hedgepeth: You’re really making it exciting here in terms of I’m listening to you and I’m hearing behavior, I’m hearing human behavior characteristics, such as chatbots and emails that will be generated automatically. But you’re saying that the AI systems not only doing schedules, not only doing crunches of lots of millions of pieces of the data to find out what the average is or something or where the direction is going on the project, but that behavior is interesting.
If you’re going to be getting an email from an AI system, we’ll call it PM bot or something, the project manager bot or something, a robot. If you’re going to get an email from somebody, I’m sitting here and all of a sudden I’m a worker, one of 5,000 workers at a shipyard.
And I’ve got a notice. We’ve got to come together tomorrow morning to meet. Okay. That came from the robot. Okay. The program manager bot.
But when I get to the meeting or maybe it’s just project managers together and I’m on a Zoom meeting with 20 other project managers and I’m talking to you and I’m talking to other people in the team, who’s in charge?
There’ll be a project manager, but if the project manager is allocated a majority of this human behavior stuff to the robot, is there going to be a Zoom screen with a little character of a robot saying, “Oh, Wanda, you got to do this next. Oliver, you’re slowing down on this.” Is that going to happen? We’re going to have a Zoom meeting with a robot picture?
Dr. Wanda Curlee: It could happen. Absolutely. Like I said, the mundane things. If I’ve got the project managers together because they’re not putting in their time or their teams are not putting in the time, I don’t need to spend my time on that.
Yeah, the robot, hopefully it would have an avatar that looks more human, would say, “Hey, Joe, your team is not entering time. And because of that, we’re unable to do what-if scenarios if we’re running ahead or behind schedule.” Or it could go to Sam and say, “Hey Sam, thanks. You’re doing a great job. Everybody’s putting in their time.”
But then on the other hand, if it was a more complex, we have an engineering problem, and we’re trying to solve that.
Ultimately, it’s the project manager’s decision on how to run the project and what to assess. It could be that AI comes up with a solution. And I know for a fact that solution can’t work because of XYZ. So I discard it. That’s where I come in as the value-add, or the team comes in with the value-add.
For example, I said that the AI came up with solution and I brought it to the team and the team would say, “We can’t do that because that doesn’t exist yet.” Or, “We’re using this contractor right now and they don’t have the capability.”
Well, if somebody came back and said, “Well, that contractor doesn’t have the capability, could we maybe contract, subcontract, with another vendor that does, that could help us through this?” So those are all the things that we need to do.
I, as the project manager, have to understand what the AI is telling me. Does it make sense? Do I present it or not? I also tell the AI what it can and cannot do with the team. I might tell it that you can’t present any risks to the team. That all goes through me. Or I might say, “The only thing you do with the schedule is input the data that you heard from the team.” Or I might tell it, “You could gather all that data, but before you input it, you have to get permission from me.”
That probably is not the best use of AI, but again, it’s ultimately up to the project manager and I would hope the project manager would be transparent on what the AI is doing because the team has to have confidence in the project manager and trust and understand what the AI is doing and what the AI is not doing.
Dr. Oliver Hedgepeth: Trust is very big. You are the project, the human project manager. They’ve got to trust you. But can the AI system, should the AI system be talking to the team members about doing things, sending those emails out? Or are you sending emails out? Or is it sending emails out under your name?
Dr. Wanda Curlee: I think, in my opinion, what should happen because people are not used to AI is maybe initially, although you’re transparent with the team, initially the emails go out under your name so that the team creates confidence that these emails are done well.
So for example, I could say, “Hey team, you’re going to be getting an email to enter your time. It’s an automatically generated by the AI system and I expect you to do it.” But as time goes on, especially if it’s a long project, I may tell the AI, “I don’t want you to send it under my name anymore. I want you just to send it out and follow up as needed.”
I would let the team know, “Hey, you’re not receiving this anymore under my name. You’re going to receive it from the chat bot or the AI system,” whatever we decide to call it so that they understand. It’s the same email, the same information, but it’s now coming from the AI. I think trust is a good thing. It’s absolutely necessary, especially on virtual teams. Studies have shown that tremendously, that that’s needed.
So I think as the team and as us as a society become more understanding of what AI is and what AI isn’t, we will see that AI will be part of the project, as I was saying, especially on mega projects, AI has gotta be there.
Initially, it might be the project manager, especially in the first five years or so until people … I think the Millennials and Gen Z generations are accepting of AI. So they wouldn’t be surprised to see an AI-generated email coming from AI.
Whereas, people in my generation might be mistrusting. They’re saying, “Oh, a computer is telling me what to do. I don’t trust this thing. It might go off the deep end.” So anyway, that’s how I see it.
But I think five to 10 years from now, companies are going to decide how the AI is going to be integrated into the project. And I don’t think the project manager or the team will have any say, but it will be up to the project manager to instill trust into the team that the AI is adaptable and understandable and will help the project.
Dr. Oliver Hedgepeth: It’s interesting, we’re talking about AI in project management, Wanda. And you just raised some good points about the team members being uncomfortable with an AI system working with you, as the project manager, to provide advice and information on what’s wrong, what’s right, what risk is happening, but you’re already living that.
Now, let’s go back project manager. You’re a government employee. Your projects were things like submarines, aircraft carriers, little things like that, where you’ve got 15,000 to 25,000 workers. So there’s a big project manager over all of those people and as a bunch of other 100 project managers, sub-project managers, there.
But all those 15,000 people got to work every day and they might need an email from somebody. You can’t send them out an email. You could say, “Here’s what we’ve got to do.” But each one of them is working on a specific project.
The AI system would really be able, I would think, to capture what each person is doing, each of those 15,000 is doing. And if they saw that Bob was, his work hours were a little slower and he’s supposed to turn a widget three times an hour and he’s doing it only one time an hour, he don’t need a personal email. The AI might be able to send him a note, “Hey, ratchet it up!” versus you might have to take a week after looking at, or a month, after looking at a monthly report. What do you think?
Dr. Wanda Curlee: Yes. I think AI will can probably do that now. Not probably. I know it can do that now, if it’s asked to go look at that and if that information is programmed.
Remember, AI is a computer program. Somebody’s got to program it. But we’ve got to balance that, as you said, with human behavior. Do we send that email out to that employee saying, “You’re turning the screwdriver only three times an hour, whereas everybody else is doing it 15 times an hour.”
It could be that person is not feeling well that day, that person is having terrible problems at home. There could be many things. And maybe they injured their hand.
But yes, we have to understand all of that, but we also have to take in human behavior. We’ll also have to take into account unions, if unions are involved. So yes, the AI could come right to that team leader of Joe, let’s say, who’s not doing the screwdriver correctly, or come to me as the project manager and say, “Joe is not doing this.”
Do we go out and chastise Joe? No, I don’t think so. That’s where the value-add of the project manager comes into account.
We go and talk to Joe, “Hey Joe, what’s going on? What’s happening? You’re steadily going down in the number of turns per hour. You’re down to three an hour. Have you got something on your mind? Can I offer you to go to employee assistance to help you out? Are you injured?” Things of that sort.
So AI and the human have to work in conjunction. We can accept that what AI tells us is correct in its mind. AI would go out and say, “Hey, Joe, you’ve got to start doing this or you’re going to be fired.”
Well, that’s not always the situation. Human behavior has got to be taken into account.
Dr. Oliver Hedgepeth: In five years, that’s long ways out. Here’s 2020. 2025, could you see a project manager being a 100% AI system for a project that may include only five people or 15,000 people with some range in there? Could there be a project of a size that could be, yeah, let an AI system run it with a human who’s kind of overseeing it and saying, “Oh, I’m looking at it. It’s doing okay.” But it doesn’t really get involved with the people at work. Do you think in five years there could be an AI project manager?
Dr. Wanda Curlee: I would say yes. And those companies that don’t will end up losing a lot of profit. Their ROI will go down. So the big companies that, as you said, if it’s a smaller project and even big companies have smaller projects, yes, the AI should run it and with a person overseeing it. That’s where I was getting at.
Now that we have AI helping us, in five years I may be able to run two mega-projects or maybe 10 small projects because I have AI. And if companies don’t start getting on the bandwagon and allowing their AI systems to interface with project management, they are severely behind and they will find out that it’s going to hurt them terribly.
In fact, the Project Management Institute did a research on AI and projects and programs, especially on mega projects and programs such as you said, building submarine or an aircraft carrier. They found that those companies that are using AI now are more profitable than those that are not that do similar things. They’re finding that in the future, if projects don’t use AI, even on small projects, they are going to be eroding their bottom line. And I don’t think any company wants to erode its bottom line.
Dr. Oliver Hedgepeth: The government will be leading this effort to build an AI project management or we’ll have to rely on companies like Apple or whatever, Microsoft, some other software firm? Who do you think will build the big AI system project manager assistant, let’s say? Who will build it, the government or some civilian organization? Who will build that first one in five years?
Dr. Wanda Curlee: I’m going to guess it’s going to be a civilian organization. Now, government may give them some grants to do it, but I think it will truly be civilian.
We’re seeing it with Accenture. Accenture already did that one little area of risk. It wouldn’t take much more to build it out. The problem is when it’s sold to a company, that company decides what that AI system can go out and retrieve. And many times, there’s leaders within companies that are somewhat protective of the data may not believe in AI yet. They’re leaving the project manager without the data it needs.
The powers-to-be in the corporation may say, “You don’t need to go out to vendor payments.” Well, what if I found out that this project, this subcontractor was sending invoices that did not correspond to what they truly did? Maybe out of ignorance, maybe out of theft, who knows? But at least it would put me as the project manager on pins and needles to understand what’s going on with that vendor.
And maybe I could then tell the AI, “I want you to really look at these vendor’s invoices.” And let’s face it, invoices sometimes are wrong. Even if you have an automated system doing it, it’s wrong. Or there may be a dispute between the two.
So like I said, companies are going to have to understand that project managers need this data to do their work efficiently. And there are companies out there, their leadership, that don’t truly believe that project managers do anything special. They just kind of lead the project and get it going. They don’t do any value-add by looking at all the information out there.
And it actually goes even before the sales cycle, if I’ve, as you said, a submarine or an aircraft carrier, if I’m bidding on it, I as the project manager need to understand all of the information that happened on previous things that we have done. Or if we haven’t done something similar, then have we built a commercial ship and how do we transport that over into the government arena and understanding government?
AI can help me with that. They can help me understand the risks so I know how to appropriately bid this project. Look at all the government projects and programs that are severely over budget and severely over schedule.
If we had AI, the government would even have a favorable outcome. So the government’s got to use it as well. They’ve got to be able to go look at all their data and say, “Gee, we’d gone to this shipbuilder five out of 10 times. And every time we went to that shipbuilder, the ship was two years late. Whereas if we went to this shipbuilder. And we’ve done it maybe four out of eight times, they have been on time, but they’ve been over budget by 100%.”
Now the AI can help the government understand, well, maybe I do things differently. Maybe I don’t just go to one general contractor. I go to several. So it’ll help on both sides. The government and with us. But I think it will be civilians that will eventually come up with a virtual project manager, so to speak.
Dr. Oliver Hedgepeth: Okay. The project manager. And you mentioned a term here, bidding. Yes, the companies, the private industry companies, they bid on the projects. I worked for a major company back in the ‘80s, and whenever we wanted to go after a major defense contractor, and we built things for the government, not a software, this happened many times or several times I was there: The company would push aside $1 million and that $1 million would hire for one year 200 different employees from around the world.
We’d get the best of the best and we would sit in a room full of computers, rent an office somewhere downtown Washington, DC for a year, and we’d have meetings with government people and other people. And we would use that year to develop the bidding. We would develop at the end result reams of papers, lots of meetings, to bid on the contract, going against other major companies in Washington, DC bidding on the contract. And they would be doing the same thing.
So we were spending a year, a million dollars to come out with the right number. Here’s what we’re going to pay you or we want to be paid to build that submarine or that aircraft carrier.
So there is another part of project management before you become the project manager, that’s the bidding part. Do you think that could be replaced with AI and save a million dollars and 200 people? You think something like that could happen in five years?
Dr. Wanda Curlee: I don’t know about five years. I would say, yes, that’s going to happen and those companies that are now using AI for bidding and going to previous projects to see what the risks were, what were the delays, what were some of the other issues that happened, what subcontractors did we use that were good, what were bad.
If they’re not doing that already and going through all that mega data and talking to project managers, then they are going to miss the boat. Because many times companies don’t save the project data. It resides on the laptop of the project manager and that doesn’t help anybody.
So companies have got to understand that they have to include the project data with all the amount of mega data that they have in their corporation. Don’t let the project manager just keep it on their laptop. Obviously, if it’s a ship or a submarine, the schedule’s probably not on somebody’s laptop. It’s just too big for one project manager.
It’s probably sitting on some server someplace, but again, AI has got to be used in all phases of the project. You could even have an AI that goes out there and understands when we think the government is going to put out an RFP on a certain thing that we’re interested in because we know it’s coming.
Well, AI can go out there and assess, well, this government agency is notoriously six months to a year late on putting out their RFPs. Or they are within three months after they put in their request for information, which is usually a precursor to the request for proposal, that between the RFI and the RFP, this agency takes 18 months. You know you’ve probably got about 18 months to do what you need to do for this project.
Dr. Oliver Hedgepeth: In terms of developing an AI system for project management, who is going to be working as a competitor or a country first — China, Germany, Finland? Who do you think might be doing that?
Dr. Wanda Curlee: China is actually doing a lot in AI, and so are some countries in Europe. As to who might do it first, that’s interesting.
Dr. Oliver Hedgepeth: But it’s there. Isn’t it real? It’s got to be completed. We’re smart people in California. We’ve got a lot of AI folks, the Alien Technology. It’s a company I know out there in California and they are out there working. I know MIT and Georgia Tech, they’re working on AI stuff all the time. Do you think there might be somebody out there getting ready to send that first thing out, project manager, try to sell it to us. China maybe, or…..?
Dr. Wanda Curlee: Well, I think China probably may develop one sooner rather than later. I think they’ll probably even have one before us.
Will we buy it? I highly doubt it, especially based on what’s going on between the U.S. and China right now. But I think we will be hot on their heels and we’ll probably develop something about the same time that any European country does. I think Germany and Finland and [the] U.S. might be on the same track for developing an AI project manager.
Dr. Oliver Hedgepeth: Now, all this sounds nice and neat. And all this software we’ve got on the computer is so nice and neat. And it’s so cheap. Oh my goodness, it’s cheap, compared to what it was in 1984, 1985. This AI PM bot, AI PM bot, call it a bot, it’s going to cost money.
And if you want to have a project management AI system that will help you out as a project manager, someone’s had to spend probably millions of dollars and decades of time developing that little AI project management bot. Are you going to be able to afford it as a project manager, to help you or to run a major part of your company? How much do you think that AI PM bot’s going to cost?
Dr. Wanda Curlee: Well, let me step back to MS Project. When MS Project first came out, consultants couldn’t afford it. It was very expensive. But as Microsoft saw that consultants who are onesy, twosy people, or maybe just a medium-sized company, weren’t using Project, they realized that they were missing a big part of the economy.
And there were many companies that refused to buy the server part of MS Project and just bought laptop versions for all of their employees. But Microsoft soon realized that if they didn’t get their software out there to be the leader and make it affordable for people, then they were going to lose the race. And they did that. They followed that.
So I can’t help to believe that there will be early adopters of the AI project management bot that most likely will be big companies. The Deloittes of the world, the AT&Ts of the world, the HPs of the world, et cetera. Facebook. I’m sure Facebook’s already using AI.
But I think that there’ll be several companies that will develop an AI bot about the same time. And one of them will be smart enough to understand that if they don’t make a smaller version for the consultants in the world that are not associated with big companies, will need it as well.
So if they don’t provide them a smaller version, then they’re going to lose the ability to say, “I’m number one in the industry.” Or, “I’m number one in sales.” Or “I’m number one in XYZ.” Like Microsoft is doing now with MS Project. I mean, who doesn’t know MS Project if you’re a project manager?
I think it’ll be the same thing with AI, AI bots. And how much will it cost? I think it will probably cost, by the time you implement it and interface it with all of the systems within a company, it will be very, very expensive, probably multi-million dollars.
But as it becomes more sophisticated and understood, and as AI is developing it, so it doesn’t take hundreds of software people to code it and develop it, it will develop itself after a while, then I think you’ll find it come down in price quantumly. I think it will probably end up, in maybe five to 10 years, there’ll be a smaller version for consultants that will be maybe a few hundred dollars.
Will the companies allow that consultant to patch into the company’s mega data? I doubt it. I highly doubt it.
So I think what we’re going to see is that when companies hire somebody that wants to use AI on the project, then they’re going to see that the person will have to use the AI within that company. And then it’s going to come down to training. They’ll have to be trained in the AIs that are out there for project management so that they can go to companies and say, “I know your training. I know which AI you’ve done. I’ve done this training.” And they’ll be able to show them.
Dr. Oliver Hedgepeth: I remember. You mentioned the early days of MS Project. I remember back in the ’80s and when I was working with private contractors, after finishing the government project management work I used to do. And I remember companies out of California spending $25 million to buy this great, big new software package that’s going to make project management so easy. And the problem was that it cost around $50 million to $75 million to train it for that company to be familiar with that company’s personality, that company’s data, that company’s expectations, that company’s past risk. It had to be trained.
So if we have these AI PM bots in the next five years, one AI PM bot for Ford Motor Company might not be useful for AI PM bot to build that ship at Newport News Shipbuilding. You think all these AI bots, don’t they have to have some training? You mentioned training. And if so, that would add a cost, wouldn’t it? And wouldn’t that be time consuming, before you even grab hold of it?
Dr. Wanda Curlee: Having done enterprise-wide software implementations, we all know that they come wrought with a lot of ups and downs and risks, and we have to put the personality of the company in there.
I think that will happen with the AI bots, the project management bots, because they’ll have to understand, well, what is it that I’m building? Is it a service? Is it a product, et cetera, et cetera.
But I think what we’ll see is that AI will do that for us. So it will come down … And the amount of money that it costs. Yeah, there’ll still be a human in there to make sure that what the AI has put into the system meets the needs of the client. But I think you’ll have AI PM bots that are flexible enough to learn the industry and the company and do it in a relatively quick fashion and reliably so that you could have it at a shipbuilding yard or you could have it at a software company.
Dr. Oliver Hedgepeth: All those characteristics behaviors have to be learned. So maybe part of an AI PM bot that you buy has to be trained early on before the project starts. I’m just trying to figure out, when would you start training it? You can’t start training that day one the project starts. We’re getting ready to build this Ford Motor Company, or we’re getting ready to build this aircraft carrier. When do you start training that robot?
Dr. Wanda Curlee: What I see happening is a consortium. And I think you see that with some software programs, where a company, maybe it’s a research area, will create a consortium with an AI system, project management system, that has shipbuilders, it has software companies, it might have social media companies, et cetera, et cetera.
And they all come together and they feed and teach this AI system so that when the AI system is getting ready to launch, this consortium has already fed all that information into the AI system so that it doesn’t start at ground zero. It understands some things.
It may not understand the people on the project, but it would be able to learn them let’s say within a week or two. And then be flexible enough so that as maybe a brand-new employee who knew nothing and was probably delivering very little or nothing on the project may turn out in six months to be one of the star performers.
The AI is so quick that it should be able to learn all of this and do all of this within a matter of a day or two.
As far as the information on understanding the industry, I think we’ll see, like I said, consortiums for different AIs, program management AIs, that impact put the information into that. I know I’ve seen that on other project management software. I’m trying to think. I think it was around risk project management software, that if you were part of the consortium, it already had risks for your industry, for the type of project you were doing, if you were doing it with government or you were doing it with commercial. So it would automatically populate all those risks, depending on what you put into it.
Dr. Oliver Hedgepeth: We’ve had a good conversation with you, Wanda. What kind of final word would you like to say to somebody who’s listening to this about AI in project management?
Dr. Wanda Curlee: Well, what I would like to say is that AI is all around us. Whether you believe it or not, it is around us. Think about Alexis or Google. Those are AI systems. They may be in their infancy, but they are AI.
Will we have something like a Terminator in the future? I hope not, but we will have project management bots. It’s coming. It’s there. If you, as a project manager, or if you’re listening, from a company and you’re a leader in that company, are not using AI to help your project, you are going to be left in the dust.
Those companies and those consultants that use AI in their everyday work will end up being the leaders in forefront and doing marvelously well in project management. Will you still have hiccups? Yeah. We all will. AI learns just like we learn. So yes, there’ll be some hiccups, but I’m excited.
Dr. Oliver Hedgepeth: Okay. Well, thank you. For you listeners out there, we’ve been talking to Dr. Wanda Curlee. She’s a Program Director at American Public University. This is a really exciting and I think all major companies should really invest in thinking about this. I’m sure they are already, as you mentioned. Wanda, thank you very much for joining us today in this episode of Innovation in the Workplace.
Dr. Wanda Curlee: Thank you.
About the Speakers
Dr. Oliver Hedgepeth is a full-time professor at American Public University (APU). He was program director of three academic programs: Reverse Logistics Management, Transportation and Logistics Management and Government Contracting. He was Chair of the Logistics Department at the University of Alaska Anchorage. Dr. Hedgepeth was the founding Director of the Army’s Artificial Intelligence Center for Logistics from 1985 to 1990, Fort Lee, Virginia.
Dr. Wanda Curlee is a Program Director at American Public University. She has over 30 years of consulting and project management experience and has worked at several Fortune 500 companies. Dr. Curlee has a Doctor of Management in Organizational Leadership from the University of Phoenix, a MBA in Technology Management from the University of Phoenix, and a M.A. and a B.A. in Spanish Studies from the University of Kentucky. She has published numerous articles and several books on project management.
Ready When You Are
At American Public University, students are priority one. We are committed to providing quality education, superior student resources, and affordable tuition. In fact, while post-secondary tuition has risen sharply nationwide, the university continues to offer affordable tuition without sacrificing academic quality.