This transcript was created with help from Ai. Please report errors to us and always refer to the video as your primary source.

Justin Brady: From the jetro Barbecue Studio, this is the Iowa Business Podcast, and of course, I’m Justin Brady. Whether you Like it or not. AI is here and it does introduce a lot, a lot of opportunity for organizations, small and large, like crazy, absolutely mind blowing stuff, which we’re going to get into today. With the opportunity though, comes a lot of risk for your organization. I’ve already seen people adopt AI tools, Very small, businesses. And then these ais are doing renegade things that they were like, oh crap. Now nothing. really, really awful destructive in those cases. But if you’re integrating this in your large organization and you don’t know what you’re doing or a medium organization, that could be you. I mean that probably already is you, if we’re being honest and to talk about that. Chris Draper. the CEO of Morrigan AI joins us today to talk about risk and integration, all this stuff. Basically how our leaders exposed is what we’re gonna be talking about. So thanks for coming in the studio. No thank

Chris Draper: Thanks for joining me. Thank you. A

Justin Brady: It’d be a really good discussion. Uh, before we get started, I had Kaylee Blaik Rajo in here from Dentons Davis Brown about immigration law changes, and of course, the a hundred thousand dollars fee you may have to pay if you want an H one B visa worker and how to get around that. Why Iowa is uniquely positioned to actually grab some top talent. We hear a lot in this state about how Iowa can’t get top talent. It’s so hard. She has a very different, very, very smart take. I would check that out if I were you at Iowa podcast.com/business. So how many, Chris, obviously appreciate you being in here. How many Iowa companies do you think are like, have a problem. with employees and people and even executives using unsanctioned AI tools in their organization?

Chris Draper: Well, I think the unsanctioned uh, usage or, so the shadow AI usage is commonly called, uh, the, your main numbers are believing that somewhere between 60 to 90% of, of usage of unsanctioned tools, uh, that’s probably accurate. When you think about what is out there, the number of people who have, uh, used. You know, Chachi, bt, or Claude, or some of these tools in some form, grok in some form and thought, oh boy, I’ve heard all these things. Maybe it could do x. Um, I would say that the number of people using ’em in some off cases is extremely high and much higher than we think. Uh, the part in some ways though, it’s probably similarly relevant, but almost more concerning in some ways, is the number of folk that have, uh, the co-pilot, uh, just added on into the Microsoft P. So the Gemini just add on to that Gemini piece and really don’t know when they’re using it, what they’re actually doing with it. And what I mean by that is, um, is, is I think. Often understood. These tools will sometimes give a feeling of certainty to something that is highly uncertain.

Justin Brady: Mm-hmm.

Chris Draper: They will often be giving you more answers that are in line with what you’re hoping to hear than what is the best thing to be thinking about. And so understanding how they work at a general level, understanding how we should be introducing them in a more strategic way that is a bit more thoughtful and really human-centric is what is really vital to be done. So it’s not necessarily the fact that these tools are being used a lot more than we think. It is the problem that we see and that we need to be addressing is what is it we want these tools to be doing? And more importantly, what is it our company really is expecting? It’s wanting done.

Justin Brady: Yeah, you said six. Was it, I just wanna make sure to put a, a exclamation point on this. You mentioned 60 to nine, somewhere between 60 to 90% of organizations are, or excuse me. Um, 60 to 90% of organizations have shadow AI projects running in their companies right now.

Chris Draper: And shadow AI being something that where an individual may be. Um, there was one that we saw that was probably less, uh, thoughtful than others where uh, they needed to restructure the presentation of their customer list for, uh, a leadership meeting. So they just. Dumped it all into Chad GBT to get out a prettier Excel format, or they decided that, uh, hey, we wanna go get a lookalike, lookalike audience. And so we’re just gonna figure out, I have this free version of Claude, so I’m just gonna like dump in. Here’s all my current customers. Find me all the next ones that are similar to that. So those types of small uses, um, are probably more common than we expect the numbers. No one wants to admit they’re doing something wrong. So saying what the number is, uh, where I’m going with those is between McKinsey and Wired and a number of different reports have gone out there to say Infotech has a lot of data on how much shadow we is used. It’s obviously very wide range. A 30% range in any number indicates it’s uncertain, but it could be incredibly high. And when you’re just noticing. When you’re working in these small to mid-sized businesses, what’s actually happening and watching the process flow, I wouldn’t be surprised if a vast majority are experimenting in ways that are going to be, um, fundamentally data breaches.

Justin Brady: I mean, it’s hard not to because the tools are so effective. And what I’ve noticed is most people, they, they’re so amazed by how they work that these AI tools that they’re almost don’t care about accuracy. It’s like, well, It’s you know, done is better than not done at all. And so I have noticed a lot of people that have this weird bias. I don’t even know if there’s a term for it, but they’re like, well, it, it. It has the perception that it mostly got the project done, so it’s probably fine. And then they just kind of submit that work. We’ve seen lawyers do this. We’ve seen people like lose their licenses over this and it doesn’t seem to bother people. So I don’t know what that breakdown, is, but maybe it’s the conversational aspect of it. But are you seeing a lot of people giving, we’re gonna talk about all the. like the risk for having AI in your organization, but focusing in on one thing particularly. Is this mostly like a privacy kind of thing? Like people, you mentioned people uploading information into these tools, co-pilot what whatnot. Is it mostly a risk of privacy? Like they’re uploading private data into these tools and therefore that data is no longer private? Is that the Big. risk?

Chris Draper: No, I think the big risk is, I, I don’t believe so. So, I mean, I think the big risk with these tools first comes down to an understanding of what they are and how they, how they work. Um, and I think that that isn’t really appreciated enough. We just had, I’m a vice president of I Coder, which writes the standards for online dispute resolution, which is, if you’re not going to court, if you’re gonna mediation and if you use digital tools in that ico, or is the backbone of the ISO standards for how you should be using ai. And we’re using these online tools for alternative dispute resolution, and AI is a big issue of it. Um, when we think about traditional technologies, we think about back in the pandemic, uh, video conferencing. The issues in video conferencing or email or Word documents is whether or not the data is now released to the wrong people or whether it’s visible to the wrong people. We think about data sovereignty and data security. Those are all data related reliability type issues. It’s is the data that I have. Exactly like what I thought it was, was it transferred? Where I think it was, was it ensured not to be edited or changed. The data accuracy was important because the use of the data directly is where I got my value, and in a lot of traditional technologies, that’s my major concern. AI is fundamentally different. AI is an informational sovereignty problem. AI is. Problem we see now is exactly like what YouTube started, and I’ll start with what AI does and how AI will these tools take its view of the information of the world. It will then with its filters and biases, create its own data set that is a model, and from that, it keeps generating these synthetic information reflections again after it. So the data itself isn’t anywhere near as important as the refactoring, the recreating, and the reconditioning and the remaking of it into new information. And so anytime we have something that takes, it’s partly broken view of information to create data that it’s now gonna make new information with. We always have to be conscious of the fact that the synthetic information is always holding that bias and that imperfection that was in its creation of it. Now, you point rightly to the fact that most people get comfort from it because it is able to get us something that is synthetic, but looks close enough, often enough. Yeah. That we get this, you know, this, this, this comfort in the data that we start to lose our need to keep rechecking

Justin Brady: It’s almost like you can kind of look through the veil. If you look at a sloppy piece of journalism, I’m a journalist myself. I’ve written for some publications. Um, but you look at a sloppy piece of journalism about something you know a lot about, and you read through it and you’re like, this isn’t, this is Like 80%. there, dude, but This you made some. huge, huge, huge errors. In a way that’s kind of what AI’s doing all the time. It’s kind of a sloppy journalist. It it does, it takes the shortcuts. It does its best job. And so what you’re saying, just to put an A period on this, is that the Bigger risk. for organizations in using AI in their organization, in their ecosystem, whether it be on-prem or in the cloud, isn’t security. It’s actually the fact that it’s corrupting all your data. You don’t even know what’s, what’s, what’s up and down.

Chris Draper: It’s corrupting the information that you’re believing you should be working with. And I think that data and information are something that is usually sort of a nerdy distinction in data scientists, but data informa data and information are two different things, and AI is forcing us to not only realizing they’re fundamentally separate, but how we address ’em are different. And it’s funny that when you use the A 80%. Most of these tools and how we’re looking to roll them in are not gonna get more than 83 to 84% accurate.

Justin Brady: Interesting.

Chris Draper: The LLM tools and just the way that they’re structured are not going to break that barrier for many of the use cases we want to use them for. Um, and that’s because what makes them feel human is the randomness. It’s when I put the same thing in five times. It gets me five different things. And so when I started off earlier, my, my, my temperature was a big item within Chad G Petit, which was a, which was the measure of the, the, the inherent randomness, which is the humanness of it. And if I have something that has a, you know, in rough math of this, which isn’t, you know, if I have roughly a 17% randomness, that means I have a 16% guaranteed wrong. Great. It’s not going to get the per fancy. It depends what your use case is, and this is, this is where the big challenge in the small to mid-size space, which is where we spend a lot of our time in, is. No plumbing company is ever gonna go bankrupt out of AI for hooking the pipes up wrong. It’s not like I’m gonna go in and Chad, GP says, alright, I’m gonna hook the hot water into the electrical board. You’re like, okay, no plumber’s gonna do that in their right mind. If they do well in this case, they’ll be electrocuted. But, you know, if, however,

Justin Brady: care of itself.

Chris Draper: of itself, problem solved in a very non, non appropriate way. However, that’s, you’re, you’re, you’re cross, you’re crossing skill sets there that’s never good, but where plumbing. Organizations always go bankrupt is on the business side. If I’ve now implemented an AI-based tool that’s supposedly going to get me a better, tighter, and more competitive margin, cause I can turn out bids faster and it’s 15% wrong on one side and five on the other, I’m bankrupt. That’s where the true risk is. The risk is in attributing something that has a, you know, a spew quality to it that is close enough. At a speed I can’t catch. So if I’m a small plumbing company, there’s only three plumbers running around and one office worker, I say, you know what I’m gonna do is I’m gonna go off and use this new AI tool to do all of my biddings and it’s gonna do the pricings. ’cause I have this new tool I’ve seen out of Omaha or something that looks at the plans and it say, it goes and estimates, okay, this is all the pieces in it and now I’m gonna, it’s gonna connect into. Home Depot and it’s gonna gimme a parts list from the parts list, gonna price it all out, and now I’m gonna know my pricing. And now I’ve just been able to sign a guaranteed pricing. And now from that I’m gonna put this margin on Bam. And it turns out a hundred bids like that in a month, and their margins are all off. I’m going bankrupt faster. Yeah, all the pipes are hooked up, right? Plumber hasn’t done a bad thing then None of them are electrocuted.

Justin Brady: I, I, already. saw this happen for myself. Now, luckily I caught it and it was very simple use, but I was using a note taking app from a popular tech company that will go be, that will go unnamed. And the client had started talking about Justin, said this would be sufficient, that this was good enough. And I’m like. I never said that. What are they talking about like that? I would never say that in a million years. So Luckily I have two, tools I run parallel. one that just. takes notes, only one that records the context of and has a transcript. Now all the clients I talk to know these tools are being used. It’s right there in Google Meet or Zoom or whatever. Uh, and it even announces itself and that kind of thing, but. What was fascinating is I went through the notes that this popular tech company provided, and sure enough, it had summarized something I had said as saying, Justin says that will be sufficient. But then I went through the context. The context and what I, actually said were way, way different. She didn’t remember what we talked about on the call, but she did read the summary, which threw her off. And So, of course I turned that tool off. But this just gives me a tiny little glimpse into a highly controlled, very isolated use of this where I saw this happen and it didn’t, and I, you know, I turned the tool off. And so what you’re saying is everyone’s, freaking out about data privacy, which is definitely, I’m assuming you agree, is a concern. but the real. problem is the absolute destruction and corruption of all data that you use to conduct Business And this wouldn’t even just be sales. This would be anywhere you reference organizational data that you need to do your job. If you let AI run amuck in your company, All of that stuff could be wrong and corrupt forever. You may not, not even have any records anymore. This is sales data. This is tax data. I guess I should ask like. What does unmanaged AI damage even look like in, in the real world? Like what scenarios are you seeing play out for Iowa companies?

Chris Draper: I mean, yeah, hopefully unmanaged AI is not occurring because when you, when you’re looking there, no tool is truly ready for that. In the same way there, there’s no test law autopilot, right. Put your hands on the wheel. Okay. And, and I think in a lot of ways, um, our bigger concern. So if we step it back a few fold, right, and this is, there was a piece of legislation, um, that was, uh, starting to get some traction within, uh, the legislature this year. And it was actually a really good piece of legislation in intent. And, uh, there was some drafting is going to be occurring to actually get it directly and where it needs to be to get it right. Because the challenge with Iowa that we’re not. Always conscious of is that we are really an information society. What separates your ability to sell a John Deere tractor from the exact same John Deere tractor is your ability to activate relationships. It’s your ability to activate the trust necessary for your customers to work with you. When you look at insurance, insurance is all about correct pricing, which is all about my model being accurate. It’s all about knowing more about the demographic that I’m gambling with than anyone else. This is, you know, I, I, I hate hearing. All the time we hear in the startup community, oh, Iowa doesn’t take risks. Our whole culture bets on the weather every year. That’s our, like our economy is built around the weather. There is no bigger risk than the weather. And you’re telling me that we don’t take risks, we just take the same risks. The fear with AI is that AI is not a traditional technology tool. AI is not like something I buy. I put it on shelf. I plug this thing in and out. AI’s most fundamental risk is that it is a human accelerant and amplifier. It is an exoskeleton technology, not a core technology. And in that, your biggest risks of it are the synthetic realities. And the spew speed at which it pumps them out. We talk about this in our book, uh, safe AI Basics, is when you’re starting to look at your business and look at what is it really about? It’s, it’s really about, again, in Iowa where we are, um, we’re, we’re a heavy commodity business. We’re all relationships. What’s. Separates our businesses from another is our human connectivity. Our risk is that we need to make sure that that unique data is something we’re able to protect and we’re able to ensure is something that remains ours, that doesn’t get ingested in. This is what the law was talking about is, um, right now there really just isn’t a lot of protection out there in general for what I were truly needs, which is an understanding of copyright exists. If I made something that’s mine, if I have a unique data set, just you putting an alteration of that data set does not make it yours. Uh, if I have these, uh, items that are fundamentally mine and core to what I am doing, we shouldn’t just be allowing California to come up with, well, here’s the best way you should do it. ’cause we’re the gonna be the state that’s damaged most our profitability. Our access to business in the state of Iowa comes down to our ability to own the information that comes out of our data. And so data privacy is massive. Knowing that if I put data in somewhere, it is going to be respected the way it is is good. But the next level up there, and we talk about the line governing artificial intelligence, is what are those other pieces of law of items that we aren’t thinking about when it comes to information? Um, what I mean by that is, for example, when I say this is a YouTube problem, um. YouTube in itself when it first came out. When we think of YouTube in the data way, and we think of technology in a package way, when YouTube came out and had a bunch of cat videos and everyone’s like, we can’t do cat videos at work. Well, you know, unless you’re, I don’t know, a veterinarian, but if you’re, you know, if there’s a reason you can’t do cat videos at work, the solution was, okay, we’re gonna block YouTube.

Justin Brady: right.

Chris Draper: Well, now the pandemic comes around and almost all of my government meetings are on

Justin Brady: On YouTube. Yep.

Chris Draper: It’s not the tool that’s the problem, it’s the context and the use of it. That’s an informational sovereignty problem. That’s an informational security problem, and we’re looking at artificial intelligence. It’s not the dataset that is the problem, it’s not the datasets, even training datasets, location of it. I mean, the, the, the training data could still be sitting in Iran for all I care anymore. Right? Because the fact is, it’s how is this tool reusing that information in the context it currently is right now, small businesses, uh, collections. If you were to use AI for your collections work for outreaches to say, Hey, by the way, your bill isn’t paid. And let’s say you’re a southern Iowa firm. So I’m a small southern Iowa, let’s, let’s say I’m a farm implement dealer. I’m in southern Iowa and I got two people, one in Iowa, one in Missouri. Neither of them are paying right now. ’cause there may be some challenges with costs, right? And that state line right there separates what that system’s legally allowed to do. And on one side of state line, I could be do something legal and on the other side, without the right context, without the right training, without the right implementation, or with, maybe it turns out his home is in Limona, but his business is in Missouri. You’re racking up a hundred thousand dollars every single time you have that ai. Make a call without you knowing at a speed that you don’t know what’s making it at these jurisdictional boundaries that are in our real world that don’t translate an AI without thought. That can get corrupted on the way it reads the information, right? If it’s going in and calling someone about a rock in the yard and we don’t realize whether that’s a rock or a dog, I’m gonna either need a shovel or leash.

Justin Brady: Yeah, yeah.

Chris Draper: Which one I choose leads to very significantly different and potentially very un unhelpful outcomes. I guess, unless you’re just trying to get the dog to stop, then shovel will work in both cases, but. advocating, just saying could work, but those little pieces of context we see every day that the synthetic realities, it makes. Randomly assembling pieces hopefully more accurately as we go, but it’s still random. There’s still opportunities to be in appro, inappropriate aligning items, and if we’re allowing the tool to work faster, then we can check it. We’re always gonna be up for, in this case, of completely unmanaged ai, real serious problems. The will attack the heart of what makes Iowa businesses Iowan, which is that human to human. Data based connection, which is the core of all of our true business

Justin Brady: I, I like the example of YouTube and there are probably a lot of different tech examples. Of this as well. But basically you’re warning and saying, look, leaders, your temptation’s going to be to take the iron fist and just say, all right, you know what? All AI is blocked. If you use any i a to AI tools at all. it’s all blocked. You can’t use it too bad. And of course, in those situations, people are just going to use it on their private devices, they’re gonna find workarounds. So it sounds like what you’re advocating is to, um, from the, from the leadership position, from the the CIO, the CTO, whatever position. You should be the one telling everybody in your company you will use AI because it’s really helpful and we are going to listen to you. We’re going to help you understand what tools you need to get your jobs done, and then we will Guide you through. that process to make sure you have the tools and make sure you’re using them correctly.

Chris Draper: I think ai, when I advise boards and CTO level individuals on how to approach this and when we help them with writing what their strategy is and when we help them figure out, uh, what they should do, the first is there needs to be a true acceptance of understanding what we want it to do. And when we look at AI as a tool that is separate from workforce, every strategy will fail. AI must be looked at as a human augmentation component. It must be looked at what it can do for increasing the ability for the human to shift focus. Away from those things that were bogging them down or away from those process p pieces away from those items that were going to, um, require them to do things. That could be in the, in the general sense, we, you know, automation in general. This gives us another layer of how do we actually help rethink where time is spent. When you look at I, I believe this was a McKinsey study, but may have been Gartner, I apologize. To which, which, which large group of dataset or not. But when you’re looking at productivity gains through efficiency improvement, you have massive leakage just in the general way of business work. So if I’m looking at a human as a cog and I have the same series process of this human did this, it creates this, which goes the next human that does this, which goes to this cog and goes to this, and if I look at that. Process there. And then if I keep everything in that linear, you know, linear series way that I do things and all I do is speed up a component of it, it really doesn’t change my process fundamentally. And I will have a significant leak in productivity where I’ll only see for a 25% increase on my ability to speed something up. I’ll really only see a 1% productivity gain that’s gonna lead to profits. Where I have my greatest opportunity is in the shifting of the focus of the employee. If the employee is still thinking about the same process and perfecting that process versus saying, why do we do this process at all? Um, I won’t see the value I could if I’m able to shift it towards. I can have less handovers by giving this same person expanded skill sets in certain areas. I could have more intakes going at the same time with the same level of human touch that I would’ve had before, which is really where major value is gained. A lot of our companies, humans still wanna talk to humans. Um, if, if, if you, if you ever wanna see the danger of ai, call Verizon customer service. Good lord. I mean, like the, the only thing I’m. The only thing I’m really happy about is if I go into a store, they don’t get anything faster either. Yeah. They’re stuck on the phone 12 hours too. Awesome. I love any company that treats their employees as bad as their customers. That is great. Right. But if you sit in that environment and say, boy, I really just want a human, AI can give your human super superhuman access to data, it can give you superhuman access to. I have a family member that will remain nameless right now, but they might not always be the most empathetic. Mm-hmm. And so if you have a problem and you want to have some type of emotional response from ’em, you just don’t approach them. Now plenty of customer service people have had a long grinding day. They’re getting just beat down, and the fact is there’s going to be little sentiments within a caller’s voice that they may not even pick up on as an opportunity to actually jump out and be like, oh, if I just did this instead, this person sounds like they might be open to

Justin Brady: Mm-hmm.

Chris Draper: Mm-hmm. That’s a perfect use for ai. That’s a perfect well contained. Human directed amplification of a skillset that human has. If I’m now using AI to an analyze a sentiment of this caller and hear something in there that I could probably test with them, that I could explore, maybe this is a solution for you. Maybe the solution is something that I just am not in a mood to hear right now, and that is something that’s able to flash up and get that customer service rep out of their, I just gotta grind through this mode. You’ve now provided everything we want, which is human to human interaction in a way that is now better, more informed, that is more useful, that is quicker to resolution by letting that tool be an amplifier. Those use cases are incredibly powerful, but we have to be thoughtful where we apply them. Mm.

Justin Brady: So manage, manage ai. I mean, honestly, I, your entire IT structure should be managed this way too, but manage AI around the individuals and the projects they’re trying to attempt rather than a wide brush and just saying, this is banned, or we can or cannot use this. It, CIOs and CTOs should be working with the other leadership team to say, what are the big projects you’re trying to accomplish? And let’s be your assistant. Let’s help you work together on accomplishing those rather than always being the oppositional gatekeepers of everything. ’cause if they’re the gatekeepers, employees are gonna find workarounds and they’re still gonna bring risk, rinse risk inside the organization. They’re gonna still corrupt your data. It’s still gonna be the same problem. So you have no other choice but to work with people. Uh, I guess final question would be. And I, I think I’m. hearing this right, but in most Iowa organizations, organizations in general, it sounds like it’s not really a matter of, uh, tech savvy versus non-tech savvy being your risk when it comes to working with AI in the workplace. It kind of sounds like everyone’s pretty much at risk unless you know what you’re doing.

Chris Draper: And I think that everyone is at risk, but they’re also open to the same opportunity. The the broader challenge is that, again, this isn’t traditional. This isn’t traditional tech. When you’re looking at, for example, our insurance companies, you have a data component, an InfoSec component. You have a tech component, but AI at its core is a human resource issue. This is a human resources technology. If we’re going to be most productive with it, we need to understand that this tool is a human amplifier and accelerant and the type of employees and the type of skills they need. We must be planning for more than the tech itself. In the same way that, and I don’t agree with this fully, I believe that there’s a lot of hype about how we aren’t getting any programmers. ’cause I, I was just,

Justin Brady: yeah, I don’t believe that. crap either.

Chris Draper: Here’s the reason, and the, the, the reason is the same as you saw with the, the math race back in the sixties with the US and the Russians. And the fact was that the US we had higher computational power, so we came out with better computations to solve the problem, whereas the Russians went to closed form solutions. Now, if I went, they went to. Did the hard work of getting to the core item and getting the core thing right. When I’m looking at software programming, um, can these 8,000 lines of code do what we wanna do? Sure. But about 9,000 of ’em go in a bunch of circles. I don’t need, they provide me risk vectors I don’t want, I’ve got a whole bunch of other challenges in that slop and a great coder knows, oh, it’s lined this to this, to this is going to solve that problem correctly safer with less issue. Similarly in business. Businesses are essentially a piece of paper that allocates wealth and the operation of it is always going to be coming down to the people. The wealth that’s gonna be created and distributed is all based on the people you have there. The technology must all be fit to the human and ai. Shows that truth more clearly. Because it is so malleable, it is not something that can be so structured that humans have to conform to it. Humans can constantly break it. They can constantly bend it. They can constantly bend the data sets we’re working with. They can use the tool. More dynamically than any other piece of technology we have, which is what’s forcing us if we open our eyes to it and as we work with boards and say, okay, where’s the real issue? And who should be responsible for this? It really should be, your implementation will be tech. That is correct. Your, your pre-processing of it will be data and it must be housed within an info security position, but it really should fall under hr if HR is approaching it. The way it needs to be. This comes down to another sort of bug of mine though, is I think that one of our challenges in Iowa, we, we talked about this very early on, like what’s the uniquely distinctly Iowan problem we have? Mm-hmm. Our distinctly Iowan problem, which we have in the startup space and the company space, and a lot of this drives down to, to how we view human resources, how we view our own understanding of risk. I joked joke about this often we say that Iowans a risk adverse, yet almost all of our economy is betting on the weather. That is bananas. We’re betting on the riskiest thing going, bro. You know, I mean, I, I think we lost Enron over the weather. Right. And we’re telling us that we aren’t a risk acceptance. We accept a lot of risk, but we like our risk in boxes. Mm-hmm. And we like our people in boxes. We like our company in boxes. We want to find the thing and say, oh, I need three to five years of this, and I need this degree. I need X, Y, and z. Often we are looking at the HR solution as what will keep me from getting sued and what looks like everyone else. Well, AI is going to allow those little added value elements within a human to be amplified and accelerated in ways where that box is wrong.

Justin Brady: True.

Chris Draper: Our boxes here, our boxes here, were always

Justin Brady: Yeah,

Chris Draper: but the fact is now. I will not ever be able to succeed in the tech department with AI as long as human resources is forcing the same boxes.

Justin Brady: So you, we need to be, I guess, to put a period on the this, we need to be very open-minded about how much AI will grow the potential of our employees. And what you’re saying is the bigger risk here. Or One. of the big risks here will be overregulating, this internally and, and being too rigid about it. We need to be, uh, proactive in working with our employees to get this right, um, because it will really empower them significantly beyond. I like how you mentioned that we kind of put employees in these little boxes, like this is what your capabilities are, this is what you can do, and very soon. That box isn’t gonna exist. And the big risk here is that we’ll be limiting our employees to such a degree by not allowing them this augmentation technology. Um, Chris Draper, the CEO of Morgan ai, um, what’s the website? And give us like kind of the elevator pitch of what your company’s doing for Iowans.

Chris Draper: Yeah, so morriganai.com. Uh, that’s, uh, M-O-R-R-I-G-A-N-A i.com. It’s not morgan.ai. That’s, that’s a Polish group, which isn’t us. We’re, we’re still working on that, however, so what we do fundamentally is we help businesses from all the way to small to midsize is where bread and butter is. Truly understand how their business is operating with and without ai. So we have a, right now a beta program called Crow, which is out. What Crow tells you is every time information flows in or out of your system, was it, uh, human initiated or automated initiated? And was it using or was it not using ai? Our Raven product, which is coming out in Alpha, which we actually got, uh, uh, budget from POCR, from IADA to actually do some, uh, noco. Oh, thank you. Well, that’s, that’s gonna be doing some no cost test cases for Iowa businesses. So if you have any listeners that wanna work with us on Crow and then our Raven Tech, which is coming out, um. Raven gives you essentially an operational risk model of your business that lets you understand how you’re truly operating with and without agents. So right now, a lot of your business consultants come in and either do interviews to say, Hey, who’s doing what in this business? Or they look at your CRM to say, Hey, your CRM says this is how your processes are. Well, there’s a massive industry to fix CRMs, so that’s probably not good data. And every time you interview people, you, someone says, well, we don’t really like Betty. So you come in and you fire Betty. It turns out Betty’s the only one doing ar, and you just erased all of your revenue. What Raven does is takes a hyper cynical approach that let’s assume we don’t know anything. It goes and allows an agent called Morgan Raven to actually take a look at the artifacts of your business. It analyzes how information coalesces and it can now tell you these are the most influential processes, influential individuals or teams. This is where your bottlenecks are to growth, or your bottlenecks are to downside risk. So it, gives you

Justin Brady: people to not, it basically allows people to get outside that box and be productive employees, but also not bring the risk and destruction to your organization. Uh, Chris Draper. the CEO of Morgan ai, thank you so much for coming on the Iowa Business Podcast. Really appreciate you being here.

Chris Draper: Thank you for the time. Appreciate Justin absolutely.