Jon Brewton, CEO of Data², Discusses Key Questions When Addressing AI

Welcome to another episode of National Leaders Talk on your host, Mark Stansbury. And I should say also this will be aired on new scape hired advisors, cohost of that with Don Vex. And today we're honored to have with us John Brunen, that's J.O.N., the R.E.W. T.O.N., Chambler of the Data Squared. And he's going to discuss with us, impairing those that are listeners today, impairing them to, in whatever discipline they may have. We're going to be talking for a bit on higher education and AI, other disciplines like that. But we wanted to make sure that all listeners are welcome to this across the different disciplines. And there's ways to look at how this could be incorporated in their daily life, as well as their corporations or whatever area of business they're in. And also how I can apply as far solutions. There are definitely a lot of missed communications going on when it comes to AI. And so for verification and capabilities, I can't think of anybody better than John Brunen to put us on the road map and pathways to get ahead that way. And I'm going to turn over to John with that. He's been on our episode of National Talk several times. This is going to be aired again on both National Talk and new scape hired advisors. And this will be something that will be an ongoing conversation. And I'll wait the next few episodes, but ongoing off and on during the year, 2022, six, our next episode. For example, Don and I will be co-hosting our new skate, hired advisors, the Chancellor of Higher Education for Oklahoma, Sean Burridge. A chancellor will discuss about the 26 universities across the state and how he's trying to incorporate or at least address AI and capabilities there as well. So here we go, John Brunen, John. Thank you, Mark. I really appreciate the opportunity to come and chat with the audience again and talk a little bit about AI, which is pretty exciting at this point. No question. Well, I really appreciate the opportunity to come and talk to everybody again today. And I'll start with a quick introduction on myself. So my name's John Brunen. I'm the founder and CEO of Data Squared. At Data Squared, we are a company that's really pioneering next generation artificial reasoning technology that enables organizations to transform how they understand and use their data. Prior to founding Data Squared, I served as an engineer and a strategy advisor in the oil and gas industry across a couple of different companies working with both BP and Chevron pretty much everywhere in the world. And Scott learned Nigeria South Africa Brazil, the UK, Thailand, Angola, Australia. You name it. I'm going to probably spend a little bit of time there. And during that experience, I really understood the value of how you take really important, but disparate data and make better decisions and drive better decision quality in your organizations by preparing that data correctly. So that really springboarded myself and the other co-founders into starting Data Squared. And ultimately, we're kind of leading the market and the development of advanced knowledge graph and AI technology that delivers fully traceable and trust worthy and explainable insights, which is very, very unique within the market. Everybody started the term hallucinations. We as a company have the only patent in the world for how to reduce hallucinations to zero and ultimately create auditable trust worthy and explainable outcomes from any AI model that you might use. Such as a little bit about me and in my background, you know, what I'm hopeful that we'll be able to get into today is kind of moving past a lot of the buzz words and the chat bot discussions that I think really in and date a lot of the day today conversation and focus really on how value is created. And I think there's a couple of key table stakes, sort of foundational items in that space and it starts with having clean, well understood data, formulating the right business questions and then using predictive models to sort of estimate what you don't know. And then applying decision logic to sort of make better or augment decisions, you know, for the better of your companies. I think what we've heard from a market perspective is that a lot of AI efforts that are being taken out by companies are failing to generate a lot of value. And it's not because the math is hard. The real problem here is that people run into issues with clear problem definition. And the sense of structure around the programs themselves really need to be thought through well and the ownership of the program of work itself needs to be clear, defined. And ultimately the people doing work need to be empowered to do that. So what I'm going to try to do for the audience is just give you a practitioner's perspective and really kind of pick up the conversation, you know, to a point where hopefully by the end of this discussion, you can be a better judge yourself. And then the companies come to you as leaders in the companies that you're running on a day to day basis and pitch new AI solutions and new AI platforms to you. You can be a better judge of the claims that they're making and the potential value that they can bring is a byproduct of the conversation that you know, we'll ultimately have today. That's wonderful. Please proceed. Great. We're all listening ready to take charge. I'm not here to sort of forecast the next model release or really argue, you know, what everyone should be doing with AI. I'm ultimately here to sort of use my experience in building AI systems at scale and the defense intelligence and the energy enterprise environments to help you better understand how to sort of parse out what's true, what isn't true. Ultimately, you know, make better decisions as decision makers, but I think to do that well, we really need to start back from, you know, where where we've come from since 2022. And you ultimately where we stand today and what it means for you as the, you know, the folks that are actually making these decisions and evaluating these opportunities and programs of work. It's really important just to understand how we've sort of evolved over the course of the last four years. So prior to 2022, AI for the way that most of us understand it was experimental. It was siloed for a concept driven most of the work that was being done by AI companies or companies that were investing in R&D and research and AI was done in partnership with academic institutions. And was really sort of siloed in a small non commercialized space. And then that all really changed in late 2022 with the release of the first chat GPT model. That was a bit of an inflection point from the market because generative AI went from nothing to consumer scale very, very quickly. And then we really forced enterprises around the world to move from and we, or how do we use this to we need to figure out a way to incorporate this into our business and our processes so that we can capitalize on the value that we can potentially gain from using these tools. Now, 24 to 26 organizations were really moving from pilots to trying to develop agentic systems. Trying to develop solutions that could reason and act autonomously with very little or no human oversight. The real question and the sort of problem that manifested was whether or not those systems, these autonomous systems, these agents could really do what they needed to do one without oversight and two without clear and well defined and managed governance boundaries. So we've kind of gone from this situation in a very short period of time where we went from a very experimental silo mostly university and sort of organization research driven expiration to commercial scale in a very, very short period of time and people are trying to sort of capitalize on everything that you can get from these systems and these solutions. I think without really doing a lot of the table stakes work that is required to enable success. And so a lot of that is really about data quality governance models and organizational readiness like those three things are more determinant of your success. And then model capabilities or quality really would be. And so that's really what we found over the course of the past couple of years. And so if you don't have these things in place as decision makers as program leaders as managers of solutions that are being deployed. And so you know it really is important to understand kind of where we came from and what it means for how you should interpret the future and how you should really plan to be successful with your own AI solutions. I'm glad you're saying all this because I'm confronted quite a bit regarding how you tie the work force along with the C suite in the board and to have everybody on the same page. And it's very hard to do that with you're looking at budgets from all different areas at different disciplines within say it's energy or higher education and making sure there's, you know, the proper tools but not so much duplication because I'm seeing quite a bit. And duplication to the point because there's a lot of smoke experimenting with how we going corporate AI. And then there's some that are really hesitant waiting for all the kinks to be worked out. And then they come in and say, now I'm ready to use incorporate. I'm not talking about small companies here, some of the larger companies that I've talked to guys, well, we're not really there yet. We're going to use the people we have in place right now to handle the data, data management, so forth. We'll incorporate AI along the way. And I'm been saying, if you don't incorporate it soon in your organization, you're going to be left behind in a big way. So what you're addressing is so important. I know you're going to be speaking at the BKDM, professional petroleum data management association, and you'll convention Houston in April. I'll be there as well, speaking on really on a panel about new energy, which is really going to be including what we're talking about today, looking at AI and how to incorporate that along with robotics and the data management as well. So this is such a broad issue right now. And it's seen from its own the news when I turned off the television all along the last thing they talked about was AI. And so you're under the right subject now. How do we get this going? And you're heading that way for us today. Thank you. Yeah. Yeah. No problem. I think two things. Your point about, you know, really incorporating or at least trying to incorporate AI today. I think it's an important point. One of my professors at Harvard is really famous for, you know, one of the quotes that he's famous for is, AI is not going to replace people, but people using AI will replace people that are not using AI. And that just sort of remains true, right? As the as the industry changes and his capabilities and competencies change, the people that are on the front end of that change and become real practitioners of that change and of that technology tend to outpace from a success perspective, people that wait, you know, that are that are ladders or, you know, slow followers. The other big thing is that most of what we hear about AI isn't necessarily wrong. It's just being told to people through the media from the perspective of people who benefit if you believe what they say. And I'll kind of get into the details of what I mean by that. But the narratives for what's happening with AI have changed so drastically, so quickly in the last three to four years. And really, it's late 23 through current. So maybe the last two and a half years that I think grounding and understanding in that can really help you understand kind of what you see and what you hear in the media on a day to day basis. Because to your point, like every news organization, every program, they're talking about AI. They're talking about implementations. Now, in many instances, they're talking about it from their perspective of how it's failed. You know, very recently, we had a strike on a children's school that was executed via a volunteer system with some backup, you know, reasoning that was being done by one of the larger language models. And, you know, mistakes are being made. And I think if you understand sort of the table stakes around how this information is being used by these systems and the narratives that are being pushed in the public way, you can start to cut through that noise a little bit. So, you know, I can start from what the world thinks happened in 23. You know, I think the world thinks that generative AI arrived and it immediately began reshaping every industry that AI was on a near-term path to replacing large portions of our workforces. And AI first became synonymous with competitive advantage. You know, each successive wave of like new technology or new capability, whether that was agents or artificial general intelligence or the infrastructure push that we're seeing today represented progress towards the same inevitable outcomes. But it's here, it's active, and it's reshaping everything that we're doing in a meaningful way that's changing our workforces and our efficiency curves as a business sort of around the world. There's an implicit belief that this information requires. The narrative there is evolving and changing because the technology is advancing, but is that really true? Like, the real question you have for people is, do you really believe that AI has changed tremendously in the last three years? You know, the answer to that question is no. Like, if we look at the first release of chat GPT in 2022 or late 22, to today, there's small and minor modifications and capabilities that have been sort of changed over that time. What's actually happened just from a market perspective, is this underlying economic and organizational reality is sort of becoming clear? And that's driving a lot of narrative change. Because value creation, systemic value creation is really lagged behind community belief. And so capital tends to need to find a new story. And what followed wasn't necessarily a clean technological evolution. We didn't see massive upgrades and capabilities of these tools and these systems. We saw sort of marginal increases and capability and marginal increases and value creation. And it really sort of tests the limits of, I think what people believe these systems are capable of inherently. And then what they find they are capable of when they do these pilot tests. And so, there's this narrative that really shapes our expectations as leaders and as managers and as people that are doing the investing into these systems that actually doesn't match with the technological evolution reality. And so, I think we have to look at how that's evolved. In 2022, I think everybody said this changes everything immediately by proxy knowledge work would sort of collapse into a series of prompts and productivity would explode with significant organizational changes, slimming down our employment base and increasing margin on all of our activities. What actually happened was we got some really neat demonstrations of potential capability that we couldn't scale. We did get faster at experimentation. And we had significant and broad accessibility. You know, like we went from nothing to being able to access multiple models very, very quickly. But structurally, there's a couple of issues with the technology and how people are using it. One is there's no persistence of context whenever you start to use these systems in a one-off capacity. There's no accountability for the outcomes. Most of these systems are black boxes. You can't actually see how they reason what data they touch, how they return the answers that they return. And because of that, there's this lack of systemic ownership of the decisions that are being made as a proxy of using these tools. Everybody's heard it and seen articles that, you know, I made this decision because Chad GPT told me to do it. That's really sort of an abdication of responsibilities and individual to sort of running these programs. And the other big thing is that a lot of these systems still to this day struggle to work on anything more than a single mode of data. So that means you could potentially load up a spreadsheet, but you can't point it at a database and say internalize all this information and now make me a better decision maker within a given workflow. Whether that's least operated expense optimization, whether that's water management from a produced water management perspective, maybe that's drilling a well better. You still have to prepare the data in a one-off capacity to engage with these systems. And so there's no real clear path for durable enterprise value creation. And so what we saw as a result of that was a massive initial adoption curve. You know, I think the last time we talked, I gave an example of the guy from Microsoft, the guy that was contracting Microsoft and he said, you know, well, I signed up 4,000 people, nobody's using it. But you see this massive initial adoption curve. There's limited production grade leverage that was created from that. And so value was really concentrated in the platform providers, not the enterprises and not the users. And so that really stressed the system because we had all this sort of emotional promise that was delivered from a lot of media coverage. And I think a lot of hyperbolic positioning on what you could actually do with these tools. And that really affected capital behavior. And so the next thing that you saw is people really looking at, how can we change the narrative here? What's the next evolution of something that's going to change people's focus and really push them to sort of expanding the capacities of these systems in their current deployments? It's agentic systems. You know, we can talk about agents, but in general, it's these autonomous workflows. It can be executed without people and executed at the same level and with the same rigor that people execute them today. And you know, we had sort of a first iteration of agentic systems that were called RPAs. And RPAs are just workflow tools that were executed little bots that would execute, you know, filling out a spreadsheet or filling out a application or something like that. So think of it like that. You know, we have these robotic process automation tools and the evolution of that is now their autonomous. You know, we can essentially use these tools to do parts of our work and make ourselves more efficient. At the high level, generative AI proved to be powerful but very brittle. And so it could generate a little bit of value and you could create use cases around it, but it couldn't operate autonomously. It could and respawn. on the questions that were asked, but it couldn't really influence an individual company's business outcomes because the aperture of the data that would be accessing at any point in time was too narrow. And so the new story really evolved in agents will come in, they will act, they will plan, they will coordinate. Autonomy is sort of the missing ingredient here. And agents are framed as this bridge from tools to labor replacement. In some settings, we've seen that manifest, especially when we start looking at software engineering as a practice. You see Amazon and other companies laying off large parts of their software engineering workforces. And that has to do with some of the immediate promise that you can generate from these tools. But it's a very, very narrow application. So what actually happened is agents increase complexity significantly faster than they increased reliability or outcome success. So the failure modes that we got from the initial systems deployments actually multiplied. And debugging these systems became significantly non-trivial. And so governance as a byproduct became harder, not easier. And it's really because we lost this sort of linear process transparency. It became obfuscated. So the net effect was agents worked in narrow constrained environmental applications, but the broad enterprise impact from agentic systems being built and deployed to this day really remains elusive. And so we had to sort of change the narrative again. And everybody that's listening to this, if they paid any attention to the news, would have heard a year ago, late 24 early 25, artificial general intelligence is right around the corner. Everybody says this. I think the anthropic CEO said two days ago that I think my systems are actually alive. They're working on their own. And if you've ever tried to apply these things, they're not. So maybe they are in a research capacity that nobody's seen. But AGI is a concept, something that is going to change the true value proposition from AI systems, true artificial general intelligence, meaning these systems can do what we do with no coaching, no guidance, and can do the job that we do on a day-to-day basis as engineers, as doctors, as other people, as lawyers. They can do it without fail at the same level and at the same capacity we can. That's just not a reality. And if you look at some of the structural limitations associated to how these systems perform, the new story that's being pushed, that artificial general intelligence is right around the corner. And the current failures in the system are just temporary growing pains. And the breakthroughs are just one model away. They don't really hold up to any rigorous analysis. And so I think people really need to understand a lot of the narrative shift that people are seeing really starts to be, what are we really trying to say here about artificial general intelligence and just AI in general. What we're trying to say is that we're transferring reasoning across domains and we have human-like capability and around the corner tomorrow, this is not going to be theoretical, it's going to be practical and we can apply it. Now, the truth is that if you want to apply that, valuations associated to the companies that are providing these solutions are inherently anchored and directly tied to the net present economics of the business and the scale of the business capacity. And so the narrative sort of increases in complexity, but it draws a clear sort of steel thread line to what we're seeing today, which is data centers as destiny. It's infrastructure as the savior for artificial intelligence systems. And these things, if you really look at them and you can understand, they've evolved because of the lack of delivery and the lack of promise and the lack of application that we've seen in each successive way, but to go from scaled-agentic systems to artificial general intelligence, the only answer is more compute capacity. And so now we sort of see this infrastructure as the savior narrative. And it's not that AI has failed, it's just resource constraint. And if we fix this bottleneck, we can scale compute, we can scale energy, we can scale data centers. The value creation itself will finally follow. And there's just sort of a scrutiny that needs to be applied to this. As we start to look at the market and what value we're actually creating, if we go back just a minute, like we haven't really seen any major technological evolution in these tools since 2023. A lot of the things that are being done today are roughly the same as they were three years ago. They might have a little bit more scaled capacity in terms of how much data they can review. We might have a little bit of sort of autonomous process as we can work in, but the reality at the end of the day is that compute enables capability, not correctness. And so that's a real problem. But yes, we can scale these systems up with more compute capacity. But does it mean they're gonna be right more often? Does it mean that they're actually gonna enable value creation in a fundamental way and a more scaled capacity? The idea that infrastructure amplifies both value and failure is important. Because right now systematically scaling intelligence does not automatically scale trust. It does not automatically scale alignment and it sure does not scale usability. And so we really need to think as decision makers, like what's going in to our programs of work that are gonna make us or at least raise our capacity for success. - I'm so glad you're addressing this because I'm the S-Hawry Poor Resign that was supposed to be the best report, accurate, reliable and so we started trying to tear it apart in a sense to make sure. And the accuracy was not 100% by any means. It was in the 90% round. But at the same time, so I get higher folks can get closer to 100% than I could the report I saw. And so I'm going, wait a second, the affordability, the reliability, the effectiveness and efficiency was what since 2023 that I've seen as well, missing in the roadmap. And it's like getting our road, it's a straight road at first, but then all of a sudden you're going off the road to the left and you come back over and try to correct it. And you go back over the ride and you try to correct these things and that's where we are in the sense. Until I came across your company and I thought, while you have some patents, you have some answers to efficiency. And I think that's what has been missing. And so that's why I wanted you to be on this episode as well as other episodes that you've been on to get us on the proper footing. Right now the proper footing is we're getting off the road back and forth and really need to have this straight pathway and to get to where we need to be in all these different disciplines. Could we talk about energy on a show? I've got a higher education on a new skate, but there are so many different disciplines from health to you to aim it. Financial institution and so forth. So I just wanted to interject that. Continue to prove. - Yeah, no, it's really important point. I think people don't really know what they don't know. And I think a lot of what people see, especially if you sort of look at an abroad macro sense, the AI investment economy seems to be booming. - Yes. - The truth of the matter is, the AI investment economy is essentially circular. Like if you really look at this and Bloomberg has done an amazing amount of reporting on this, large parts of the AI investment cycle right now and the economy are circular financing. We essentially see money flowing through interconnected deals among dominant players and flading valuations without clear end market demand. And that's like not a very good imperative for technological evolution, which I think actually underpins the point that I was making before. Like we haven't seen a lot of technological evolution from 23 to 26. And a lot of it is because the factors that determine the capacity for technological evolution are sort of focused in in customer demand. And at the end of the day, we see a lot of artificial investment demand driven by circular financing. And what that sort of looks like in reality is in videos investing in open AI, then open AI buys in video hardware, then orica builds infrastructure around that hardware and revenue cycles back to the same capital bulls for those three entities. And as a result, you'll see valuations outpacing revenues, roles start to blur. You have these insiders creating force, sort of self reinforcing loops. And it's reminiscent of what we saw in the late 1990s with the dot com era. At the end of the day, it doesn't necessarily mean this is a bubble, but it does set sort of a very clear distinction but financial and incentives in the market right now are driven to amplify valuations and The perception of demand and so we're over building capacity right now to demand But if we're ever going to deploy agentic systems at scale and all the stuff we need it But the result is money's kind of flowing into these these loops and that has indirectly or really directly Impacted technological evolution over the past couple of years And so I think a lot of people think well God there's so many dollars flowing into the market right now and into this sort of AI economy that you know the rate of change in the pace of change from a technological perspective should be tremendous and the truth is there's a gap between that economic reality and how capital is being Alligated today and so that's really where you start to see these and at the end of the day The real reason for that is because of the factors that can Tribute to scaled application of AI solutions for any business are outside of the scope and control of the providers of hardware infrastructure models and systems It has to do with the enterprises themselves and it has to do with several key things but data gravity is probably the most important thing driving this right now and Data gravity is a concept is really interesting because the idea is that as data grows in size and importance and connectivity it becomes increasingly hard and more costly to move and Instead it will pull applications compute and decisions towards that data Much like a massive object will pull things via gravity Towards itself right or the core idea is that simple Small data sets are easy to move large complex business critical data sets are not and Over time the idea is that systems will adapt by moving computation to the data and so Instead of people asking where should we run this AI model? The real question should be where does the data already live and exist and how do we reason within that space? You know, I think there's been pushes for the last decade and a half mark and you would have seen it and every other person listening to this would know Data lakes data lakes data lakes. We got to aggregate all of our intelligence and that was sort of trying to Indirectly manifest data gravity right and it's it's not Something that actually proved to be super useful at scale and then data lakes are useful But like in certain applications and in a managed capacity You know data gravity itself especially for AI is a thing that really Henders scaled capacity and scale value creation and so the truth is that Context for any problem space that we're operating in doesn't really travel well when we move data from one repository to another We tend to lose a little bit of fidelity in that information So we lose the understanding of the relationships the history the proven odds and sort of the overall business meaning of that information And if AI models are trained or run without that context they hallucinate Significantly more they make very brittle recommendations and quite frankly the end users will lose trust in those solutions quite quickly Now what that means for the businesses that are probably listening to this conversation regulated industry in general finance health care defense energy Data residency laws dictate how much movement we can do audit requirements and force how we you know We actually manage that information Chain of custody rules determine how much we can move what we can move when we can move it and that actually Increases data gravity and so if we don't sort of solve for being able to deploy AI systems Into where the data already exists with the context it exists with We start to lose fidelity and we start to lose the capacity for scale and so what actually happens is Model gravity becomes the determining factor on success And it's like where the model sits and so the closer to where the data is and the importance of that data And the ability to connect into more than one system where that data resides Is the key competitive advantage for building trusted solutions? So data gravity is a really big piece of this equation The other thing that I think really needs to be addressed and this is sort of a derivative of that is that Enterprise adoption bottlenecks are Really stalled on trust governance and integration So it's not really intelligence. It's not really the capacity to make better conclusions It's that if we can't trust Where these systems are accessing the data the governance around how they access that data and the rules that are enforced and how they manage that interaction And ultimately how to integrate data without moving it Then we're kind of bottlenecking ourselves, right? And so I think where a lot of people say we need to build a new data warehouse for a new data lake You know ultimately what happens when you do that stuff agenda doubling your cost you lose fidelity and You sort of reduce everything to a new structural reality that might not be fully indicative of how you run your business on a day-to-day basis And so that really impacts the likelihood of success for when you deploy these systems and when you build them and so You know, it's just like table stakes stuff people need to understand A AI can solve a lot of problems and it does have tremendous promise if it didn't I wouldn't have started a company to do it But you have to understand where the data comes from you have to understand who's accountable for the decisions that are being made You have to understand what happens when the system is wrong like how do you manage those interactions? And you have to be able to define whether or not the solution that you're building is Scaling trust or scaling risk and you have to be able to do those things if you can't do those things You're not really going to be able to effectively evaluate AI systems and the claims that are being made from these systems, you know without that sort of framework and thought process around it It's a good point to reflect what you just said and to put it in a summary form if you will because there's a lot of information provided today There's those that are placed in different disciplines different leadership roles that each one has whether a student or a CEO You know those that are listening to this can make application of what you've you've talked about as well as reflect on How they're going to take their leadership roles and in incorporating AI their daily life and so if you would John There's so many more things we cover today. That's a good stopping point to reflect But in a summary if you would yeah, and then hopefully get you back soon. Yeah, perfect. I think AI misinformation sits on two extremes right now. It's either an existential threat That's going to either kill us all or destroy all of our jobs or It's this magic intern that we don't have to really do anything with it's just going to do our jobs for us And like most things the truth is somewhere in the middle right those are those are sort of two polar extremes That's how people view these things But a practical framework that we can use as decision makers as leaders and evaluators for these things is really A couple of questions. Where does the data come from? You know like what is the data provenance What is it grounded in what is the governance around that and how do we use those things to determine trust worthiness? You know without a for a date of sources outputs tend to become indefensible So if we can go in and say we understand where the data comes from who's accountable for it and ultimately What happens when the system is wrong? We're in a really good place So I think if you can just sort of think about it within that framework. Who's accountable for the decisions that we'd be making with these systems? What happens when this system is wrong? And does that the answer to these three questions scale Trust or does it scale risk you can answer those questions as a decision maker when you're evaluating these things You will put yourself your teams in your investment and a much much much stronger position To create value and to create success for your company And it's really as simple as that like if we can answer those questions We can really start to make better decisions about how we invest in AI systems What we expect to get out of them And ultimately what happens when they don't perform the way that we think they're supposed to perform and that ultimately If we get those things right is going to scale trust and value quickly So I'll leave you with that and ultimately look forward to the next conversation That's wonderful job Yes, I look forward to the next conversation and thank you jog for I'm delivering this in a way that we can all take it and put into practice And to know at least those questions and there's more to come I'm sure But those at least get things wrong whatever position you're in You may have been using AI for some time You may just be able to start with those questions Apply to every every scale of the operation. So appreciate that John Thank you to your team. Thanks to Jordan. As a producer, we will get to have this on National Energy Talk as an episode as well as New Escape Parties of Rogers as an episode and then we'll come back together hopefully soon in the next couple of weeks or so and follow up on where individuals are. Hopefully get some feedback. We love to have questions. I love to have input from those who are listening and you've been listening to National Energy Talk and New Escape Parties of Rogers on your host, Mark Steyasbury, stay tuned for upcoming episodes. Thank you, Jordan.