E058 - AI or Not - LeAnn Oliver and Pamela Isom

Welcome to "AI or Not," the podcast where we explore the intersection of digital transformation and real-world wisdom, hosted by the accomplished Pamela Isom. With over 25 years of experience guiding leaders in corporate, public, and private sectors, Pamela, the CEO and Founder of IsAdvice & Consulting LLC, is a veteran in successfully navigating the complex realms of artificial intelligence, innovation, cyber issues, governance, data management, and ethical decision-making.

AI can save hours of work, but it can also scale mistakes faster than any team can react. That’s why we sit down with independent consultant and former DOE leader LeAnn Oliver to get practical about AI governance, not as paperwork, but as risk management for real systems that touch real people.

We start with LeAnn’s path from public finance into enterprise technology leadership, including automation work that created massive efficiency gains and also introduced hard lessons about change management, data privacy, and handling PII. From there, we dig into what comprehensive AI governance actually means: aligning enterprise architecture, cybersecurity, and policy so autonomous AI agents know what they can do, what they must never do, and when they have to hand control back to a human. We also talk about why quantum and ever-faster compute raise the stakes by accelerating both good outcomes and bad ones.

Then we pressure-test the ideas with real examples: an AI tool that wiped data during development, AI-generated “recommended books” that didn’t exist, and why high-stakes domains like medical interventions and critical infrastructure should have clear human decision points. We also explore the uncomfortable truth behind the race to AI: lots of spending, uneven ROI, and growing legal liability, which is exactly why boards and CEOs need to treat AI governance as strategic oversight, not an afterthought.

If you lead a business, run technology, or advise executives, listen through and share this conversation with someone who’s pushing AI into production. 

Transcript

[00:00] Pamela Isom: This podcast is for informational purposes only.

[00:26] Personal views and opinions expressed by our podcast guests are their own and not legal advice,

[00:34] neither health, tax, nor professional nor official statements by their organizations.

[00:42] Guest views may not be those of the host.

[00:50] Welcome to AI or not, the podcast where business leaders from around the globe share wisdom and insights that are needed right now to address issues and guide success in your artificial intelligence and your digital transformation journey.

[01:05] I am Pamela Isom and I am the podcast host and we have this really cool guest with us today.

[01:12] Her name is LeAnn Oliver. LeAnn is an independent consultant and former DOE employee, so we work together,

[01:20] the Department of Energy.

[01:21] She has some strong opinions and perspectives when it comes to governance and business insights and wisdom to share.

[01:30] LeAnn, thanks for joining me today and welcome to AI Or Not.

[01:35] LeAnn Oliver: Well, thank you very much for having me. It's great to talk to you again.

[01:39] Pamela Isom: I'm excited to hear what you have to say.

[01:42] So let's start with you talking about yourself, your career journey,

[01:46] and I'm curious about your experiences in and outside of government.

[01:53] LeAnn Oliver: Well,

[01:55] I can honestly say that when I was starting out, it never would have occurred to me that I would be involved in a podcast like this. This on this subject,

[02:05] I grew up in a really small town in far Northern California,

[02:10] and for the vast majority of my career it was tangential and crucial,

[02:17] but it was tangential to what else I was doing.

[02:21] I was a political science major undergrad, and I did take one class, a computer class. I learned how to program in Fortran,

[02:30] so I'm kind of dating myself there. Then I didn't really have another encounter with it until I went to grad school, got a master's degree in public finance, and was working in the Small business administration in D.C.

[02:43] for a loan program.

[02:46] But my boss found out that one of his colleagues was getting a wide area network and he decided that he wanted one even though he didn't really know what it was.

[02:56] And I got that as an other duty as assigned.

[02:59] So so I set that up and then over the years I worked my way up through the ranks at sba. I ended up working a lot with IT to support what we were doing.

[03:13] Originally, all the loans were processed in 50 local district offices and we had a budget crunch that resulted in not having enough people. So the decision was made to automate and I was instrumental in that automation.

[03:29] This involved getting loan applications from thousands of banks through a portal,

[03:36] which was challenging because some of the banks were very IT capable and others were not.

[03:44] But it was my first encounter with the change management issues that go along with automating things.

[03:52] And so I learned that lesson. It wasn't very pretty, but we did manage to do that automation.

[03:59] As a result of that, we ended up with a big database that included PII and business sensitive information.

[04:08] And so I had to learn how to handle those databases.

[04:12] And I had another job that also involved a large database of entrepreneurial training.

[04:19] For another part of the Small Business Administration.

[04:23] I went to the Department of Agriculture and that had a very non.

[04:29] There just wasn't a lot of it there. So I didn't really have anything to do with it there. But when I came to doe,

[04:36] I moved to DOE to run the grant programs that were part of the Recovery Act.

[04:41] And I ended up working on that for a while.

[04:45] I then ended up in the CIO's office where I was the deputy for Policy and governance.

[04:52] But my last job was the Director of Corporate Business Systems in the CFO's office where I was responsible for the 32 systems that did all of the basic admin functions.

[05:06] HR, accounting, procurement,

[05:09] all of the stuff that basically makes the entire department run. I used to tell my staff, look, all of the really cool scientific stuff that gets done in this agency wouldn't happen if you guys didn't do your job every day,

[05:23] day in and day out, to make sure all those admin functions functioned because people tend not to think about those. But if they don't work, it becomes very apparent very fast in that role.

[05:36] We were starting to use AI and I was trying to figure out how to benefit. We made the accountants very happy because we did a system where we call.

[05:48] We automated the closeout of a bunch of accounts.

[05:53] It used to take them two hours to do one account. And we got it down to six minutes. And for five of those minutes they could be working on something else.

[06:02] They had to make the decision to start that process, but once they did, they could move on to something else. So it saved a lot of time. And it was a very good example of how AI.

[06:13] And that was kind of lower level AI, but it was still AI that could work.

[06:18] I've been retired for about a year now,

[06:21] although I have been looking at various things to do and governance. I always liked governance, I think maybe because I was a political science major undergrad. And when I went to the CIO's office, the first assignment that I had was to do a new governance system.

[06:38] And after about three weeks I walked into my boss's office and I said, you know, this is not an IT problem.

[06:45] This is a Poli Sci 101 problem. We want a bunch of people to do something that they don't really want to do.

[06:50] And so that kind of,

[06:53] that was a significant part of my thought process about governance. Because if you do not have good governance, you run the risk of having all kinds of other issues that you do not want to have.

[07:07] It gave me a broad appreciation of what AI can do as well as the overall perspective, the broader perspective, because as I said, most of my career was not in,

[07:21] was in other things.

[07:23] And I can understand the uses of AI within that context,

[07:29] but I also understand what happens if you don't have good governance because you're going to end up in a bad place.

[07:37] Pamela Isom: That's an interesting background. And I personally know you from those acts. Right. So the, particularly the research act that you were involved with. And so that's how we met. And that was interesting.

[07:49] Had a great relationship and great experiences together.

[07:52] So I want to move on, but I do remember those times and I appreciate the opportunity that we had to work together and also the opportunity to continue to work with you as life goes on.

[08:03] So I want to ask you, because that is your passion, is governance, I want to ask you about AI governance.

[08:11] Give me some perspectives on what comprehensive AI governance looks like to you.

[08:18] LeAnn Oliver: Sure.

[08:18] AI governance is pretty broad.

[08:22] I think it includes the architecture of your systems,

[08:29] cybersecurity,

[08:30] because if your cybersecurity,

[08:33] if you're granting developer authority to an AI and you don't tell it what it can and cannot do, you could possibly run the risk of it having a security breach where they get asked for data by something or someone and they give it to them without,

[08:52] I mean that, you know, you have that risk.

[08:55] I think when you combine that. And this is more in the future, but Quantum is coming.

[09:01] And what that means is that whatever the AI does, it's going to be able to do it even a lot faster than it can do it now.

[09:08] And so if you have bad governance or no governance,

[09:12] you're going to end up with a bad outcome.

[09:16] It's going to really replicate really fast.

[09:20] And I think you also have to have overall policy because maybe you allow the AI some flexibility,

[09:30] but you need to define it so that it doesn't have the flexibility to do just anything that it feels like doing or thinks it should do or maybe even doesn't think it should do.

[09:42] It just does it without thinking. And that, I think is one of the reasons that you need to have good governance, because I think that's a Possibility.

[09:50] Pamela Isom: You think that that's difficult? I think it seems to be difficult for us to figure out, like, where those decision points should be and how to go about developing and integrating policy to accommodate the autonomous behavior of AI.

[10:13] What do you think?

[10:15] LeAnn Oliver: I totally agree with that.

[10:17] I think it's very hard.

[10:19] It's kind of hard to even do it in a more limited way, which is why you have some of the failures that you have. In the broader way, it's even more problematic because there are a lot.

[10:30] You can't anticipate every single path that the AI might follow.

[10:34] And so you need.

[10:37] There's a push pull because you don't want to limit it too much because it can do a lot of things for you that are good in the sense that it's not.

[10:48] You're not spending your time doing kind of things that have to be done but aren't particularly fun to do and aren't.

[10:56] You have to balance that, which is why I think you have to think through at a bigger picture level what kind of constraints you want to put on it so that it doesn't kill someone or it doesn't, you know, make some mistake that is going to be really terrible.

[11:14] I mean, I read not too long ago about this.

[11:17] Somebody asked. It was in a medical network, and they asked the AI to identify all of the devices that were all the medical devices, like laser surgical instruments and that kind of thing.

[11:33] Fortunately, they did not.

[11:35] Those devices weren't told to do anything. It just was asked to identify them.

[11:40] But it did identify them, and some of them were being used in active surgery at the exact moment that that identification happened. So if the AI had decided to do something that it shouldn't have,

[11:53] it could have been very, very bad for the people who were getting eye surgery or whatever else was going on.

[12:01] And so I think you have to really think through what it is that you're trying to do and what the capability of the AI is.

[12:11] You need to be smart in terms of cordoning off the things that you don't want it doing that you definitely don't want it doing,

[12:20] as well as allowing it the flexibility to figure out better ways of doing something that isn't going to kill somebody.

[12:28] Pamela Isom: Okay, so comprehensive AI governance to you is looking at the layers of the infrastructure. So it's scaling the infrastructure. I heard you mention the architecture,

[12:41] you mentioned cybersecurity,

[12:43] and I could just hear in what you were saying that it needs to be examined from the lens of all of the stacks and potentially the entire AI life cycle.

[12:54] Did I hear you right?

[12:56] LeAnn Oliver: You did.

[12:57] I think that you absolutely have to do that.

[13:00] And that's the kind of thing that's a real challenge because it's kind of tedious,

[13:05] depending on what the area is.

[13:07] And you have to think about it carefully. You can't be sloppy about it because that's how you're going to end up with problems later on.

[13:15] And so it's a tedious process. It's not the fun stuff that you have by having AI take away a half an hour of work for you and saving money, which is what a lot of people want to use AI for, to do things that save money, which is, you know,

[13:34] I mean, I'm big on saving money, but the governance is part of, you know, making sure that your governance is appropriate is really a major thing that you need to look at, even if it is time consuming and kind of tough.

[13:51] Pamela Isom: So when we talk about those key indicators.

[13:55] So like key indicators that humans should be the decision makers in the day of agents and humanoids and robotics at large.

[14:07] I was saying that I think it's difficult for us to figure all this out, but it's something that must be done. So I have two points that I want to make.

[14:16] One,

[14:17] I was on a call today. I was teaching a class today, right? Because I teach.

[14:21] And one of the things that I pointed out, and I'm curious as to your perspective,

[14:26] but when we worry about AI taking jobs or taking jobs away from us,

[14:33] I inform my class that, listen,

[14:36] governance is not going to be replaceable by AI. So if you can build up your skills in this domain,

[14:44] no matter how much it augments the work that we're doing,

[14:49] this is an area that needs to be sharpened and sharpened by humans.

[14:55] What's your take on that?

[14:57] Before I make my second point,

[14:59] I

[14:59] LeAnn Oliver: totally agree with that. Because basically any of your students should be focused on things that the AI can't do. Because if there's something that needs to be done and the AI cannot do it, and there are things that it cannot do or you don't want it doing, I mean,

[15:15] then those are the places that you want to work. So you're absolutely right about that.

[15:20] Pamela Isom: Okay, so that's what I told them. So I just want to know, because I might have to go back and say, wait, wait, wait, wait, wait. Let me clarify.

[15:27] LeAnn Oliver: No, Pam, you gave them good advice, like I would have expected you to, knowing you.

[15:31] Pamela Isom: So,

[15:32] and then the second part is when we're thinking about those areas and you are starting to talk to it earlier,

[15:43] where Decisions can't be undone. For instance,

[15:47] like if it's about a medical situation, a medical intervention that would need to occur by a human, if it's a critical infrastructure action like the grid should be shut down or I deal a lot with micro grids.

[16:05] So the micro grid islanding and it's under uncertainty.

[16:10] So that's not where I would want an AI to decide for me. Right. So the medical interventions,

[16:18] the national security,

[16:19] although we're looking at integrating AI more in our security efforts.

[16:26] Can you help me understand,

[16:28] like one of the communications that I have is that this is, this is complicated as we start to talk about,

[16:34] but these are opportunities for us to just take the time to start thinking through these types of things.

[16:41] And also like other areas as I don't want to say too much here because it's you talking.

[16:47] So, but what do you think about that and where should we be as far as society and as far as looking into and paying more attention to humanoids and agents and these decision making layers?

[17:05] LeAnn Oliver: I think the way to approach that is to think about what you want the AI to do and then what are consequence. I mean, one way of looking at it is to say, okay, we want the AI to do X and then to think about all the consequences that could happen with X.

[17:26] And that would then narrow down because if none of them are fatal or potentially fatal or have any really serious ramifications,

[17:36] then you can say to the AI, go for it, do whatever.

[17:40] But if it has implications that are significant in any way, whether it's human life, whether it's the environment,

[17:51] I mean, one of the things that AI,

[17:55] I have a background in economic development too. And so I can see that when you have data centers and you're putting data centers and you're going to have some issues because you're going to need the data centers to handle all the data and do all the AI stuff.

[18:10] I mean, there's no question about that.

[18:12] So that's another factor here.

[18:14] So you have to look at that whole broad picture,

[18:18] but for the specific AI target, if you will,

[18:23] you have to look at it and say, okay, can this kill somebody? Is this going to poison somebody? Is this going to do whatever and then make a decision about how much, you know, what kind of strain you want to put on the AI.

[18:38] Pamela Isom: It makes me think that there needs to be,

[18:42] as a part of the governance, there should be some deep education or heightened education or literacy on trade offs. Right?

[18:53] LeAnn Oliver: Absolutely.

[18:55] Pamela Isom: Yeah. And that's what I'm kind of getting to, because that's Kind of what's like in my brain is we gotta get there because I think people are stuck, right? And when we are, when we make it past it, the incident has already transpired.

[19:09] Mm.

[19:10] Now,

[19:11] speaking of that, you mentioned to me earlier about an incident that we actually talked about several incidents, but there was one where some data was wiped out. Tell me more about that and how you think AI governance can, can assist with, with preventing this.

[19:27] LeAnn Oliver: This was an article I just read about this CEO of a tech firm and he was using AI to work on a system and the AI that he was using wiped out his data.

[19:42] I would think that part of the guardrail there should have been do not wipe out anybody's data. Hey, AI do not wipe out data.

[19:53] You can sequester it if you want, put it, you know, park it somewhere where it's not going to mess with anything else you're trying to do,

[20:01] but do not ever wipe out data.

[20:03] And that obviously wasn't part of the AI.

[20:07] The guy who developed the AI apologized for it and the guy who,

[20:14] the CEO basically figured out how to,

[20:19] how to recreate his data.

[20:21] So it wasn't the end of the world. And it, you know, and he was working on it as a, you know, he was, it was in development.

[20:27] So it wasn't a system that was being used in real, in the real world. Real world on a production basis.

[20:35] So I think that those are the basic guardrail that you give to every AI and you can have the, you can tell the AI to go back to a human, it's human developer, and ask if he, he or she wants it wiped out.

[20:52] And then that's okay. If the human says yes, go ahead and wipe it out.

[20:57] But I think, you know, those, that's a basic kind of thing that when I read that, I was thinking I would be really upset if that happened to me.

[21:06] And I think that that should have been a basic. Don't do this. This should be a basic rule in the guardrails.

[21:13] Pamela Isom: And then there was another one where we were discussing reading material and some books were recommended that didn't exist or something like that. What was that all about?

[21:24] LeAnn Oliver: The Chicago Sun Times and I just saw this. The book editor,

[21:28] I think it was the book editor, but somebody at the Chicago Sun Times asked AI to come up with a list of recommended summer reading for people who are going to the beach or whatever.

[21:40] And the thing came up with a list of 15 or 20 books. But then it turned out that several of these books didn't exist and they.

[21:50] There was one that was supposedly written by Elizabeth Allende, who's this very well known author who writes great books. And this book didn't exist.

[22:01] So not the end of the world. Nobody died as a result of this. But you really don't want to have to question everything that you run across,

[22:11] or at least I don't want to question everything that I run every piece of information that I run across.

[22:17] And if AI is generating incorrect information,

[22:21] that is not a good thing.

[22:23] Pamela Isom: So that seems like another skill that we need is to be able to quickly detect whether it is legitimate or not. Right.

[22:32] LeAnn Oliver: Well,

[22:33] detecting whether it's legit is one thing, but AI shouldn't be generating stuff that's not legit, I don't think.

[22:42] I mean,

[22:44] it's like, what is the point of it if it's going to lie to you? Basically,

[22:49] it needs to not do that in the first place.

[22:52] I would need, I would like to be able to trust what it tells me. If I go in and say, find me a list of whatever.

[23:00] Pamela Isom: So, yeah,

[23:01] as I was getting to, these are hard times,

[23:05] these are tough times.

[23:07] And we have a push, in fact, a race to AI.

[23:11] What is your take on that?

[23:15] LeAnn Oliver: Well, there's a huge push. AI is the hot thing right now.

[23:21] I mean, it is every time you turn around, every place you go, everything you see is AI this, AI that.

[23:30] But it hasn't really delivered to the extent that it probably could or will in the future. I mean, it's, you know, it's like any new thing. I mean, remember your first PC?

[23:46] It was pretty basic. It didn't really do a lot.

[23:49] It was, you know, your first cell phone.

[23:52] Going back a little bit further, the Model T cars, which fortunately neither you or I are old enough to remember in person, but we,

[24:03] you know, so there's a.

[24:05] All technology has kind of a growth pattern and AI is like that in its infancy. And there's a lot of cool things that it should be able to do, and I don't doubt that it will,

[24:18] but, you know, it needs to be a little more mature.

[24:22] I participate with this research group and every single time we have a discussion about AI,

[24:32] the guy who runs this asked the question about AI paying for itself. And I have yet to hear where anybody says,

[24:42] yes, it's making a huge difference. It's making a difference at the margins because there are certain things that it does well that are good things to have done,

[24:50] but it's not nobody on these calls. And these are pretty senior people across large tech companies and they don't they're all going, you know, they can see the potential. They know it's going to be there.

[25:04] They.

[25:05] I don't even know what to call it, but it's the sort of the limbing approach where everybody is following it because.

[25:15] And I think that it's not a bad approach because I do think that there is going to be something there at some point in the future, even though we're not there yet.

[25:28] And so these people are all afraid of being left behind, which is a legitimate concern and kind of an unsettled situation because there's not a huge payback on it and it costs a huge amount of money.

[25:44] You know, you need servers and you need storage and you need develop, you know,

[25:51] you need governance on the development.

[25:55] And so you need to,

[25:56] you know,

[25:59] it is a changing situation and I'm sure it will sort itself out going forward. But I hope that people are very careful about the governance because some of the things that it could do have the potential to be very dangerous.

[26:13] And I don't really think that the Terminator is going to show up on our doorstep, but it's, you know, possible, I guess.

[26:22] So you think that, don't want that.

[26:25] Pamela Isom: So you think that the push is real and you think that there is in fact a race to AI and you think we can survive it if we have good governance in place.

[26:36] Did I hear that right?

[26:37] LeAnn Oliver: Yes. That's a very nice summary. Thank you.

[26:40] Pamela Isom: So what's your take, that being said on the Genesis mission?

[26:45] LeAnn Oliver: Well, that is,

[26:46] I think, a very interesting approach and a very interesting, I mean, you know,

[26:53] from having worked there, DOE is the science agency in the government, basically.

[27:00] And so it is a good thing for DOE to be in charge of that.

[27:07] And I think there's a lot of money being spent on it, which probably is necessary because DOE is in a position to cover those kinds of things that are not viable in the private sector because those costs, those basic costs,

[27:28] are not something that the private sector can do on their own. And that's always been what doe's role has been. I mean, doe, if you recall all this research that they did, once it got to the point where it was potentially commercially viable,

[27:46] that's when DOE moved on to something else. And so Genesis is kind of in that same,

[27:51] the same history there. And so that's good. I mean, it's. But once again, they need to be careful about governance.

[27:58] Pamela Isom: What would you want to see as the agency is progressing with this effort? Is there anything from, from a governance perspective? What? Give me an example. Of what you mean by that?

[28:14] LeAnn Oliver: Well, this gets back to what is a specific genesis is going to get broken up into a bunch of different projects that do different things.

[28:22] There's a huge number of things that AI can do. Well,

[28:27] you know, it can, it can do.

[28:29] I mean, one of the things that I found fascinating or I find fascinating about it is that it can do chemical calculations that it can run through in a day.

[28:41] Calculations that would have taken people using calculators or whatever they were doing, spreadsheets, whatever they were using weeks or months to do. And it can.

[28:53] So if it's just limited, I mean,

[28:58] using that as an example, if it's just limited to hypotheticals and it's not doing anything where it's actually creating a particular chemical,

[29:07] that's fine. I mean, that's pretty safe.

[29:10] But when you get to the point where it is driving the manufacturing of something that has the potential to be dangerous,

[29:20] then your level of governance needs to step up a little bit to an appropriate level. I think maybe the way to think about it is looking at what you're doing with it and then as I said earlier,

[29:34] figuring out what level of governance is appropriate. What guardrails do you want to put on that because of the risk? It's a risk assessment, basically.

[29:45] It is a total risk assessment. And if something isn't that risky because it's hypothetical,

[29:51] go for it. If it is something that has real world consequences or could have real world consequences,

[29:59] then you need to be a little more careful and you need to be appropriately careful.

[30:04] Pamela Isom: I love this conversation because really it's centering upon governance, true enough, AI governance, but also paying a close,

[30:12] close attention to understanding the consequences.

[30:17] And I would add testing to testing for those consequences. Testing the consequences. So we were talking about edge cases.

[30:27] And so, so the example that I like to use is take Hawaii for instance.

[30:35] Hawaii is a sunny state. You get rain,

[30:38] you probably aren't going to get any snow.

[30:40] So. But you'll have. But you know, we're likely to have electric vehicles there and autonomous vehicles there and just vehicles, period. Right.

[30:51] So when we test the resiliency of the vehicles,

[30:58] are we only testing in rainy conditions,

[31:03] wet conditions,

[31:06] where they're testing to see how quick the vehicles will rust and that's it, Right? You know what I mean?

[31:13] How about testing to make sure they'll withstand snow?

[31:17] But one would say, well, why would we do that? Because it doesn't snow in Hawaii, because this is the day and time that we live.

[31:24] There are so many things that are unexpected that are happening.

[31:29] There are so many consequences of if it does not perform,

[31:35] if that happens, until I think we need to think and plan for those edge cases.

[31:41] And so I tell my,

[31:43] my colleagues and my students this, like,

[31:46] because when,

[31:47] when it comes to AI,

[31:49] you don't know. So you have to think about these edge cases. And a lot of times what we test for and what we bank on is what we think will happen.

[32:01] LeAnn Oliver: Right, that's, that's why you need a wide perspective to think.

[32:07] You need to put whatever it is that you're looking at in the bigger holistic perspective and say, okay,

[32:15] what is the risk now? Maybe if you're never going to sell the car outside of Hawaii, you don't really care if it rusts. But if you think that that car could then get sold in other parts of the world,

[32:30] then you've got to do the broader test. Now that's not to say you can't say, okay, I want to, I'm going to sell these cars mostly in Hawaii initially.

[32:39] So fine, we'll test for Hawaiian,

[32:41] Hawaiian weather. But then the next phase may be, I mean, you don't have to do this all at once if you don't have a reason to do it all at once.

[32:50] But that doesn't mean that you get a pass on it. Because if you then say, oh, well, these worked really well there, so let's sell them in Alaska,

[32:58] well, yeah, that might be an issue.

[33:01] Want to test them for Alaskan circumstances before you,

[33:06] before you expand. I mean, see, and that gets back to looking at it holistically, because that's very logical. If you look at it holistically and you go, okay, I'm just going to do it here and under these circumstances and then if I want to expand,

[33:21] then I have to think about the other circumstances. But that's okay. I mean, there's nothing wrong with doing it that way.

[33:28] You just need to like not rush them to market and have them fail, which would be a bad thing, which

[33:34] Pamela Isom: is happening with AI.

[33:36] Yes, so that was the point. Right. So I have another question for you. So I've heard this statement and I just need to know from you, is this a myth?

[33:46] Do you think it's reality?

[33:48] What should we be doing about it?

[33:50] But the statement that I've heard more than once is anytime you're talking about governance, and especially AI governance, it doesn't pay.

[33:59] So what's your take on that?

[34:01] LeAnn Oliver: Well, I think the person who said that is being very short sighted.

[34:06] Because if you don't have governance, the downside risk and once again, this gets back to what I've been saying all along, which is if you are doing something that isn't particularly risky, that doesn't have any obvious negative consequences.

[34:22] Okay.

[34:23] I mean, and you do, you know, I mean, AI is kind of quirky.

[34:28] Some of the things that have happened in AI are things that nobody expected.

[34:33] And so you do have to kind of go, well, what happens? What could happen here?

[34:39] And I can understand the trade off because it is expensive. Governance is risk management.

[34:45] You have to have people who understand what the risk is. You have to take some time to think about it. And I think maybe part of that statement is coming from people who really want to move forward really fast, make the move fast and break things.

[35:01] People which in some environments is fine.

[35:05] I mean, that's fine. But if you are messing with something that has real world consequences on real people,

[35:14] then it's a different standard.

[35:17] I mean, even if you're going to say, well this is something that is low risk and we're not going to worry about it,

[35:23] fine. But you need to think about that before you flip the switch and have it go.

[35:29] And then you need to keep monitoring because if something odd happens, you need to stop it because if it replicates and I mean, I don't really want to sound like I'm writing a script for some sci fi horror movie,

[35:44] but you really want to pay attention and you don't really want the sci fi horror script.

[35:53] Pamela Isom: I say we can't afford not to. So we can't afford to not institute good governance and continuously evolve as society is evolving. Right. So we just. Absolutely.

[36:07] LeAnn Oliver: That is your ability to just hone in on the truth is really, really good. Pam.

[36:14] Pamela Isom: I agree with you though. All right, so I have enjoyed this conversation.

[36:20] I am appreciating your emphasis on the consequences.

[36:26] I understanding the consequences, like I said, like that ought to be a class or that's, that's just always zero in on or something that the board, the boardroom workshop. Right.

[36:36] So we're. That's what they're dealing with for a day. Let's have a work, let's have an off site and let's do understand the consequences and mitigate the risk. But that's just my take at this point of the talk.

[36:47] I would like to know, is there anything else that you want to share? So there. A lot of times my guests will have thought of something by the time we get to this place.

[36:56] So if there's anything else that you want to share with us, let me know that and then Finally, I'd like to know if you have any words of wisdom or call to action or all of the above.

[37:09] LeAnn Oliver: Well, I'm going to say all of the above because,

[37:14] you know, I appreciate you having me on here.

[37:18] It's been a real pleasure to talk to you and to talk about this because I do believe,

[37:25] given my background,

[37:27] that governance is really important and broad governance. It's not just specific governance. Although if you don't do the specific governance, right,

[37:37] you're going to have a broader problem,

[37:39] possibly.

[37:40] And they're going to be problems that you really don't want to have,

[37:45] like wiping out data. Wiping out data,

[37:48] wiping out processes,

[37:50] wiping out people.

[37:56] I mean,

[37:57] yeah, I mean, that's maybe a little much, but as we were talking earlier, people have been killed,

[38:03] so we don't really want to replicate that.

[38:07] And as a, you know, and in order to avoid that, I think my bottom line is you really need good governance, you need smart governance. I mean, you don't want to unnecessarily constrain the AI process,

[38:23] but at the same time, you need to have something that is going to be reasonable.

[38:30] And yeah, it's probably gonna be expensive because if you don't do it, you aren't spending any money on it until you have to pay for the consequences that you maybe should have seen.

[38:41] I mean, we haven't talked about legal liability,

[38:44] but I think that that's an issue that will have some ripple effects through society. I mean, you know, just this week I think,

[38:54] I know it was meta and somebody else, they. They lost a lawsuit.

[38:58] And I think that that's probably kind of a precursor for some other stuff that's going to happen here. Just because there is somebody should be responsible for this.

[39:09] Yeah, I mean, I think it has to be the companies that are, or the people and the companies who are responsible for the AI. And so if your AI runs amok,

[39:21] I mean,

[39:22] gonna get messy and people are gonna come after you in a lawsuit, in court.

[39:29] And you need to think about that. When you think about your risk,

[39:33] if there's any, you know,

[39:35] any boards of directors or CEOs or whoever listening to this, you really need to come at this from a risk perspective and adjust your AI governance appropriately.

[39:47] Pamela Isom: Okay, so this though, is for business leaders who are looking to advance and grow their business and cultivate what they have going on so that they may survive and thrive in this AI era.

[40:01] Right. And just figure out what to do in this digital transformation world to keep pace, to keep sharp,

[40:09] and also to just strengthen our brands. Right. Or pivot. Right.

[40:14] So maybe we need to start thinking about pivoting.

[40:17] And so the conversation is about that, and it's intended to help businesses through that process. And so what you have shared today, I think would be very instrumental for,

[40:27] actually for the boardroom. I think the boardroom leaders should take a listen to this call and pay attention to some things that were said here, because we gave some specific examples of how things work well,

[40:41] as well as how things. You did how things work well and how things can easily go astray.

[40:49] And so I appreciate you taking the time today to share your knowledge and to share your wisdom and to help us move forward in this complex world.

[41:01] And I just. I just really appreciate it. So thank you for taking the time and for being here.

[41:07] LeAnn Oliver: Thank you for having me. I thoroughly enjoyed this. And I think your point about boards,

[41:14] you know, that is a really good thing for a board to look at. I mean, the CEO and the staff need to think about it, too. But really,

[41:23] in a large corporation with a board,

[41:27] they can be really helpful because they may be aware of other angles of things.

[41:34] I think part of the reason that I look at this the way I do is because I didn't spend my entire career in it.

[41:41] I did other things that had that tangentially hit it.

[41:47] But because of my background, it's easier for me to take a look at the bigger picture and identify it, because I've always looked at whatever I'm doing in that bigger context.

[42:00] And it helps make better decisions, I think.

[42:04] And that's the kind of thing that a board should have a bigger context than just what the company. That company is doing. I mean, that's why you have boards of directors.

[42:14] So this is a really good thing for boards to think about.

[42:19] Pamela Isom: I understand. I sure do appreciate it. So thank you again,