Big Tech Enters Defense, Tracking AI Disruption, and Startup Success with Robert Shelton

In this episode, Robert Shelton, a retired tech industry professional, discusses his passion for artificial intelligence (AI) and its impact on various industries. He shares his background working at Microsoft and the company's role in the AI revolution. Shelton also highlights the transformation of the tech industry in the federal space, particularly in national security and intelligence. He emphasizes the importance of working with startups and small companies, as they often drive innovation and bring fresh perspectives to the market. Shelton explores the potential impact of AI on knowledge worker jobs and discusses the changing job landscape. He concludes with advice for startups, encouraging them to get their ideas out there and leverage platforms like YouTube and podcasts to gain visibility and support.

00:00 Introduction and Passion for AI
Callye Keen introduces Robert Shelton, who shares his post-retirement passion for artificial intelligence and programming, highlighting the rapid advancements in generative AI.

01:16 Shelton's Tech Background and Microsoft's AI Initiatives
Robert delves into his extensive background in technology, focusing on his time at Microsoft and their significant contributions to the AI revolution.

02:09 Microsoft's Role in Defense and Intelligence
The conversation shifts to Microsoft's proactive stance in the defense sector, especially when others stepped back, and how startups began contributing to national security.

02:59 The Transformation of the Tech Industry Towards Federal Space
Discussing the tech industry's evolution, Robert and Callye explore the impact of startups and innovation on national defense and security solutions.

11:09 The Impact of AI on the Workforce and Industry
The dialogue turns to the broader implications of AI on jobs, the workforce, and how it's changing the landscape for programmers and other knowledge workers.

29:36 Future of AI and Advice for Startups
Robert shares his perspective on the future possibilities with AI in national security and offers valuable advice for startups looking to make an impact.

Takeaways

• Artificial intelligence (AI) has the potential to revolutionize various industries, including national security, mental health, and data analysis.
• Startups and small companies play a crucial role in driving innovation and bringing fresh perspectives to the market.
• The adoption of AI may lead to changes in the job landscape, with certain career fields being reduced or eliminated while new opportunities emerge.
• The accessibility of AI tools and platforms allows startups to compete with larger companies and create impactful solutions at a fraction of the cost.
• Entrepreneurs should focus on getting their ideas out there and leveraging platforms like YouTube and podcasts to gain visibility and support.

Transcript

Speaker 1: 0:00

Welcome to the Startup Defense. My name is Callie Keene. Today I have Robert Shelton. Now Robert is happily retired now, but Robert is a great entrepreneurial thought leader and really played a very pivotal role at a large organization that I think most of us are fairly familiar with. But before we get into that, robert, what are you passionate about right now?

Speaker 2: 0:29

Yeah, even though I'm retired, I spend way too many hours every day reading about tech, and the area that I focus on, or have focused on over the last probably two years maybe three has been artificial intelligence, and generative AI has completely changed even what I thought was advanced two years ago. So I spent a bunch of time in my retirement time learning how to program in Python and doing things with artificial intelligence.

Speaker 1: 1:03

Well, that's kind of the exciting topic of the day right now injecting ai into everything. Uh, can you speak a little bit about your background, because your former employer is really pushing a lot of that ai revolution right now and just injecting it into everything. But this, this will give a good perspective of who is Robert and tech wizard, tech titan, robert Shelton.

Speaker 2: 1:34

Give us a little bit of context here. So I spent 30 years in the industry. I did everything from software development DBA to networking and various levels of management. My last job was at Microsoft, and so I worked for Microsoft in the federal industry, working specifically with national security and intelligence.

Speaker 1: 2:25

Yeah, this is great because I think I've been around technology and certainly startups long enough to remember just a applying those tech talents towards critical infrastructure, big government projects. I think everybody remembers Google's project, maven, them backing out of that. But while other people backed out, microsoft stepped up and is in a really good place right now because they're not a small fish in general but they really helped move that market and you were a critical part of that move, of providing all that tech talent and knowledge and working towards national security problems, right Applying what Microsoft has and modernizing that equipment. I know that there's a lot of very important walls and barriers of what we can and can't talk about with that, but it's just an interesting narrative in general. Can you maybe talk about that transformation that the tech industry has had? Or that kind of coming of age moment is like, hey, we can go into this federal space, like what does that even look like?

Speaker 2: 3:30

Yeah, when I started in that business it was very immature Steve Ballmer was the CEO at that time and you're right when Google backed out of Project Maven. One of the proudest moments, at least in my team's time, is when our current CEO, microsoft current CEO, satya Nadella, stepped up and said that Microsoft would absolutely support the federal government. At the time we had been selling the government Microsoft technology for a long while, but it was, you know, office and windows and some very custom solution for mission and mission enablement things that obviously I can't talk too much about. But the Microsoft has been around for a long time in the intelligence space on all fabrics.

Speaker 2: 4:24

When I started, it was all about selling licenses and trying to get Windows on every client desktop, but towards the latter part of my time at Microsoft, microsoft really started focusing on the mission. It had just put out a secret and top-secret cloud. Amazon was there, of course, first, and I believe Google and Oracle also have reached the other fabrics, but Microsoft is bringing the full bag of tricks to the defense and intel agencies. I know that, like OpenAI's, azure implementation is one of the things that the government's very interested in, from everything from the standard government with the GCC high all the way up to. You know I'm sure that it's going to roll out on top secret if it already has, and I've been gone for about a year year and a half at this point so I don't know what the current status of it is, but I know that was something the government was very interested in.

Speaker 1: 5:38

Yeah, absolutely secure LLM implementation. So it's inside of that enclave and use those tools and bake those tools in. It is really interesting because you know obviously some somebody like myself. I see all these fun tools that people use and I can't use them day to day on my computer.

Speaker 1: 6:00

So even the promise of hey down the road, being able to use some of those or even a similar tool with my work, it's very appealing. But yeah, it's been such an interesting road to see that leadership from some of the bigger companies and then I feel like that attracted startups and that attracted funding and that's kind of how we met is through working with startups and I know you have a big passion for entrepreneurship and passing lessons learned and yeah, I think that's a great analog for the entire market. Right Is technology people applying their discipline to big problems in our space national defense and then working with entrepreneurs to bring them into that mix as well. So we get that really passionate, interested, emerging startup vibe to solve things in a new way. That's what really gets me excited. I know that you share that excitement, that tech excitement. But how.

Speaker 2: 7:14

I just saw that one real quick, one of the interesting things I found once I got on the other side into SKIFs and talked to the government prior to getting cleared and getting in that space, I would have thought that the most innovation came from the large systems integrator. It turned out that I mean, if you were thinking like you know how do I deliver email to administrative assistant or some government executive, that was generally the big systems integrators that were implementing that. It was Microsoft tech, but they would win the massive you know 10-year, you know hundreds of millions of dollars. I contracted before that, but when it came to like the mission or building something that was extremely unique, it was always these 10-man firms or, you know, three-person firms or, in a lot of cases, like one or two people who got a small personal services contract and they were implementing things that, even working for Microsoft, I was like whoa, you took the technology which was meant for one thing and used it for this completely other thing, and that was the thing that was most impressive to me.

Speaker 2: 8:36

So during that period of time I understood that you know, for the mega big implementations, they were always going to go to Microsoft anyway. Right, because you know Word and you know they were really big competitors to Office and Windows. There were some along the way, but they're you know, they all kind of went away. But anything that Microsoft wanted to do that was going to be truly impactful was going to be small, probably 80-90%. So, yeah, I have great affinity for that segment of these systems in the small.

Speaker 1: 9:19

And is that a driving force of why, now that you've retired, you have kind of engaged that and worked with the startup community, accelerators, entrepreneurs?

Speaker 2: 9:31

Yeah, I mean, you know this yourself the difficulty when you're dealing with defense and intel. Everything has to be accredited and it has to go through a lot of security reviews and nothing gets a larger product. Again, microsoft actually has its own consulting. That's a, you know, sort of a. It's a part of Microsoft, but a different division than the group and what I worked with. They won contracts and that sort of thing. So nothing against the bigger guys, although Microsoft would be considered bigger, but they're bigger than a lot of the companies that I'm talking to and working with.

Speaker 2: 10:12

The timeframe between idea to pilot, proof of concept to deployment is typically in years. Often the government has the delay, but it's also the delay that's built into companies having to be concerned about stock price and the red tape and the engine's got to churn. I like the smaller companies' ability to quickly innovate and get something out there. Now, even at Microsoft, my group kind of specialized in doing that sort of thing as well innovation. But it's very rare to see that in bigger companies. The smaller companies are the ones who I think are making the big change, even in broader tech. I mean, if you look at OpenAI as compared to Google, both were doing AI, google obviously before OpenAI, but it took the small company OpenAI to move the market ahead, and I would say orders of magnitude. That's why I like small companies.

Speaker 1: 11:26

Yeah, I think we share that affinity for sure. If you want to stay abreast of the changes, even you think somehow open AI or these other competitive solutions they've maintained that speed but over time there's going to be other innovative solutions. And staying in the trenches with the startups you also get to see all the fun tech right, talk to smart people and see you know how, how they're thinking, and you get to see things maybe a couple years before everybody thinks. I mean, we've been working with ai for you know, more than half a decade like deploying it for product and and so I. It gets me really excited when everyone else is excited about things. But also, like you know, we've had people in um, in the accelerator programs with, essentially you know, different models. But what you would look at then you'd think, hey, isn't that chat gpt? And like no, it's, this is another thing. We've been deploying it in defense for, you know, five, six years in into product and that's kind of behind the scenes piece. But then now everybody's excited. You get this proliferation of all technologists jumping in and adding their piece and adding their piece and it just accelerates and snowballs from there you can hear. It gets me a little excited because you get to see something new every day. I really enjoy that piece.

Speaker 1: 13:03

I'd love to get your take because I know you are following the AI revolution very closely, so one. I'll ask you a more boring question what do you think that it's going to do for knowledge worker jobs? And then I want to get your opinion just on what are fun things that you see in the market. But I mean, it really is changing what people do, how they do it, what jobs are available. It's already changing what programmers are getting hired, and it's not even very good at programming yet. So, in general, there are Devin. You probably saw the Devin demo.

Speaker 2: 13:43

That's interesting yeah.

Speaker 1: 13:48

But how do you think that's going to impact the workforce? For you know, we're both coming out of a knowledge worker base. Like what do you think that that's going to do for the tech industry?

Speaker 2: 13:59

Yeah, you know I was having a conversation with somebody that actually they were in education and they were asking a very similar question and my response is you know, I actually posted this on LinkedIn, I think, right after the conversation about what happened to the horse and buggy industry when the automobile released and you know, a lot of jobs went away. And you know, I want to say, if I remember correctly, correctly, like hundreds of thousands of people in the coach business one way, and then it was replaced by millions of people in the art industry, right? So there's, as innovation comes out, there's always the replacement of one type of job and typically an increase of another type of job. When I was in high school, they promised you know, I remember watching movies where they had office pools of secretaries that would just say the type things you know and do you know to make duplicates and triplicates and deploy all of these? You know, no one does that, right. So I think AI is going to, interestingly enough, go towards the white collar work and I think the generative AI attitude provides making you know the conversation about blue collar relevant maybe a year or so from now. So I think it's going to drastically change. Certain career fields will go away or get severely reduced.

Speaker 2: 15:40

I think in the software industry which I came from, when I started out, people got paid reasonable incomes for, you know, fresh out of college people with college, people with IT backgrounds. And then I remember, as a manager I used to be an executive at USA Today I remember people fresh out of college demanding $100,000 and there was nothing I could do. I had to pay the $100,000. The last thing and these are typically developers the last thing I wanted them to do was write a code, because we had to now teach them how to write code professionally. It's different when you're in school, you know, when you get into a corporate environment and you have large groups of professional developers, it's sort of like apprenticeship. You put them with a senior developer. The senior developer often didn't make that much more than the junior developer or fresh developer would make.

Speaker 2: 16:36

I think with AI, I think it's going to not eliminate software development. Chairman of NVIDIA said I think it will adjust how software development will pay and I think it will flush how software development will pay and I think it will flush out a lot of the people who are not passionate. You know, people who get computer science degrees, learn how to program, really don't care about software development Right, and they come in, they do their years and then they try to go into management and become a management. I think that will change. You're going to actually have to be, because AI makes things a lot easier and will replace certain aspects.

Speaker 2: 17:17

I believe it's going to go back to when I started in the industry, which was you have to be really good, because the basic stuff the AI just did, and so what it doesn't do, you're going to need actually to be good at software development or networking or infrastructure or name any of those you know, because the AI is going to be able to do the things that journeymen are doing today for a lot more money. That's my. I think it's going to be the same thing in other spaces, one of the big ones that I'm looking at now is AI is used in mental health.

Speaker 2: 17:56

That it can do. It can provide people with real time support right now. If you want a therapist, by the time you get one online and work through the insurance. Actually get to talk to them, unless it's an emergency. You're looking at weeks right Again, especially because of the insurance. Actually get to talk to them unless it's an emergency. You're looking at weeks right Again, especially because of the insurance. You can pay $20, get on chat, gpt and get a lot of good, you know insight. It's not quite the same, but it's better than weeks. So I think that will affect, you know, that industry as well, and so many others.

Speaker 1: 18:33

Yeah, I can see this leveling up happening. We get involved in a lot of data sensor fusion or you can think of connected devices be another application of this. And there's an element, the rote work of this. I could easily see AI helping you write a Python script that does you know X, turns a camera on, sends data. But to know what to build, what technology stack to use, how to interpret the data, a lot of the connected concepts or new concepts of this.

Speaker 1: 19:13

We're quite a bit away from AI just out of the box being able to provide those higher order thinking but the execution function of that concept, so not knowing what to pitch, but then knowing how to do it, it's a massive enabler or a time saving. So maybe now a senior developer or a systems engineer can work as a senior developer or a senior developer can act like they have five junior developers on their team and be able to turn a product or a MVP very quickly. But it's not going to magically come up with an idea that somebody wants and then build all of it and then make it appear. At least not, at least not yet, and I'm like I'm not out of a job yet and like when.

Speaker 2: 20:07

Devin was pretty impressive for what it was.

Speaker 1: 20:09

It is pretty impressive. That's right. Thankfully, thankfully for me, right now, until figure really gets out there, somebody has got to turn screws and make that product a real thing and they've got to go put it in a place to make sure it all works. But it's the perspective. I. I think what you're saying is true. It's an enablement. The perspective, I think what you're saying is true. It's an enablement.

Speaker 1: 20:37

We just don't know what that job is, because I use the horse and buggy analogy frequently with people. Also, a fun bit of US history is when kerosene became more prevalent, the whole whaling industry fell out within a couple of decades. So it was the largest industry in the United States. And then kerosene comes along. You don't have to hunt whales anymore. So this large industry goes to nothing within a couple of decades, leaving entire towns gone. Right, they just closed up shop. There's villages in New England that just completely shuttered.

Speaker 1: 21:15

But we did okay, right, mostly we just didn't kill whales as much anymore. So it worked out well for everybody. In the short term. Not that great right? Uh, seemed not productive. In the long run worked out really well. We got gasoline as a byproduct of creating kerosene, which enabled us to have the, the combustion engine, which enabled us to have that automotive replace the horse and buggy. So you know, this is just how technology works, is that when it hits, it changes the world. And at the interim time, you think, hey, I have the most prosperous city in America and now it's in ruins. But another decade after that, people have just moved on, they're just doing something else.

Speaker 2: 22:10

I think that will happen here. As you were telling the story, I remember having a conversation and I can't remember her name. I ended up doing a couple of presentations with her at an industry conference. She was an executive for NGA, executive for NGA, and what she was saying and we were doing artificial intelligence. Back then, what we thought was artificial intelligence until we saw general AI and I realized that that was DOS and you know, and general AI is, like you know, a graph for user interface.

Speaker 2: 22:44

But she said something that always struck me. She said you know, we get petabytes of data a day Petabytes and are only able to look at a very small percentage of it. A lot of the rest of it gets thrown on the floor. And I asked you know, why is that? And she says well, there's not enough analysts in the world. You know the requirements you have to be in order to become an analyst and a lot of that comes from experience. When you get experience is by being a thing, and then you know doing it for a certain period of time, and so we engaged in like a proof of concept to do something for it.

Speaker 2: 23:27

But that wasn't the important part. I actually got to spend time watching analysts do their job and one of the things I found out was like a large percentage of it was getting data in one format, running scripts and massaging the data to try to get it in another format to then load it into some disparate system that does a thing that associates that data with another piece of data to which they get the output from, and then they massage that into and ultimately it ends up in Excel right, which is you know, I joke Microsoft's number one database product is Microsoft Excel. And then, once it's in Excel, an analyst can then start looking for things or sometimes it's a querying system, you know, some sort of in-depth search system but a good chunk of that time is just like trying to kick the data into a point where a human can actually read it. And AI doesn't have that particular property Not that it doesn't have other things that it can't do, but the ability to look at mass amounts of data in disparate formats without having to munch it and look for common occurrences or things that are not very common in that data.

Speaker 2: 24:44

That's going to be a game changer. So now an analyst can actually look at the summary data. It doesn't have to do all of the munging to get it to a point where they can look at it. So I think about, especially when it comes to national security. Having gone through all the training, I worry about the data that got thrown on the floor.

Speaker 2: 25:08

I mean, you saw what just happened in Russia. You know there were warnings and there are political things that go along with that, but I'm sure that their analysts got the same sort of warnings that the United States gave them and, you know, sort of like our 9-11, they just didn't rise to a level of importance because they're looking at petabytes of data a day. How do I know which one of the things to focus on? Well, ai doesn't sleep, it doesn't take coffee breaks or breakfast breaks, it doesn't argue with his wife the night before and, you know, doesn't have the focus that it should. It just does its thing, 24 by 7, and can scale up. So petabytes get looked at versus, you know, a few megabytes. So, anyway, that's when you were talking. That's the first thing that came to my mind was the conversation I had with that executive at NGA, right, yeah, and you think that's a Palantir solution?

Speaker 1: 26:07

right, it's a billion-dollar, really bespoke, super-powerful solution. It's been around for a long time, it's done some real impactful work. And I think about this often is they can't go out and say, hey, we, we did this, we did that. You know, our system prevented this, prevented that, obviously. But the power of being able to mine all that data, look at disparate data and say, hey, here's the markers of someone who's planning an attack or is going to take this action, that that's kind of getting dumbed down to the point where, yeah, now I can take an excel database, a, like you know, postgres, sql database and a bunch of word documents and smash them all together and then ask a question and get answers out of it and not really even have to figure out how, how, how does that even work, right, you know, like, take a knowledge, a knowledge base and query a knowledge base of just articles and get an actual answer out of it. And so the new generation of this AI, a bit with data, we've we've had a couple of people on the show that are doing plain language queries of different databases. So there's a couple of very interesting products out there, but of disparate data solutions and bringing it in together.

Speaker 1: 27:42

I think, man, a decade ago that was a billion dollars. Today it's something cool. I just saw in Product Hunt. That was a billion dollars Today. It's something cool. I just saw in product hunt and that's pretty frightening or pretty exciting, whichever side of the fence. I'm kind of a technology maximalist, so I have to think that I really don't need 100,000 horses in New York City. I think that we can move on beyond that, but time will tell.

Speaker 2: 28:14

When we started Palantir. I remember when they first entered the space and they were just a small, really small group of incredibly talented people Like stupid talent and they created a product that we all know is extraordinarily powerful even today. But an interesting thing happened Because of the innovation that they had, they have to pay for extraordinarily talented people. Volunteers are not inexpensive. One of the days of the government, I remember complaining to Microsoft. You know the price cap. We're expensive, microsoft's expensive, but a volunteer is like an order of magnitude more expensive. But it does what it does extraordinarily well, so well that if you were a small company 10 people your ability to even remotely get close to what they do like forget it. Two years ago I would have said game over Even a big company like Microsoft. The only way to compete with them would be to do a, but now, with their vision, capabilities that you see in some of these tools and mission and speech recognition.

Speaker 1: 29:51

I mean there are a lot of things that a small company could do that within a niche could absolutely compete with a palantir, certainly at like a 60 or 80% level of competition, but at a hundredth of the price. That's really kind of what we're seeing is $20 a month tool that I can just plug and play. Now I have a machine vision system that can I can train to recognize faces and I spent 10 minutes building it.

Speaker 2: 30:11

Without the overhead of having a PhD? Who? Who has a degree in, you know, in vision science?

Speaker 1: 30:16

exactly, you know, I just need to know how to call the api and it you know, it's just me goofing around like one afternoon and I can build that thing and it's incredibly powerful for the government because, as a government executive, I might have the ability to approve $100,000.

Speaker 2: 30:34

What are you going to do with Palantir for $100,000? You're not going to even get a phone call, yeah.

Speaker 1: 30:43

No, it's just yeah. Yeah, they're sticking to the premium model because I know that they've moved up market similarly in the past couple of years as well. But it's just interesting from a tech perspective is to see what we would consider unbelievable. 10 years we worked on license plate readers, facial recognition. We've got a couple of people in gun detection, like we've had Chad Green on here, you know, building handgun detection. You think, man, that's really cool. There's a couple people in the space just doing this. Really well, but what I could do today is I think I could build this Versus. Yeah, like you said. Like you said, get it, get a phd, get a couple of senior devs in here, go through a year of building hardware, uh, building a hardware software solution, going and doing field tests. I can just. I can just buy a board from blaze and they have that model already in their AI studio. I can just drag and drop it and have it like right now. You know what I'm saying? For nothing.

Speaker 2: 31:56

It's a crazy. And the interesting thing is, you know some people would say wait a minute. You know that's going to totally disrupt the industry and you have two ways of looking at it. Like yeah, that's like problematic. If you're problematic, if you're one of these high-end developers or a specialist where you're sought after because you're one of five people who know how to do it. But at the same time, imagine that across the entire security apparatus you had that ability and instead of costing a million dollars, it costs a hundred thousand dollars. Right, and now we can all be safer. I remember doing a presentation for TSA years ago and we were.

Speaker 2: 32:42

They were talking about the data that they get from the scanners and we were talking about proof of concept that would allow them to take that data and then correlate it to find if you and I were traveling from two different locations and planned on hijacking or blowing up a plane, and somewhere in the data there's a relationship between us that they've picked up, in that you have a part of the weapon that you know by itself is innocuous, right, you could just be taking it to a gun show. Or you know, and I have a part of a weapon, and three other people have a part of a weapon and by the end of themselves they don't throw a scam of themselves. They don't throw a scam. But if we all travel within a certain period of time going to a location and we each have a part of the weapon, the ability for the TSA to be able to identify that and then pull us off. You know this was I think Admiral Neffinger was the administrator at that time.

Speaker 1: 33:45

So this is.

Speaker 2: 33:47

I think this may be almost seven eight years ago. The ability to do that today is actually API. The ability to do it back then we were talking about putting together the proof of concept. Even proof of concept would have been a million plus dollars Right, and maybe 10% effect. So you know, it's incredible that that kind of capability may exist right now, but that you could do that now and back then it would have been a happy thing.

Speaker 1: 34:18

Wow, robert, I really appreciate you spending some time with me today Before we go. Do you have any parting thoughts like a gold nugget for our audience?

Speaker 2: 34:30

Yeah, One of the things that, especially now that I've spent a little more time working with startups, working for a large company even though for my group getting internal funding wasn't easy, but the money was there I put together a business case case and it's been a lot of cases the company says sure, here's $100,000 or $1 million, don't do anything stupid with it.

Speaker 2: 34:59

Working with some of these startups now, I realize how many great ideas could go forward if but for $100,000. I want to say especially to ideas could go forward if but for $100,000. And I want to say especially to the startups who have a good idea that I'm happy that the tools are becoming inexpensive enough for you to be able to at least have a shot at using it and to work your butt off just to get it out there. Whatever your idea is, get it in GitHub. Put together a YouTube video. You never know who is going to see that.

Speaker 2: 35:42

That YouTube video that goes viral could be the difference between you never having a business and you being a multi-billion dollar business. Especially because I work in national work in national intel and defense. I'm hoping that a lot of you will spend time on making sure that planes in the United States don't go boom. You know that's my number. One thing is to make sure that our country is secure, and I think a lot of the innovation will come from a small business. I just wish that they would continue to push forward and get their proof-of-concept pilot idea out there on platforms like YouTube and podcasts like yours. However, you need to get it out there. Get it out there so it can be seen by someone as soon as possible.

Speaker 1: 36:29

Yes, absolutely. There's more people out in the world waiting to support you than you think. Right, if you start building those relationships, you'll find that there's somebody at Microsoft that's running a program that they want to co-pilot a thing with you. There's somebody at the DoD that will give you a non-dilutive capital to help build that prototype or help field it, or there is a world of support out there that's just beyond you know, hiding in your room thinking about an idea versus getting it out there. I think that's great advice, robert. Thanks again, thank you.

Speaker 2: 37:10

This was fun.

Speaker 1: 37:15

My name is Kali Keene and this has been the Startup Defense.