How AI Can Change the Game
March 5, 2025
Watch the Video
Read the Transcript
This transcript was generated with the assistance of AI. Please report inconsistencies to comms@afa.org.
Maj. Gen. Kimberly Crider, USAF (Ret.):
Good morning. So today, as we face an increasingly complex battlespace, artificial intelligence is rapidly emerging as a decisive factor in gaining warfighter advantage in the critical domains of air and space. In the air domain, AI revolutionizes combat through autonomous systems from next generation drones to unmanned combat aircraft, allowing for faster target identification, adaptive mission planning, and real-time threat response. AI-powered algorithms can process vast amounts of sensor data, providing commanders with instant insights that sharpen decision-making and increase mission success.
In space, AI is equally transformative with threats from anti-satellite weapons and growing space debris, AI-driven systems can track, catalog and predict the movement of tens of thousands of objects in orbit. It enhances space domain awareness and space-based ISR, giving warfighters a clearer, faster picture of emerging threats and rapidly provides meaningful counter options. AI can autonomously orchestrate and reconfigure and reconstitute constellations to assure the availability of vital communications, navigations, missile warning and sensing capabilities to advantage joint warfighters around the globe, including for homeland defense.
Now, we know our competitors are also aggressively pursuing AI to dominate the air and space domain. We also know if we fail to harness AI’s full potential, we’ll risk ceding control of the very battlespaces that underpin modern warfare. Let’s turn to a short video to show us what we’re up against.
Thank you.
Well, to help us address this challenge and to talk about how AI is certainly changing the nature of warfare today, I’m pleased to be joined on the stage here, today, with some panelists that have been on the forefront of the AI conversation for some time. Major General Bucky Butow, the deputy director at DIU Mr. Jason Droege, the Chief Strategy Officer at Scale AI. And Mr. Joel Minton, the Chief Technology Officer at Google Public Services. I’m going to take a minute and turn to our panelists and allow them to introduce themselves to you. Tell them a little bit about what they do and how they see AI changing the nature of warfare in light of the threats that we face. Bucky?
Maj. Gen. Steve “Bucky” Butow:
Well, good morning. The Defense Innovation Unit, our mission is to accelerate the adoption of commercial technology at speed and scale. And there’s no finer example of the work that we do than our guests to the left because our job is really to make it easy for commercial companies, the innovators who really are doing disruptive things backed by the private investment at capital that’s driving our modern industry, and bringing them into the Department of Defense to help solve really tough problems. So thank you, both, for what you and your companies do.
Jason Droege:
Thank you.
Maj. Gen. Steve “Bucky” Butow:
And others who are here as well. So I look at the video and the video’s obviously a little bit dark, but I’ll tell you, the talent that we have and the ability within our department to adopt and apply these technologies is really paramount to modern warfighting. But I’ll stop there and pass it off to Jason.
Jason Droege:
Sure. Thanks for having me. Excited to be here. And thanks to the major general for the partnership with us on DIU, where we’re going to build some AI agentic systems to move some of the AI initiatives forward for our military. I’m Jason Droege, I’m the Chief strategy Officer for Scale AI. You can think of Scale as the first and last mile for AI. We provide data sets to the major model builders, almost all of them in the world, outside of China of course, and that helps make the models more performant. And then on the last mile side of things, we help governments and private industry take the models that the model builders build and put them to productive use in private industry. That might be to make a company more efficient. In the case of the military, it might be something more around agentic warfare, agentic governments, agentic space, using this technology to accomplish whatever mission is necessary to be accomplished.
I watched that video and I understand what you’re saying about the gravity of it, but I think this is real. I’ve been starting companies in tech since 1997, and the only time that I have felt that there is about it to be as much of a sea change as there is now, is in the ’90s when the internet started. So, I do think that this is a very pivotal moment, not just for private enterprises, not just for consumers, but for just the nature of how governments interact with each other, through policy, through military, through action. And it’s really important that we get this right. I’m excited to be a part of it. Excited to be here. I’ll stop there.
Joel Minton:
Hey everybody. My name’s Joel Minton. I lead technology for Google Public Sector, have a background in building large-scale platforms for both on the commercial side as well as for the White House. I spent about two and a half years in the White House building login.gov. So thank you for everything that you do for building software in the government. I know it’s really hard and I appreciate everything you do. At Google, we spend a lot of time, people think of us as a search company, which we are. People think of it as a consumer internet company, which we are, but we also have a massive infrastructure that we use to build those systems and that massive infrastructure from our network to our public cloud, which is Google public… Google. Oh, sorry, sorry, sorry. Oh, that was bad.
Jason Droege:
Speaking of AI.
Joel Minton:
I’m sorry. I’ll try to use G from now on. So, our public cloud infrastructure allows customers to use our compute, our AI services, as well as our infrastructure to be able to deliver these outcomes. We are a data company and we’re constantly known as the data cloud, and that’s what we do.
In terms of the video and where we are today, I see the U.S. as having amazing resources in the research area. I see us with the best AI researchers in the world. I see us with the best AI companies in the world. The Frontier data foundation models that are being built are in the U.S. and that is all great resources. We also have the infrastructure, the power, the compute, the storage, the massive accelerators. We have all of that in the U.S. to be able to give us that advantage. But there are challenges as well. From a challenge perspective, we don’t have all the data in one place that need in order to be able to build these models. We also don’t always move as fast as we need to. We don’t have the urgency that we need to. So, a lot of what I’m going to talk about today is how we can improve those things.
Maj. Gen. Kimberly Crider, USAF (Ret.):
Great, on all these points. So thanks for those great opening remarks. All right. So Bucky, let’s circle back a little bit now on the threat. You’re right, the video certainly opened with an ominous tone, but it really is meant to reinforce that there is a looming threat out there that our competitors, China in particular, are investing very heavily in artificial intelligence capabilities. They have made it very clear that they want to dominate in the AI space. They certainly have made it clear that they intend to apply AI in a variety of ways and in many ways, to try to develop an asymmetric advantage. From your perspective, where do you see our competitors? You can speak to China in particular if you like, advancing AI in warfighting, and how is that driving the priorities that you put forward and the capabilities that you prioritize that we need to be investing in?
Maj. Gen. Steve “Bucky” Butow:
Well, the first thing, which I know is a concern of every warfighter in the room is that we train and we organize, train and equip to fight the unfair fight. So everything we do, including this effort, is focused on gaining an asymmetric advantage. And this is going to be our asymmetric advantage in the information realm. So, we can’t give that up. And for most of my flying career, we mission planned on desktop systems. We shared information via media, but the vast majority of the operational strategic planning was all done using Microsoft Office, PowerPoint, Excel spreadsheets. And the problem with that, it still continues to stay. And what you do is you trap data and you’re looking at legacy old information, and it’s not really shareable with the architectures that are being put together today.
So, we know that to continue doing what we’re doing and actually have that not just the decisional advantage in having information at our fingertips at the speed of the warfighter, but also, increasing the situational awareness all the way down to the Airmen at the tactical level. That’s going to require us to shift and use live data and bring in a lot of rich content from lots of different sources. And therein lies the challenge because our systems are all federated. We’re not connected to the cloud. If you use ChatGPT or Grok or any of those AI platforms, you’re doing it on the open internet. That’s not the same thing that we’re doing for DoD. So we have to make all the data that we use for warfighting accessible and more so, we have to build APIs that allow us to tie all the critical software that we use to plan, execute, and then evaluate our missions. We need those APIs to tie in to this new infrastructure that we’re assembling for AI. But I’ll pass it off to these guys.
Maj. Gen. Kimberly Crider, USAF (Ret.):
Do you want to comment on that?
Jason Droege:
Yeah, yeah. Just real quick. Part of our partnership, part of our job is to bring best in class technology where it doesn’t exist either in the government or within Scale, and then implement it, build all the custom solutions that will service the needs of the military. So wherever those models are, anything that’s best in class commercial, it’s our job to figure out how to bring it into the government.
Joel Minton:
Yeah. And to add on to Jason and Bucky’s points there, I think the infrastructure is really the key. You need to be able to have the infrastructure to be able to run these large scale models, whether it be in the public cloud like you mentioned, Bucky, or whether it actually is in the private cloud as well. You need to be able to build the models as close to the war fighter and the service member as you can. You need to be able to build those models to allow decisions to be made lightning quick. And what you’re going to find is the country that has the highest throughput, the fastest decisions, the ability to move data around, it’s going to be a data and an AI fight. And we need to make sure that the compute and the storage and the ability to build models and run models are there in the field to be able to drive that advantage. So it’s a really interesting and opportunistic time where, now’s the time for us to build the infrastructure so we’re ready when that happens.
Maj. Gen. Kimberly Crider, USAF (Ret.):
So Jason, let’s dive a little bit deeper into one area where Scale has been particularly vocal and certainly, has raised a lot of awareness on a very important topic. Alex Wang, the CEO of Scale AI uncovered just a few months ago, was probably one of the first to point out that China was developing DeepSeek and he called it an earth-shattering model of artificial intelligence. It certainly has upended the markets and it certainly has been identified as a potential threat coming from China. Tell the audience a little bit more about DeepSeek. Give us your take on it now and what’s different about DeepSeek. How is DeepSeek potentially creating an advantage for China, if it is? And how has it shifted the landscape on AI?
Jason Droege:
Yeah, absolutely. Look, I think it’s a wake-up call and it’s a wake-up call that America… And I’ve been in private industry in Silicon Valley for a long time. I was at Uber, I was at a bunch of startups, a bunch of scale companies. And there can be a mentality sometimes, certainly in Silicon Valley that all the biggest, best technology starts here. And that’s a bias. I think what DeepSeek was a reminder of is that there is a global community that is most accentuated in China that is competing with us aggressively. I think that DeepSeek for most people, came out of nowhere. I think it was an alarm bell that went off, which is why Alex put it out. There’s a bunch of stories about how it was created. I don’t know which of those are 100% true, but if you look at how…
So, how is it changing the landscape? If you look at how we use AI today, think about how you use ChatGPT or any of these products today. You use it, you ask it questions, you ask it questions about your daily life, you ask it questions about how should I treat a cold? You ask it questions about American history. So when we looked at the model, the model clearly has a PRC bias. So if our collective memory is going into these models and everything that comes out of the models has a lens of whatever government initiatives or bias or desire there is to put forward a history that is not favorable to the U.S., I think it’s absolutely alarming. And changing the hearts and minds, as you all know way better than me, is a very, very big part of winning the war. So, if you’re laying the groundwork that Tiananmen Square maybe didn’t happen or it happened in a way that is not our current understanding, yeah, that’s really bad for us.
So, I think there are other superpowers in the world, primarily China, who want to compete and win as they’ve stated. And as we’ve clearly seen, there’s other initiatives, there’s lots of other smart people out there. And if you think about the decision making, you have effectively, an autocratic government who can push out change. And we live in a democracy. And so, our challenge is going to be if data powers the effectiveness of our models and the models power the effectiveness of our actions and our military, then we need to get access to the data. There’s not a choice. So speed is of the essence here.
And in private industry you deal with this too. I was at Uber for a long time. I started the Uber Eats business there, which scaled very, very quickly. And I’ve started a lot of companies. And everything in companies at the earliest stage is, how do you outrun your competition at the beginning because the larger companies that are there are trying to constantly squash you, but they’re slower. And so, if we think about speed as being a primary part of this game, we got to figure out a way to get access to the right data, get into the right formats, build the right tools, and get it deployed. Super, super critical.
Maj. Gen. Kimberly Crider, USAF (Ret.):
Yeah. So, Alex went one step further and he said it’s not just an AI race anymore. It’s an AI war.
Jason Droege:
Yeah, absolutely. Yeah. China is trying to change the overall landscape of the world in their favor. That is what they’re trying to do. That is war. War doesn’t always have to be bullets and missiles. It can be all of the things that we’ve talked about. Let’s take an example of how the war could play out. So, if right now, we make decisions in 30 days, just pick a topic. I’m going to be over simplistic just to use the example. We make a decision in 30 days and China can make that same decision in 29 days and you make 10 decisions per year. Well, you can catch up if you’re behind. You make a decision one day slower, that’s 3%. You compound that over a year. It’s not great, but it’s solvable.
Now, imagine if all of the collective knowledge of hundreds of years of information in the U.S. and hundreds of years of information in China is available to them and they start making decisions and analysis on every single move and every single plan on a daily basis. Now, a 3% difference gets compounded 365 times a year. So in a very, very short… I don’t know what the math is, but 3% compounded over 365 days, you’re already way behind. And we don’t want to end up in that situation. We want to be on the other side of it. And with these models like split-second decisions, even when you deploy them into weapon systems, if a drone can make a decision a split-second faster than an enemy drone, that’s a gigantic advantage, isn’t it? I’m not the expert here on that topic, but…
Maj. Gen. Steve “Bucky” Butow:
I should probably add that in the realm of great power competition, which is the lens that we look at this situation, it’s quite possible that we would never have a kinetic shot, mill to mill. But all the other instruments of national power, we talked economically, we’re constantly under attack there. Cyber informational, now AI. So we have to be aware of what our adversaries doing across all these fronts and how is that going to tip and cue us to how they’re going to fight us kinetically, if it gets to that.
Joel Minton:
Yeah. And just to add onto that a little bit, I think speed is absolutely key. Everything that Jason said, everything that Bucky said is absolutely critical. We need to make sure that our infrastructure is in the right places so we can actually make those decisions quickly, whether that’s in the edge in some remote location in INDOPACOM, whether that’s here on U.S. soil. We need to make sure that the infrastructure is where the decision needs to be made so you’re not constantly transiting data across the entire worldwide service. So really important for us to think about how the infrastructure plays with the data, plays with the models, because all of it has to work in concert.
Maj. Gen. Kimberly Crider, USAF (Ret.):
Yeah. And as you pointed out earlier, Joel, that has certainly been a challenge for some time. So let’s talk about that a little bit more. How can the DoD work with government and commercial partners to create that foundational AI data infrastructure that it needs to meet the need?
Joel Minton:
Really, I think it’s the training the government employees first that they can use it. And here’s how you use it by partnering… We were talking before about public and private partnership, bringing government employees together with private industry and actually having them sit shoulder to shoulder and actually build models together, build use cases together. One of the things that I did, not recently, was go out to Morristown, New Jersey and actually meet with the service members who are running these systems. They’re looking at these screens, they’re trying to manage these screens and manage the fight, and they have all kinds of things that are hard for them to do because they have to go through 20 button clicks to do something.
Couldn’t we make AI to allow them to streamline that process? Couldn’t we make it easier for them to do what they need to do to be able to deliver outcomes quickly? So, when I think of building these types of models, I say, start with a model that’s going to make the service member’s life easier. Start with a model that’s going to allow them to remove that 3% or that 50% of their time doing mundane work and let them spend their time doing more advanced work. That’s really important I think, if we want to have the speed that Jason’s talking about.
Jason Droege:
Yeah. Oh, sorry. I was going to say real quick, the example that I gave was assuming that both parties are effectively using AI, even a marginal difference there. If you don’t use AI and someone can make decisions 30 times faster than you-
Maj. Gen. Kimberly Crider, USAF (Ret.):
Forget it.
Jason Droege:
Big problem.
Maj. Gen. Kimberly Crider, USAF (Ret.):
Game over.
Maj. Gen. Steve “Bucky” Butow:
Just really quick for DoD on a governance structure. So, our strategic partner in the department is CDAO, that’s the geez, Critical Information and AI Office. And CDAO basically manages the governance. What the enterprise level across DoD, and what they’re doing is they’re also setting what the standards for data, working on API formats and also, very critical, is that ATOs your Authority To Operate on a network.
We have two components here, the Air Force and the Space Force. You don’t want the Air Force to get an ATO that the Navy can’t use and vice versa. And then DIU is focused on those technical solutions for the joint force. We just announced today ThunderForge, which is going to support INDOPACOM and UCOM with operational-level tools, AI tools for planning. And then down at the tactical level, we partner with in this community AFWERX, SpaceWERX and others. So we have this whole continuum across the innovation communities of interest, our DICE, to make sure that we get that speed and scale and make the lift easier for individual Airmen, Guardians and others in the joint force.
Maj. Gen. Kimberly Crider, USAF (Ret.):
So I wanted to circle back a bit on what Joel was saying there about use cases. You mentioned use cases and certainly, that’s the place to start. Oftentimes, AI gets stalled just from the very beginning because there’s not a good understanding of where to apply it, how to apply it, how am I going to use it, and does it have to be a big, big thing? Can I get started on something small? What are your thoughts on that?
Joel Minton:
Yeah. I come from the consumer internet background. And consumer internet, we’re always thinking about how to make things a little faster. How to make things a little more optimized. How to increase conversion rate to allow the users to get through the process more quickly. When I go in and I meet with service members who are working on these systems, it just becomes clear as day to me what we need to do from an AI perspective.
I actually asked these people to give me, hey, give me a list of what you think would be helpful. They came with 12 different things that they thought would be helpful to make their jobs more effective from an AI perspective that they don’t have today. And so, I went through all of them with them and not all of them were able to be done, but a good portion of them were basically like, yes, let’s do that tomorrow. And then once we do identify those things, let’s get an engineer and a product manager sitting down together with that government individual and building the model tomorrow. Nothing happens until the first line of code is written. And so, instead of spending a month, three months on working groups to figure out what’s possible or how we should think about it, let’s start writing code and actually move to where we need to be, which is speed and agility.
Maj. Gen. Kimberly Crider, USAF (Ret.):
Yeah. Jason, I want to talk a little bit about a new topic that’s been pretty current in the news and Scale has been talking about this a lot, agentic AI. So help us understand what is agentic AI, how does it differ from autonomous AI and how can the U.S., the Air Force and the Space Force in particular best take advantage of it?
Jason Droege:
Yeah, absolutely. I think we’ve talked about a little bit of it here. I’ll go into more detail. Agentic AI systems are AI applications. They’re AI systems. They can pull in new data, live data, historical data and help you make decisions, take actions, surface decisions and actions to humans, because you want to do it safely within our Air Force military. And so, an example would be, if China makes a gray zone move in the South China Sea, any piece of new data that could be interesting or relevant, but it’s not a shot fired. How do you take that move and then map it against every other known time that that happened? How do you allow an operator to ask the next question of, “Well, if they did this, what else did they do?” How does it come up with ideas? How does it then say, “Here’s potentially missing from your analysis.” Those are things that might take days or weeks that can now happen in seconds or minutes.
And if you imagine that example, very small example, applied over thousands and thousands of people, many, many times a day for every single input, every single sensor input around the world, that’s what an agentic AI system can do. It can process it all without human bias. Humans are amazing. We’re amazing filters. We have tons of knowledge, we have tons of expertise, but we all have our own blind spots, which is why we have teams of people to say, “Oh, did you think about this? Did you think about that?”
These age agentic AI systems do all that with the latest information near instantly. When you talk about autonomous systems, autonomous systems are making decisions really based on a simpler set of pre-known conditions. And while those are amazing, they’re not doing it nearly as dynamically.
Maj. Gen. Kimberly Crider, USAF (Ret.):
Yeah, good. And do you see any applications of that specifically in the Air Force or Space Force?
Jason Droege:
Some of those we’re going to be working on together. Like the example that I gave, I think is something as just a hypothetical that we’ve discussed. I don’t know that I can talk about the specific use cases necessarily. In private industry, we see this all the time. One thing I do want to point out that Joel mentioned was around speed and getting access to the data. Sometimes there’s an assumption that complicated, high-powered capabilities mean long timelines, mean difficulty. And we’re at a point where actually, you can get the best of both worlds. Which, if you can hear it in my voice, why I’m a little charged up about this and why I joined Scale is because the power of the technology is there. You can build these applications in days, weeks, months. The long pole in the tent is the change management. And if that’s what stands between us and success, then we should know that.
Maj. Gen. Kimberly Crider, USAF (Ret.):
Okay.
Maj. Gen. Steve “Bucky” Butow:
I just want to reinforce this. The culture is usually the hardest thing for us to change, especially in an institution with rich heritage like the Air Force, Space Force. And so, adoption is going to be really, really critical. And if you leave here with no other thought, you should leave here with what am I going to do today to adopt and use AI in my workflow tomorrow? And put the pressure on us and leadership, to help you get there.
You can’t show up at the 11th hour, at the fight and turn on AI and use it. You have to train with it. You have to use it every day and the information, you’re actually helping to train the model. So there’s a training piece with this as well, and it’s an input output balance, but we really need the adoption and the department of the Air Force is really leaning far forward on this because we recognize that the world’s gotten smaller. The threats can be on top of us faster, and we need that informational advantage to have that asymmetric advantage going forward. And to deter, which is really the main thing that we want to do.
Maj. Gen. Kimberly Crider, USAF (Ret.):
Bucky, are there some things that DIU is doing to try to help accelerate some of that adoption? AI applications, AI enabled systems, they can hit a valley of death just like anything else. So, as you guys build out some of these capabilities in support of the combatant commands, how do you help think through the adoption, the integration of this so that you can help with the transition?
Maj. Gen. Steve “Bucky” Butow:
Well, it’s really simple. The worst thing we can do for the companies that are helping us is not allow them to interact with the Airmen who need it. So the way that we have the greatest rate of success was actually the example that you brought up earlier, which is, we need to put our coders, our people right next to the point of need because if you get that situational understanding, that’s terrific.
I should also say that AI is going to make us super skilled at other things. If you always wanted the right code and you just don’t have the skill, AI is going to help you to do things that weren’t able to do before. So you’re going to be able to optimize and get more performance out of your day using these things. But we really have to make sure that we remove all the barriers so we can get the engineers directly in contact with the people who are going to need this capability the most.
Joel Minton:
Yeah. And when I was in government, I spent a lot of time thinking about how to influence and how to get through all of the process and regulations and everything else. And one thing that I realized really quickly is that you can just go do things and you can just go do things that are going to make a big difference and try to run it like a business in every single scenario that you can. You’re always going to run into some process or something that makes it difficult, but just run and go as fast as you can, run it like a business. And if your hand gets slapped a little bit, that happens, but you need to just go and move as fast as you can.
Jason Droege:
As it pertains to your question about use cases, I would imagine, currently, there is an enormous amount of data inside of our military, our Air Force that is not being utilized for productive purposes today. And in this world, it can almost all be used for productive purposes. If we know the state of every single satellite in our fleet, in China’s fleet, and we can analyze every single move, every single change there without assigning a person to it. It can literally just automatically run a report, run a dashboard, and someone can look at it and say, “Is this important? Is this something that I can escalate?” That’s actually technologically quite simple and extremely powerful because otherwise, you would’ve had to have a human being comb through all the information. So there is probably, even without the live information, even without all this, just the existing data that’s already there can probably push us forward dramatically.
Maj. Gen. Kimberly Crider, USAF (Ret.):
Yeah, I think that’s a really great point because I think we often get stuck. I was the Air Force chief data officer a while back, and I pushed hard on use cases as we were trying to build out the data environment. The infrastructure, absolutely we need it. It has to be accessible and secure and available and all the things that you require and scalable, but you’re not going to get that on day one. So you have to build these things in parallel, but you drive through with your use cases, with your problem sets, with the things that you really want to go solve, you figure out what is it I want to go solve? What data do I need to go solve it with? Get the best data I can get, knowing that it’s probably not going to be perfect. Start building out the models and working through it and cleaning things up as you go and adding more infrastructure as you go.
That was my experience then, and I think we made some headway doing it. I know since that time, there’s been a lot of significant improvement, but more to do. I just want to know, from an industry standpoint, is that how you guys tackle it? You can’t solve these problems all at once. Everybody would like to start with all the best data. It’s super clean, it’s all available. I got this great compute infrastructure ready go. I have all this investment, I have everything I require. But you can’t necessarily get at it that way and the threat doesn’t wait.
Joel Minton:
Yeah, great question, Kim. So, what I’ve always tried to do when I’ve led engineering teams is big vision, small bites, right? You have a big vision of here’s what the architecture looks like, here’s how I will use the cloud to bring all my data together. But you don’t have that be the end all, be all, of bringing all your data together. Have the outcome be, let’s deliver that machine learning model for the service member. Let’s deliver that machine learning model to make a difference and let’s grab all the data, but put it into that architecture that we said is the architecture we want. So big vision of the entire architecture and then small bites to deliver outcomes as quickly as you can.
Jason Droege:
Yeah. And my entire history is in private industry, and so I’ll give you an example for Uber Eats. And while food delivery is not as important as everything we’re talking about here, Uber Eats was actually the last food delivery business to launch in the world, and we ended up being, if you exclude China, the biggest or tied for the biggest. And the important thing there to this point is that being first matters, right?
Whenever we got to a market first, whenever we got to a customer first, that was maybe the equivalent of the high ground. We were setting the standard, we were establishing how the business should be run. We were the ones setting the pricing, and then anyone who came after us, it was much more expensive. Anytime we entered a market where we were a fast follow number two, it was more expensive for us than them but overcomable. If we were a distant number two, it was often unovercomable. And if we were the third, we were dead. And I’m purely talking about private market competitive dynamics. But the speed is important because we want to be faster. But it’s important because you want to get to the important strategic points first.
Maj. Gen. Steve “Bucky” Butow:
I would just say that the theme of this conference, we should be thinking about how we do this for war fighting, it’d be great if we could incorporate this into our Air Force generation cycle for deployments where, we know what digital information we use when we go to fight. When I was a line pilot, we had, hey, if you don’t have data link, you are not going to get to the fight. And we’re going to have pretty close to the point where, if you’re not AI enabled, frankly, you won’t be survivable in the fight.
So we need to really press the test on that and go and start using AI and each of these cycles and exercises and doing our tactical training back at home station and really fleshing out what needs to be. Fortunately for us, in the tactical realm, we have a lot of structured data. And so, unlocking the potential of that structured data should be first order. And as I mentioned at the beginning, what we have to run away from is our reliance on all the unstructured data that doesn’t go anywhere. PowerPoint, right? All these things. And so, our ability to adopt and use live data, use the tools and make them better, that’s going to be an asymmetric advantage that we absolutely want our Airmen to have and Guardians.
Maj. Gen. Kimberly Crider, USAF (Ret.):
Okay, so we’re coming up to the end. So this is our last point. This is the call to action moment. Speed matters and we want to be first. We got to win. So what would you tell this audience right now, today? What can they go do today, to get after this, to seize the moment and really begin to leverage AI to our advantage? Joel, we’ll start with you. So
Joel Minton:
I would start by getting on the cloud consoles today, get on there and actually talk to the war fighter and understand their needs and start writing code this week, right? Today’s what? Wednesday. So by Friday, could you actually understand a mission that you could go and do? And could you actually start building something in a cloud console this week? To Jason’s earlier point, all the tools are there, all the infrastructure is there, all of the AI models are there. Get after it this week before the week’s out, at least start that effort.
Jason Droege:
Yeah, I think mine, I would obviously echo what Joel said, but I think mine would be, use it yourself. Understand the power of these models because when you research a topic, a deep topic, I was researching a medical topic for my wife a few weeks ago, and it is remarkable if you have not done something like this, how deep you can get. You can get expertise at the level of a doctor today. So, just have that conviction yourself. And then understand in your organizations, and I don’t understand the change management processes well enough yet, but understand the friction points and start working on those friction points today to get the data into these systems because if you don’t have the access, it’s all irrelevant.
Maj. Gen. Steve “Bucky” Butow:
If you’re deploying to INDOPACOM or UCOM in the next six, 12, 18 months, we’re going to have AI tools there and you should figure out how to make use of them.
Maj. Gen. Kimberly Crider, USAF (Ret.):
Great. You heard it here today, guys. AI is changing the game. It is our game to win. Thank you so much for being part of this panel.
Jason Droege:
Thank you.
Joel Minton:
Thank you.