Operationalizing AI for National Security With Lt. Gen. Kirk S. Pierce

Lt. Gen. Kirk S. Pierce is the Commander, Continental U.S. North American Aerospace Defense Command Region - 1st Air Force. He joins Mark Valentine, Head of Scale Federal to discuss the potential of AI to provide direct assistance to warfighters within the contexts of Information Dominance and Decision Superiority. They explore the operational risks that could slow down the adoption of AI to support broader joint All Domain Command and Control. Delivering artificial intelligence to warfighters is a strategic imperative but there are still many questions that remain unanswered. What will the adoption of AI mean for the United States military? How can military organizations take advantage of this emerging technology, as they look towards current and future national security imperatives? Join this session to hear an example-rich discussion of the opportunities for AI to support national defense.

Mark Valentine (00:08): All right, good afternoon, everyone, or I guess good morning, depending on where you are in the world. Uh, welcome to transform acts again. My name is Mark Valentine and I'm honored to lead the federal team here at scaling. Hi, I hope you've all had a wonderful event so far. I know I've already learned a ton from all the content and the speakers, and I know this next session is going to continue to trend and hopefully amplify it, uh, in the session. We're going to continue our exploration of the nexus between artificial intelligence and national security and focus a bit more tightly on how to operationalize AI capabilities, uh, for the war fighter or for the frontline operator to do this. We have a very special guest and close personal friend of mine who might have the honor of serving alongside for several years, if not decades.

Mark Valentine (00:54): Uh, and that is, uh, air force, Lieutenant general, Kurt Pearce, or as I've known him, uh, tech, uh, we, we go way back, but, uh, Joel peers is the commander, uh, first air force, as well as the commander of the continental region of the north American aerospace defense command or commonly known as Nora. So in other words, he's got a really big job and he responsible for all the coalition and joint air defense in the United States, as well as being responsible for planning for, and responding to a coordinating us military efforts, uh, for disaster response here in the continental United States. So he's been serving in the U S air force since receiving his commission in 1988 and served in numerous roles as fighter pilot, a staff officer, and a commander at the squadron group and wing levels. And now the number of airports level, uh, he's also, uh, serving combat during multiple combat deployments to Iraq and Afghanistan. And he holds a bachelor of science degree in aerospace engineering from the university of Notre Dame and his, he attended just about every professional military education experience that one can including earning a master of science and national security strategy from the national war college. He also has had multiple post-graduate experiences from outstanding institutions across the U S including Syracuse university, MIT and Harvard university. So Joel Pierce, thanks for taking the time to be here today and welcome to transform X.

Lt. Gen. Kirk S. Pierce (06:26): Mark. It's a, it's, I'm really happy to be here and thank you for the opportunity to talk to transform ex uh, hopefully you can take the word from, uh, a dumb fighter pilot, as I'm trying to work through all the issues of command and control and how do we increase our decision space to senior leaders, both horizontally, you know, people that I work with and the team works with and, and up through to the secretary of defense and ultimately to the president United States who makes a lot of the decisions, but, um, how do we work through those things? Uh, both at Nora and us, north comm, and now our newest combat command air force, uh, that we're supporting is the air force component US-based command. So, thanks again. I appreciate that.

Mark Valentine (07:06): And so for your dumb fighter pilot comment, I resemble that remark as well. This is going to be a great conversation. Uh, so actually to kick that off, uh, it brings up a great point, cause I'd like to start off by looking at some of the potential use cases for artificial intelligence and machine learning or AI and ML and the operational military. And I say that in the context of just about every time I pick up the paper these days, or for accurately, I read a news story online. Uh, anything about the department of defense, I'm reading these grants, weeping concepts, like joint, all domain command and control, or the advanced battle management system also known as AVMs. So these things they're at least related to AI and ML technology, at least above them to believe that the news reports. So can you share with me what some of these concepts are and why it's important to develop them in the first place and again, with the context of the operational war fighter, that frontline operator.

Lt. Gen. Kirk S. Pierce (08:04): Yeah, so that, um, I'll try to walk through that for us. So here at first air force, obviously, uh, we're really concerned with Homeland defense and Homeland defense across the spectrum with, uh, uh, a definitive role in the air domain. But, you know, it goes across cyber, it goes across space. You don't do anything, uh, in the air domain, that's not having an effect there or underwater or on land. Uh, how I look at it is, you know, we have this concept of joint, all domain command control. And again, it's a concept that the joint staff came out with a war fighting doctrine and tried to establish where we're going, which is good because that helps the services understand, you know, we need to have interoperability. You cannot continue to have data stovepipes and tremendous amount, wherever that data comes from, uh, whether it's a sensor, uh, above water, below water in space, whatever it is, we want to have the data to be able to make decisions for here at first air force with Homeland defense, being our primary mission, peer near peer, violent, extremist, whatever it ends up being that for air defense and air sovereignty.

Lt. Gen. Kirk S. Pierce (09:05): And that's really on the continental NORAD side, or it can be with north comm. Uh, we look at how could you use AIML and, and you're right. I read the same things, probably reading the same articles and too often, uh, from an operational standpoint, you know, I understand what we're trying to do, but from an operational standpoint, people will throw out the easy button, oh, AI and ML will do this in AI and ML will do that. Um, I've tried to go get smarter talking to different federally funded research centers, uh, Northern and north comm. The headquarters had done a lot of work on this. I really give a lot of credit to the team out there. Uh, it started under general Shaughnessy and then, uh, continued on with general van Herc. And they're really trying to push for all domain awareness, which that's part of jet too, which I'll get a little bit more in depth with that.

Lt. Gen. Kirk S. Pierce (09:52): And then really they talk about it as, uh, information dominance in decision superiority. You know, we turn it into air terms to an extent, but really what that means is we want to be able to bring in all the data that's already out there under this Jed C2 concept. You know, so take, for instance, in Homeland defense, we have several radars, both in Canada, so nav Canada, uh, or in the U S the FAA radars that feed into our command and control system. And then we have military radars that feed into the command control system. We've never had the bandwidth to take every radar that you see out there, a radar on an air force or army range, a radar that's used by maybe DHS for, uh, counter-narcotics and bring all of that data together and try to fuse it into a single picture, which has nothing to do with AIML.

Lt. Gen. Kirk S. Pierce (10:40): So we're still working down that path of trying to bring all that data together, but we're AIML could help with this is if you have that volume of data and artificial intelligence and just simple. I mean, it's a little bit more of the machine learning, but algorithms can look at the data before we spend more money on more sensors, which there is a case for it. But before we do that, let's see all the different sensors that are out there, whether it's an acoustic sensor, whether it's a radar sensor and I'll stay in that one to explain, um, or whether it's a, an electro-optical or infrared sensor, um, there are data that we just basically leave on the cutting room floor, like, because the systems weren't built to ingest all that because they were built for a specific purpose. So most of the radars were built for, uh, trying to detect, you know, like an FAA radar is trying to detect, uh, an airliner, but it's really looking for a transponder or a code in the radar is the skin paint backup to that.

Lt. Gen. Kirk S. Pierce (11:37): It was never built for a low observable, you know, low radar cross-section, whether that's a weapon like a, an advanced cruise missile that the Russians and the Chinese have developed, or just, uh, a home built aircraft. Um, they weren't built for that. So we try to that, but the capability is inside there's radars. It's just that, that wasn't the design spec. So way back when, you know, the FAA did pay for more, uh, on the military side, we didn't design it pay for more. So how do we use machine learning to look and go through that and try to build better algorithms and bring it, and then once you bring it together, how does artificial intelligence help us with the art? And I'm less about what the presentation is. Uh, I'm more about the command and control aspect, cause that's really what I control and what are my team here at first air force.

Lt. Gen. Kirk S. Pierce (12:26): And the 601st air operation center is trying to do for the continental United States in our NORAD hat and north America in our us north comm or SpaceCom US-based combat. So that's really what we're trying to get after and look at. So purposely have gone out and put people in that, you know, the air force research lab, uh, we've put people into some of the FDRC places. We've put people at, uh, uh, the headquarters for acquisition, so Sapphire IQ to try to influence it. So what we don't end up like what you're reading is this super common operating picture, although very helpful. That's nice, but it's not really command and control. And, and that's the piece that we're trying to get after. So, um, in a Homeland defense scenario, if we could fuse the data into a unified data library, something that we can clean and have a metric to figure out, is it good data, bad data, because even bad data we want cause it could show trends, right?

Lt. Gen. Kirk S. Pierce (13:25): And the, and right now it's no kidding the human being in the loop trying to figure out, okay, that's a good track of interest or that sell very low quality track of interest. But I have five different radars giving me a low Trop, low quality track of interest within four miles of each other. Is it something, or is it a non-organic, you know, like an aircraft is an organic, like a flock of birds or, uh, is it just an anomaly with the radars? Cause it's a weather anomaly. Well that consumes a lot of brain power or I think artificial intelligence could look at that. It also could look at trends, Hey, during this timing due to solar flares of weather, uh, anomaly, whatever it is along the Appalachian mountains, outside DC, you know, every five days at this time we get this anomaly, well, we have a historical reference for that, but maybe that anomaly when it shows up on the third day, really wasn't an anomaly, but we got trained to think it wasn't anomaly.

Lt. Gen. Kirk S. Pierce (14:22): So that's where I really think artificial intelligence being inside of a joint, all domain concept. And the piece that gets forgotten is what the C2 is. Joint, all domain command and control, not just join all of the main, like when we grow up a blue force tracker that shows everything, oh, that's great. Um, you know, some common operating picture that shows all the known, uh, or, uh, unidentified aircraft. That's very helpful too, but it doesn't get after the, so what of artificial intelligence, bringing the things together and go, this has a 95% chance of being a possible threat. That would be nice to know.

Mark Valentine (15:01): Yeah, that's, that's intriguing. And so what I'm hearing from you is that this idea of the ultimate goal is to achieve information dominance and decision superiority. And although I am not a data scientist, I've taken through rudimentary courses on it. I think back to some basic concepts in data science, which is to take, start with data from data determined insights and from insights inform action. So it sounds to me that that is the, the continuum that you're describing there. So to help maybe me and maybe the audience with a specific example, are there any like specific operational opportunities that you see coming from some of these systems? And if you want to speak to me in my old terms about what it would mean to me back in my old fighter airplane, or maybe sit in the chaos, uh, that would be helpful.

Lt. Gen. Kirk S. Pierce (15:49): Yeah, I can do, I can do both. So I'll start out with a vignette Homeland defense to there started there, and now I'd like to go over, uh, defense support to civil authorities disco, or, you know, you talk about human, uh, humanitarian assistance and disaster relief. That's outside the company and the states. So like what we did with Bahamas back during Dorian, but I'll just talk about that as discussed. So let's go to Homeland defense and just do a quick vignette on that. So, um, your a, a fighter pilot doesn't matter where you are, uh, you could be in Alaska, you could be in Canada, uh, w I'll talk kind of in the United States. And let's talk instead of, you know, what we've been doing for a long period of time is, uh, worried about violent extremists, which is still a factor let's, let's talk about, uh, Russia or China, really an existential threat.

Lt. Gen. Kirk S. Pierce (16:31): So there's a continuum of competition and you know, you, as you go through, it's all a competition and people get stuck in heroin, steady state, or we, it at a crisis, are we at conflict? Um, but if you're a fighter pilot or an air battle manager and an AWACS, or at a, uh, in a ground station like the Eastern or Western air defense sector, or the ASC, what you really care about is the information that's flowing in. So in that case, um, if we're able to bring together, let's say there's a long range, Russian aviation, that's gonna fly down the west coast, which has occurred in the past, on the 4th of July. Uh, we want to being start out at the lower earth orbit and have whatever data we can get from a space, whether it's Leo, geo, MEO, wherever it's at, uh, it could be some form of event.

Lt. Gen. Kirk S. Pierce (17:23): It could be communication. It could be, you know, signal. It could be, uh, um, EOIR. It could be whatever, uh, same kind of things from the terrestrial land perspective, radars that may be able to do that. Uh, if there was things that were heard in open source, you know, just cause some people talk on the radios and you could hear stuff there, an airliner that was going, uh, you know, from over Alaska, into Japan and in long range, aviation reports it right now that doesn't really get fused into the common operating picture. So we'd want to get all those pieces. And then anything else that's out there from a Navy sensor to a Marine sensor, to just aviation platform that's flying that hopefully could be all brought in together. And that would enable us to then use artificial intelligence machine learning and connect the dots.

Lt. Gen. Kirk S. Pierce (18:13): And again, getting back to kind of the art. So you're a fighter pilot and you launch out and we've had some indications in warning because that data's been fused because artificial intelligence or machine learning, depending on what really we need to have, cause it's not all an AI thing can, can see the data stream that's coming in, uh, look at the different data libraries fused together and go, okay, this looks like it's long range aviation. It looks like it might be, uh, you know, a Russian bear or a, a blackjack, uh, just doing, uh, some kind of, you know, information campaign against the United States because the geo poll and everything else, and it could have those aspects. It's looking at all the tensions in open source that are out there. Hey, did something get said with something said today in Russia that, you know, uh, that the head of the Russian air force, you know, made some comment that unless you're, uh, you know, just sitting there watching it, you wouldn't know, and it might come through an intelligence lane in two to three weeks, but no kidding.

Lt. Gen. Kirk S. Pierce (19:13): That gets fused in. You see open source, you see the other sensors where they're classified or unclassified or civilian partners and, uh, the AI can help go. Yup. We think that this is a, uh, Russian aviation. Here's the, uh, part of the command control. We can then launch out of the west coast or Alaska fighters and airborne warning to try to go intercept, uh, for our, um, responsibility to identify at their defense identification zone, uh, or we can make a decision not to, you know, we don't always want to respond. You don't need to respond to everything. Um, but we will have the posture to be able to do that. So, so that's where artificial intelligence could help us that there's a bigger piece to it though, because it can also help with logistics, like, okay, if we kept doing this on a 24 hour cycle, again, not we're just in competition, we're not at a crisis or conflict.

Lt. Gen. Kirk S. Pierce (20:05): Um, we can continue to look at what are the logistic requirements, what are the fuel requirements? Is there, uh, you know, with COVID going on right now? Is there something that could be occurring due to COVID, which was never built into our command and control structure to have a pandemic that's going on 18 months and could be affecting the, uh, resiliency or the response of a certain location that we weren't really relying on? Cause I get the response, I get all the information on, Hey, the fighter squadrons, ready to go. The tankers are ready to go the command and control systems ready to go. And unbeknownst to me, the FAA just had to shut down the LA tower because they had a COVID scare of a positive. And now that data source might be gone. That has the one thing of, oh no, those are American. [inaudible] coming back and they were stimulating a Russian aviation for a Navy exercise. Oh, great. That'd be great.

Mark Valentine (21:01): Yeah. I can totally see this idea of decisions for your authority come from what you're talking about, as well as the old adage that I know you and I have heard where rookies talked, tactics, experts talk logistics. So, yeah, that's great. So I want to pull the thread a little bit. So based on some of the things you were talking about and I'll call it the military operational world, I can see some analogies on the defense support of civil Courtney's world, where you've got, again, all the different actors, whether they're from different federal departments and agencies, as well as weather, uh, all these things that, that we can't control. Uh, so can you give us a sense of like some specific use cases you see some of these capabilities use for, and that's?

Lt. Gen. Kirk S. Pierce (21:43): Yeah, definitely Kennan in north comes, worked on us a lot, um, to give them a lot of credit. Uh, and they've been working with other combat commands and they've been trying to use for, uh, start with like peer stuff and I'll get to the discussion, but peer stuff, they've been doing the global information dominance experiments to try to bring AI stuff in. How do we bring in logistics, operational and, and go back to fighter pasta. You know, if you go back to that joint engagement sequence, how do I, uh, get into be able to sense detect track, characterize, identify, and then worry about a kinetic and non-kinetic effect in disco. A great example is, has been COVID. I mean, like you said, there's many missions that we're doing here at first air force in north, comm's doing, and it's going in all the services are doing, uh, I'll just talk COVID right now.

Lt. Gen. Kirk S. Pierce (22:28): Uh, not wildfires hurricanes, but think about COVID could be inside of all of those responses also, but for COVID responses, we have air expositionary groups, uh, last fall, and they're back out again now, which are, uh, comprised of medical readiness teams. And those teams are titled 10 airmen that I've gone out to support at the request of re responses from states. So the state's requested federal assistance that goes up through that and it comes down to the services and they get parsed out across the different services to respond. So right now, you know, we have teams in Louisiana, uh, Alabama, Mississippi in the air force. There's several others around the country and we've been doing that on and off whether it was vaccination or COVID response in hospitals right now, it's back to COVID response in hospitals. So what I really could help us a lot with, and it's been the same way before.

Lt. Gen. Kirk S. Pierce (23:16): I mean, imagine a world where, uh, the airmen goes out with a sensor and the sensor could be just an iPad, uh, with the right tool sets on it, that they get a common operating picture around them. They get the weather. So hurricane Ida, just coming up and we had Baton Rouge and Alexandria. So they're getting all the inputs and they know how long they can stay at the hospital. They can share with the inter-agency people that are working there. They can share with, Hey, do we need to shelter in place here? Or we can work up and get real time information. That's that's doable right now. Uh, then you add in, uh, the COVID logistics that might be going on. Are we share, um, ICU beds? Are we short on supplies? Like, do we have shortcuts in, uh, liquid oxygen at the hospital so that you can think ahead and help them out, uh, when you're working there as a medical team and you bring that piece in and then add in any other piece that might be a significant impact, especially after a hurricane or some other wildfire or, you know, a flood or anything that's occurred that might impact that defense support to civil authority.

Lt. Gen. Kirk S. Pierce (24:18): So artificial intelligence could look at that and go, okay, we have X number of people that are working today in Baton Rouge, uh, at a hospital. And then while they're working there at Baton Rouge, it's looking across the inter-agency and it's looking at as long as the state or, or the hospital is allowed us into the systems, which, you know, that's an authority and they'd have to approve it, but we could analyze that and go, Hey, in three days, we think you're going to have an oxygen problem. Or in three days, we think you're going to have, uh, you know, personal protection equipment, PPE problem, because this is what we're seeing in the logistic train coming here, or we're going to have that problem with the military people that we sent there. Um, and that let's just get ahead. Like you just talked about, get ahead of the logistics to ensure that people who are on point trying to do the job and support our taxpayers and really help out people in need have four to thinking, and they're not spending, you know, two or three people, the commander or the detachment commander there is trying to work through all this stuff and they can get through artificial intelligence.

Lt. Gen. Kirk S. Pierce (25:16): Hey, here's your three coasts. Uh, one keep doing what you're doing and hope it all works out to, uh, be reactive and I'm being simplistic, be reactive. And when a problem shows up, use that great airmanship and a great professional development that we gave you, uh, from an airmen all the way up to a Colonel and apply that science and make it work out or three, Hey, here's the, all the things that are trending, uh, we've tried to coordinate, and these are the actions that have occurred. If this action, you know, if then statements and artificial intelligence would work all through that. If this branch SQLs it, then this will occur. We have a 95% chance that it's going to work out. The PPE will arrive in the next four hours. You have a 30% chance of getting oxygen. If you do run out, uh, we've pulsed the other hospitals and other military bases around Baton Rouge. And there is a compatible because you mean you could go to Barksdale and they have liquid oxygen, but it's not compatible. I mean, I don't know. Um, but you would have that instead of the human being, trying to do their absolute best in a situation where like in Baton Rouge, you have power at the hospital, but you may not have power at the hotel that you're staying at, and you're going to work 24 hours because the people want you to work 24 hours and your dedication to the mission. Um, that would be,

Mark Valentine (26:37): And a lot of ways, this reminds me of that point in our flying careers when new, like the F 22 started coming out with more advanced, uh, automation, I'll say, and cause what you're describing at least to me sounds like AI as an augmentation of human cognition so that a human can focus on what humans are good at and the machines can do some of the nug work that they're good at. And I remember, you know, when some of these advanced airplanes started coming out, uh, being incensed initially because it seemed like they were taking over part of our jobs. Uh, but then later on I realized that, Hey, they're just allowing me not to have to concentrate on those things and be able to focus on those things that I'm good at, like positioning my horses. I, instead of trying to read an IFF interrogation or something like that.

Lt. Gen. Kirk S. Pierce (27:27): So I completely agree it to kind of riff on that. Um, I, uh, spouse, whenever I engage, I mean, I'm always trying to engage on this, um, again, stupid guy. So if I can get it, hopefully others get it. This can't be a bolt-on, uh, AI and machine learning should not be a bolt on. And what do I mean by bolt-on we take the same air force process or DOD process or inter-agency process or whatever it is. And we use PowerPoint and we use Excel and no knock on those. They're great. They have their use. And then we just take AI and ML and we added over top and sprinkling on and go, okay, we did all this stuff. And then your new presentation of PowerPoint gives you the data and stuff that is not helpful, that does not lower manpower costs. It doesn't, uh, speed up, you know, that OODA loop, the observe orient decide act, uh, all that does is it, it takes an incremental step forward, um, and to relay kind of re put it in perspective for what you just talked about, you know, that was giving us a pod on the F 16, that the AOC could call long range over the horizon and go, Hey, I need you to move here.

Lt. Gen. Kirk S. Pierce (28:38): Okay. It was incremental change. That's great because I was out of communication and I just didn't evolve, you know, in my, my pace plan, I didn't go from primary to alternate to continuously to emergency and go, okay, no, one's been talking to me. I can't get ahold of anybody. I'm gonna go to this location and see who's there. Um, okay. Incremental change, uh, but an F 22 or an F 35, or any of those commercial applications, we're not looking for that incremental change. We need to change if we want to be relevant and we want to make decisions at speed. And we want to have factual data and proven validated assumptions. Cause we're not always going to have the factual data to have that. That's where I see AI and ML. So that's where an F 35, 22 analogy works compared to, and it has nothing to do with it being an F 16 or an F 15, but bolting on a piece of technology, helpful, not the big leap that we need to do for not only pure competition, but just in normal decision-making.

Lt. Gen. Kirk S. Pierce (29:34): So back to the defense support of civil authorities, uh, it should be you flip, open the, you know, in the Nirvana, you know, you would flip, open your iPad, you'd flip, open, whatever the product is. Um, you know, um, it could be a surface pro I don't really care. Um, but you flip, open this thing, uh, and you look in it and it has the computing power, and it can give you a common operating picture, but on top of all that built into it and into the processes, it's already got artificial intelligence and machine learning that is working through courses of action. It's continuing to go back to those databases to see if there's been a change, a trend, an input, uh, an indication, a warning, a critical thing from FEMA. Hey, everything's good. Uh, in Baton Rouge except, uh, a new policy came out from DOD that says, uh, nobody in the air force wearing blue can, uh, work in a civilian hospital, uh, unless they have Tri-Care select Tri-Care prime. We know it's a glitch, but it stopped everything. And you're like, okay,

Mark Valentine (30:35): Sorry, the process that you're describing, uh, and I'm using the word process very deliberately because it starts somewhere. Some things happen and we marched through through the process, uh, is dependent upon data. And so you've said that several times, and that's consistent with many of the guests that we've had here at transform max. So this, these data are important, but you've also mentioned several different organizations and I'll call them stovepipes, uh, that have to cooperate with the data they own. Uh, how are we doing on that? And what are some potential ways we can make those people work better together? Because it appears to me that the incentives right now, uh, support holding onto one's data versus sharing that data.

Lt. Gen. Kirk S. Pierce (31:22): Yeah. So I, I think so, um, I agree with the perception that people are holding on to data. Um, to an extent I have seen more and more information sharing, um, you know, we're working on right now, uh, with the, uh, operation, uh, inside the United States and the customs and border patrol have been great about, you know, sharing data they can share, but there are legal barriers in the inter-agency between what the department of defense can do and see, and what a federal bureau of investigation or a law enforcement or DHS and the customs and border protection can share. So some of that we'll have to figure out as we get to those points and not wait reactive, but, you know, a year from now go, Hey, right now we need legislative reform at Congress. And here's the reason why, and make a good case.

Lt. Gen. Kirk S. Pierce (32:11): And for us in the air force, we have to be able to articulate that case of why do you need an authority, or why do you need to see someone else's data? Um, I think another significant challenge, which is unfortunate, it's just a bureaucratic challenge, which the joint staff is trying to get after, by, you know, leaning out and going, okay, here's what we mean by joint, all domain command control, here's the joint war fighting concept. Here's why it's important. And, and trying to lead that charge is like it, or don't like it, the bureaucracy is the services all revolve around building a budget. And when you build a budget, you are trying to, uh, you know, you're, you're doing down in end of what does the air force think it needs to be a world power and be able to support all operations in the joint force that it's responsible for supporting or thinks it needs to support, you know, implied or specified task.

Lt. Gen. Kirk S. Pierce (33:04): Uh, then it has, okay, what, what do I need to do for combat and commands? Uh, not just, you know, joint force inter-operability, but what I need to do for combat commands and make the case at the combat command level and how do we support their requirements and then operate, you know, when we're working with secretary of defense and try to get through OSD, then try to get it through Congress and, you know, ONB and all the different things. So it becomes a problem when you're on this cycle that you're trying to defend what you think for the air force, it's air battles, you know, AVMs Aerobahn air battle management system. Yeah. I get tongue tied there at tab. Um, it's a challenge to articulate that all the way through and make the trade-offs and consistently stay exactly on message when we're trying to build that speed, which, uh, DOD is still learning.

Lt. Gen. Kirk S. Pierce (33:56): I mean, that's an open learning, you knew it when w when you were still in service in, in your years of support that you've done out in the commercial sector, that we're still trying to speed up how we acquire, how do we run fast and fail? Um, how do we do those things that is a little bit easier to do, because that's how a commercial company is built, but you can't fail on a no-fail mission. So, you know, you have programs or record. So there becomes points in time in that process of budgeting. Do I shift money from a program or record and hope that it's good enough for now and sustained good enough for now? And nothing happens to get to the new thing, you know, the new thing here, being a command and control architecture, that's informed with AI and ML, you know, and has that information technology backbone underneath it, which is, I mean, that's a, a prerequisite before we even get to AI and ML.

Lt. Gen. Kirk S. Pierce (34:53): So, so how do you do that when constant pressures are on trade space of Nope, you need to buy more F 30 fives, or Nope, you need to put more, you know, you need to do, uh, manpower increase or Nope. You know, the, the, the payment is going to go up X percent, um, that that's the challenge to have a consistent path forward with any program or record. Um, and especially when you're building something like, and I say, concept Jad, C2 is still a concept that has the service pieces under, as umbrella, you know, that's the umbrella concept. So that's a challenge. Yes.

Mark Valentine (35:31): Yeah. Complex, for sure. So we've kind of, uh, made the transition from talking about what AI and ML can do, uh, for the department of defense and more specifically, you know, operational actors at the operational level, whether it's for fighting or disaster relief. And we've kind of made the transition to talking about barriers that are keeping us from getting there. And you mentioned policy barriers around data sharing, uh, the budgeting process, et cetera, but there are tons of AI and ML efforts going on across the department of defense right now. Uh, when you look at some of those, what do you think are the differentiators between the ones that are, or will be successful and the ones that aren't, or will not be successful?

Lt. Gen. Kirk S. Pierce (36:16): Uh, I honestly think the ones that are more successful are tied into, um, less of talking to a staff and more of talking to operators, uh, you know, and it could be an operational staff like we're at, which is kind of in between. So I got you, I have a strategic level, and then you have operational and the tactical level, um, you need to take into account the tactical level inputs of what do I need to be able to do the job. And, uh, how does AI or ML helped me do that? And then the operational art, because, you know, like you could have something where it says, Hey, the chances of success for this is 30%. And the chance of doing something else is 80%, but something else will only meet one of your 10 objectives. Well, depending on what the, what the risk is, the risk to force the risk to the mission set that you're doing, you may choose the 30% solution.

Lt. Gen. Kirk S. Pierce (37:11): Um, and that's operational art because, you know, uh, how it's ranked and in the 10 things that you were trying to evaluate and do some kind of analysis to and red team, um, you may push one of those things further down and make a decision there. So that's where I think we could do better working. Um, and I, and I see a lot of this, uh, of commercial companies that are doing a really good job of reaching out, uh, the ones that are successful, get the tactical level inputs, understand that, but then they kind of rise up a little bit and get to the operational level. And I'll just say, typography, but, you know, like when you look at it topography, they look at, Hey, what are the mission sets that you're trying to get to? So you don't end up with the perfect thing for one user.

Lt. Gen. Kirk S. Pierce (37:57): So, you know, you end up for this perfect thing for Mark Valentine, uh, as 16 fighter pilot Xtrordinair, and it makes him the greatest ever, however, no one else in the continental United States, uh, can use that thing to the same effect. And it doesn't necessarily help out mark Valentines, you know, simulated brother in the 15 or the F twenty-two or an [inaudible] to, to, to support that mission set. So by, by working at that operational level and understanding, Hey, here are the limitations, here are the trade-offs, but Hey, what's the commander's intent. What is the end state we're trying to get to? Uh, I've seen that work pretty well. And, and again, I I'm, uh, I'm biased because of the ecosystem I live in, uh, you know, north comm has done a very good job Northern and north comm, but they've done a very good job of taking commander's intent and going, okay, what does that mean?

Lt. Gen. Kirk S. Pierce (38:53): And trying to do these, these guide, you know, these experiments to show how you break down the stove pipes and to get Homeland defense operations. Uh, and honestly, disco is a part of Homeland defense operations, because whether it's manmade or natural, the resiliency of north America is dependent on the U S to protect itself, to project forces forward, to compete. And then, you know, ultimately if we need to, to deter defeat a threat, so they've been pretty well at going, okay, how do we break down stovepipes across combat commands? How do we use artificial intelligence and do an experiment? Not a, not just a, one-off a, this look, this is what's in the realm of possible, but how do we do an experiment and then get it into DOD or a service lane to be funded, to be an actual thing. So those commercial companies that have been part of that, um, I think they have, uh, an advantage, not just because they see that, but it could be with anyone into Paycom, Yukon.

Lt. Gen. Kirk S. Pierce (39:56): It could be with any service. Um, but because they're hearing from a tactical level to an operational level, and they're getting the overall intent, it makes it a little bit easier for them to drive and sustain and go, Hey, we're going to commit to doing this. We're going to put our own research money into this. And we see benefits. You know, there's a lot of people that work in the, you know, you look at in space and there's benefits to the things that, whether it's blue origin or space X, or, you know, whichever one that's working on it, they see commercial benefits that helps with their stock. You know, there's, there's, there's people were buying their stock, they see government benefits and they see overall, how does this work into a business case? And then they can stay the course because they've worked, you know, tactical and operational level.

Mark Valentine (40:42): Right? Yeah. So this leads to one of my last questions and you kind of hinted at it already, but obviously the, the department of defense, the federal government writ large are on this campaign to learn as much as they can and sort of incorporating the benefits of AI and ML into their operations, whatever the operation might be. Uh, what can we do, uh, those of us who perhaps formerly served or now find ourselves sort of in this space, in the private tech community, what can we do to help?

Lt. Gen. Kirk S. Pierce (41:13): I think you're already doing it by reaching out like this so we can have conversations and continue to have these, uh, and I'll talk to, you know, you know, me, I'll pretty much talk to anybody, so I'll talk to any company and you'll get dumb fighter pilot one-on-one with me. Um, but the commercial engagements are very helpful because, uh, as we built, you know, the internet department defense way out ahead with the internet, I mean, basically we had developed an internal to the department defense and then it built out. And then I look at things that NASA have done. And now we look and where is NASA? They're behind on some things in the commercial environment and where's department defense, we're behind on some things, which is a good give and take. So the engagements with commercial, uh, companies and, uh, you know, a civilian vendor of, of what's in the realm of possible and some of their, uh, data and how they influence their data production and how they get that data layer and bring it to the, you know, for me to the air force, uh, that's helpful that they're investing their money into it, but, but coming forward with, you know, not just I have this solution, make it fit into your machine.

Lt. Gen. Kirk S. Pierce (42:18): Like I have these capabilities, I'm designing these capabilities. Uh, I see a lot of promise for these capabilities in the commercial sector. I think there are the same in the department defense and how do we integrate that and how do we bring it on? And, uh, you know, as a department, I, I can't speak for that. Um, but how, how do we here at first air force, um, try to test it out, try to figure out if it's it's helpful and then, you know, that leads to other things. So I think those engagements right now, because we don't have all the answers and that's helpful.

Mark Valentine (42:53): Excellent. Well, I will bring a team down to your facility as soon as possible. We'll continue this conversation, uh, and before I let you go, uh, cause I, I find your position really intriguing, uh, because when I, when I read documents now around the department of defense use of AI and ML, uh, this idea of AI ML ethics, uh, is, is being inculcated across the board. And I personally think that is extremely important. Uh, but what I find interesting and intriguing about your role is that not only are you committing a war-fighting command, but as you mentioned before, you're doing disaster response resilience, uh, kind of H a D R type activities, and you're doing it within the continental United States. So what are some of your thoughts on, uh, how do you think about adopting these technologies in the context of that, Hey, I'm working within the continental us, uh, with some very specific laws and prescriptions against them.

Lt. Gen. Kirk S. Pierce (43:49): Yeah. Well, I mean, I'll be, I'll be upfront about that. We're, we're very careful on knowing what our authorities are and, and we look at what roles and missions that we're supporting and ensuring that we know our authorities. We know what we can and can't do in, um, the United States, because it's not just here though. I mean, everybody gets anchored about, you know, specifically for me the continental United States, but for the United States writ large, but it's also our, our, our partners. So in NORAD, we have our binational partner with Canada. Um, obviously we have NATO and the Alliance and all the countries in Europe, and then we have in India, Paycom alliances. And so it, it's not the, um, we've always had an ethical bent to everything that we're trying to do. And the department of defense, uh, whether it was in Iraq or Afghanistan or, you know, or in Europe or at home at home, we have specific us laws that we have to ensure that we're compliant with.

Lt. Gen. Kirk S. Pierce (44:42): And we're always going to comply with those. Um, the helpful part with AI is by having specific discussions, and it's great that OSD, you know, off secretary defense and the joint staff are working on this, uh, that we have an ethical application. I mean, just like you and I grew up with the laws of armed conflict in what is ethical. What's not ethical. Uh, you don't always have, uh, an authority. You don't have always have a direct guidance. You may have implied guidance, but having an ethical boundary and understanding what those are is important, because it's just another ability for you to make decisions and use that, you know, education that the us air force gave us to be airman enlisted, or officers or civilians and, and make good decisions. So by having, uh, an ethical rule set that we're using with artificial intelligence and machine learning, that's just bringing really together with what, how we already operate.

Lt. Gen. Kirk S. Pierce (45:41): Uh, so in the Homeland specifically, you know, we don't do intelligence surveillance and reconnaissance. So, you know, if there, if there's a hurricane that's damaged Louisiana and we need, uh, at the request of the state to, to get imagery so they can understand what levy is broke, or, you know, what's up there. Cause the sensors there are out of power and the generator, they had set up, ran out of gasoline and everything was working until, you know, five hours ago. Um, we're going to work within the confines. So at first air force, the civil air patrol, uh, has it reporting up through our, uh, cap USAF piece here at first air force. So we use the civil air patrol and go out and they're not collecting on Americans are just using your flying overhead and, you know, reporting back and they have some EO and electro-optical, uh, capability to, to give those images to law enforcement or to the state.

Lt. Gen. Kirk S. Pierce (46:33): If we needed to do something with an air force asset that we would be using for, um, uh, really assessment of damage, then we'd go back and work the secretary of defense and make sure we have a proper use memorandum and make sure it's the right need to use that and make sure that we're not collecting things in retaining it, uh, you know, information or data. So, uh, we worked through all those things. I mean, small UAS is a great example of that, so that you ensure that you're not violating someone's freedom, but if there is a no fly zone for small UAS and people are flying in it, how do we take the data that we sense or a visual ops rate observation and give it to the local law enforcement and have them deal with it at the state or federal level. So we try to ensure all that.

Lt. Gen. Kirk S. Pierce (47:19): So by AI already having that built in, and I mean, I think about it, you give them this big chunk of data and go, Hey, here's all the things that you can do. They would have it, you know, that, that entity, I would say vague, but you know, the entity or the software, whatever is working through that. And we'll retain that way better than I ever could. Or maybe my lawyer. I mean like JAG, who's phenomenal, but you know, it's always updating, you'd always have the most current, you wouldn't be caught off guard of, oh, no, someone got published a month ago. And I didn't realize that. And Hey, there's conflicting guidance between, uh, department of defense and department of Homeland security and a state that would never happen. Right. So they could always be on that. And then it would look at like, okay, here's how we apply that.

Lt. Gen. Kirk S. Pierce (48:04): And here's your options, you know, here here's a 72 hour life or limb. Here's what you can do underneath that. Here's what you could do, uh, underneath state guidance. But if you do it under state guidance, you need to get approval from the governor or the emergency management or, or at the federal level, you got to go back to FEMA or you gotta go back to department of Homeland security, whatever the pieces, CBP, um, you know, us marshals, but it would all be laid out there and it could bring those pieces together and go, okay, then this is, here's your courses of action moving forward, which right now we do that, but it'd be nice if this was just, it's always on tap.

Mark Valentine (48:40): Sure. I, I really love that answer. And I like how you drew in the laws of armed conflict, because I do see a lot of parallels there. And obviously this is totally rhetorical because it's above your pay grade. Uh, but you know, in some ways, you know, I wonder if these states these ethical AI standards that we were building in the private sector, uh, in the department of defense for one day come together and be somewhat of a new Geneva convention, uh, related to AI. So it's, it's really a very interesting area right now. Uh, so Joel Pierce, uh, before I let you go, uh, one final question, uh, because, you know, obviously I work in the technology space now and there is always, it seems like a translation error between what the government says they want and what often we hear, uh, at the private sector level. So if you could, on the way out share with us something you want the technology community to know about your mission, uh, that perhaps they don't know so that we can better help.

Lt. Gen. Kirk S. Pierce (49:42): Yeah. That's a, that's an interesting one. There's, there's so many places to start. Um, but I appreciate it. Thank you. Thank you for the opportunity for the intelligence, for the like artificial intelligence and machine learning. And I get there's different sectors and different commercial companies that are working on different pieces inside there. Uh, the implications of what you're bringing. I mean, I see it, and I'm just a, you know, an older general officer com origin. I see this as a, uh, opportunity to, to reset fundamentally how we look at information, how we aggregate information and be able to evaluate and assess the validity of the information and, and find those very rare one-offs that may be an outlier just bad data, or it might actually be something that we would have never observed. Um, and the companies that are working on that are gonna change how we fundamentally as senior leaders make decisions, whether we make the decision to scramble fighters, uh, because a presidential temporary flight restriction has been violated, uh, and the FAA doesn't know it, but they're in communication with a tower.

Lt. Gen. Kirk S. Pierce (50:54): It's just, it hasn't worked through the system fast enough, uh, or, uh, Russian, uh, long range aviation, um, through open source has been talking about a bunch of different things and we can figure out like, oh, here's the indications or warning. They said it right out and right out in the open, they were going to fly on, you know, July 12th, 2022, because it's somebody's birthday. I don't know why. Um, but that work that you're doing, um, especially as you're coming in, in talking to the air force and talking to senior leaders, keep fighting for it to not be an add-on, but to be baked in. So a little of that will be working whether the company has a platform or they work with other companies on the platform so that the artificial intelligence is not just, okay, give me your PowerPoint slides, give me your Excel slides, give me, you know, your Google documents, whatever that the data venue is.

Lt. Gen. Kirk S. Pierce (51:44): And I'll just, you know, whip a little AI over top of it and hope for the best, um, incrementally that probably would help, but it, but it wouldn't help us to get to a senior leader. And I'm not a senior leader, you know, a combat commander, the secretary of defense, being able to assess and know how valid the data is and make a good decision better in more time and have more decision space, cause a lot a decision to not make a decision is still a decision. So it's tough when you have, you have three minutes to decide it's much easier. If I have four hours to decide now as a staff that can be painful because it's four hours of questions, but with artificial intelligence and ML, um, that would allow, Hey, we're 99% sure. Here's the factors and assumptions and your staff can tell you about all that, right?

Lt. Gen. Kirk S. Pierce (52:37): That, that this is the proper course of action or actions that you can take. Um, I think as the, as, as you, as a commercial entity or anybody, that's a transform X approaches department of defense, keep trying to fight for it to be built in. I mean, it goes back to cybersecurity, you know, those kinds of things. Uh, when we add them on to old systems, you know, you add them on to an F 16 that was designed in the seventies, produced in the eighties because, uh, cyber wasn't a thing yet. Right. It's effective. It's not optimal.

Mark Valentine (53:10): That's great

Lt. Gen. Kirk S. Pierce (53:12): If we can be there.

Mark Valentine (53:15): Oh, absolutely. Yeah. Thank you. Uh, so Joel piers, thank you so much for spending your valuable time here today. And more importantly, thank you for your service to the air force and to the United States of America. And we are not too far removed from the 20th anniversary of nine 11. So I want to acknowledge, uh, your leadership and service on that day in DC, which was one of the first days that I've met you, uh, about 20 years ago. And it was my honor to serve with you on that day, uh, and for many days thereafter. So thank you again.

Lt. Gen. Kirk S. Pierce (53:45): Well, mark, thanks. I, I appreciate it. And, uh, hopefully this is helpful for the people that watch the transform episode. I appreciate everything that all the different, you know, vendors, companies everybody's looking at, uh, it's going to make the United States more competitive commercially, and it's going to help us in competition. So thank you.

Mark Valentine (54:03): Thank you, General Pierce, and thank you to our audience. And once again, welcome to transform X. I wish we were doing this in person, but I hope to see you all soon.